...making Linux just a little more fun!

January 2004 (#98):

The Mailbag

HELP WANTED : Article Ideas
Submit comments about articles, or articles themselves (after reading our guidelines) to The Editors of Linux Gazette, and technical answers and tips about Linux to The Answer Gang.

xine problem?

Mon, 08 Dec 2003 13:07:24 +0530
Vinod Tirur (v.vinodkumar from licindia.com)


I am using RH8 linux and successfully installed xine for video play. Video cds(.dat format) are functioning well with xine. But I cannot play the video files (in .dat format) copied to hard disk. The following errors shows on 'xine /home/temp/AVSEQ14.DAT'

-xine engine error
There is no available demuxer plugin to handle /home/temp/AVSEQ14.DAT.

Kindly note that same file on VCD play well in my system in xine. Please advise me.


NT server refused connection while trying to make backup from Linux Server

Fri, 12 Dec 2003 17:29:29 +0100
Imarhiagbe Maxwell (MIM from iris.ch)


I have tried several times to backup my new Linux server into an existing Windows NT server but each time I tried the connection is often refused. I install the webadmin which makes it much easier but the same problem happens

I tried with the command below from the option in webadmin:

  1. smbclient //zeus/home/remote/Abacus/Abac -A samba-domain-pw -D trans >dump.tar
  2. smbclient //terabyte_server/e
  3. rdump -0u -f server: /dev/nst0 /dump

*where server = name of NT server and /dump a directory in NT server Etc, all same stories.

Could you please advice me what to do as the Linux and NT Server are of same domain and with possible solution.

Thanks for your help.

newsfeeds for linuxgazette?

Wed, 10 Dec 2003 16:39:02 -0500
macewan (macewan from macewan.org)
Forwarded from the mirrors@ mailbox. -- Ben

rdf's? will you be using rdf formatted xml feeds at any time?

Cheers, macewan


There's an RSS feed at http://linuxgazette.net/lg.rss
Every time I try to wrap my head around RDF, my brain ends up screaming in frustration. Dictionaries with 2-dimensional keys, XML namespaces up the gazoo, etc. If you can provide a _simple_ description of the format this particular RDF file requires, and what it does that RSS doesn't do, we can provide the feed. -- Mike


Response to your Defectors article...

Thu, 11 Dec 2003 18:25:39 -0800 (PST)
Dave Bechtel (kingneutron from yahoo.com)

Re: http://linuxgazette.net/issue97/defectors2.html by Tom Brown

In Windows, each user has an entry in the Documents and Settings directory on the "C" drive.

In Windows XP, I know this is true; however, it is not true for Win98 or previous versions. Dunno about 2000 or ME, don't have experience with those.

Never login as root (the Linux equivalent of the Administrator in Windows)! Always login as yourself and use the "su" command to give yourself root privileges for specific commands.

I login as root all the time, and have only once trashed a filesystem (typed /dev/hda when I meant /dev/fd0.) You shouldn't do everything as root, but I do a lot of system maint every time I'm using linux, and it's just easier than typing sudo all the time. The biggest concern seems to be file deletion, which is easily abrogated by ONLY using 'mc' to delete files.

Better advice might be something like: "Do all your normal tasks - web browsing, listening to music, playing movies, word processing, etc - as a non-root user, and use root for system config and maintenance. DO NOT run a GUI for very long as root, it increases the chance of destabilizing the system. Wherever possible, use command-line or ncurses-based tools (like mc) when running as root."

/sbin: Programs and scripts used by system itself, and by users to administer the system.

You may want to add: "Statically compiled / Standalone binaries that don't depend on external libraries to run. In other words, critical Programs that can be run even when certain filesystems (like /usr, if it's mounted on a separate partition) are unavailable." That may be a bit technical tho.

/dev: Each "file" inside this directory represents a hardware device on the computer.

The /dev dir is full of stuff that doesn't necessarily exist on your machine, however. They put it all in there up front so the device file doesn't have to be created later if you plug one in. (New users might be confused if the do ' ls /dev -l ' and expect all those devices to really exist.)

Side note: Did you ever fix the 1GB memory and Promise problems? Recompiling the latest 2.4 kernel might give you some new options. Aside from that, I have to pass ' mem=511M ' on my 512MB AMD Duron box when booting Knoppix. Hdinstalled systems are fine w/o it. With the Promise controller you might be able to pass an "ide=" parm, or it may have a native Linux driver by now.

At the risk of offending everybody, vi.

Personally, I use jstar. ;-) It's provided by the "joe" package. All the common/major distros supply it (it's even in Mepis now, after I tweaked Warren's arm.)

Anyone who finds a console mode editor that resembles the interface of older DOS' EDIT.EXE command is welcome to send in a 2 Cent Tip. -- Heather

You can't just eject the CD-ROM as you do in Windows. You have to unmount it first.

That directly depends on whether you have a file open from the disc; the hardware will be advised to resist the user pressing the button if you're running a program or reading a file from it at the time. The noticeable difference is that looking at its directories in Explorer isn't holding the filesystem locked open. -- Heather

FYI, if you type 'eject /dev/blah' at a command prompt it will umount it for you. (Depending on fstab permissions, you might have to root-run it tho.) See 'man eject'. One of the caveats for unmounting is that no-one's current directory can be the intended ejection point (if your $PWD is /mnt/cdrom and you try to eject it, it'll probably fail. CD to another directory 1st. If eject still fails, do 'lsof|grep $mountpoint' and see who's holding it up.)

If you haven't already, I suggest you try the following:

Knoppix bootable and installable Live-CD (Debian):

BitTorrent download: http://torrent.unix-ag.uni-kl.de:6969/

Support forum:

The Next Big Thing in bootable Linux Live-CD's (we need beta testers!):

My Linuxtips page:

BTW, thanks for supporting and contributing to the "new" LG.net. Avoid the .com.

Contents above ThisLine (C)ThisYear KingNeutron Ltd.
===== Check out KNOPPIX Debian/Linux 700MB Live CD:
===== http://www.knopper.net/knoppix/index-old-en.html
Normally I trim out sig blocks entirely, but leaving that in seems the right thing to do... I've made the logical assumption that he wants the world to see his note, but folks, if you have copyright notes, please clarify our permission to publish your words (with or without your name attached, etc.) -- Heather

Thankyou for your efforts

Tue, 2 Dec 2003 00:01:45 -0800
Adam Kosmin (akosmin from nyc.rr.com)

Hello Heather,

I just wanted to send you a very sincere bit of thanks for your efforts in resisting the "CMSment" of LG. While I know that there are obviously many others who contributed to the cause, your name was mentioned to me just the other night by a friend here in NYC and from that news, I got the impression that you really drove the campaign. I am actually in the middle of reading this month's issue and love your statement regarding the CMS proposal:

"I think it's a solution to a problem we don't actually have, and "yet another slashdot" is not a unique magazine on the scene."


Yeah, I figured Slashdot's a great thing, but somebody else already does that...

I can't say I was the most annoyed about it, but I was probably one of the clearest in saying what I didn't like.

I could not agree more! Thankyou once again and keep up the great work. I hope you'll extend my gratitude to the other members of the team who helped save this incredible resource :)

I certainly shall! If you're inclined to do so yourself before I get to bouncing a copy that way, you can send mail to tag@lists.linuxgazette.net, and you'll reach the current Answer Gang.

All the best,

Adam Kosmin WindowsRefund.net

# chown -R linux.GNU world

Top notch.

Thu, 27 Nov 2003 22:32:03 +0000
Dean (dean.wilson3 from virgin.net)

Just wanted to say that Linux Gazette has been an excellent read since i stumbled on it a good three years ago. You guys do an excellent job and deserve kudos.

Good luck dealing with SSC.


[Thomas] Thanks, Dean, and thank you to all the other readers who have sent in their kind regards and thoughts over this. On behalf of all of us at LG, we are delighted.
There were far too many letters of this sort for us to dream of publishing them all. Just so you know. -- Heather

new look LG rocks

Sun, 07 Dec 2003 14:14:57 +0530
Raj Shekhar (rajshekhar from hotpop.com)

Hello TAG! I was reading through the new issue of LG and it is really nicely done. The new look is very pleasing and IMHO an example of a well balanced design.

My congratulations to Tougher, Mike and Ben.

Raj Shekhar

Strict or not strict editing

Sun, 07 Dec 2003 18:13:58 +0100
Thomas Kappler (thomas.kappler from stud.uni-karlsruhe.de)

Hello Linux Gazette!

From the Mailbag of issue 97:

If you think LG is too strict -- or not strict enough -- in its article selection, please let us know.

Please don't make it less strict! It is right at the moment, sometimes even not strict enough. Perhaps you would like to read a few thoughts about this.

What I like very much about the Linux Gazette is that it is a "real magazine" online, not like a news site that gathers seemingly random bits of information. There is a given time each week when it will be published, the structure is always the same etc. Another element of that "magazine feel" is of course the quality of the articles. Please keep up that up strictly, it is a crucial element of the success you have so far. A new issue is something that people look forward to. They won't do that if a new issue means they have to scan the articles first to find the mediocre ones.

I don't think you have to worry about information kind of "getting lost" when you reject articles. Linux has gone beyond the crititical mass concerning representation and information availability on the net. In fact, I believe it has gone so far beyond that critical mass that it is an important function of an edited magazine to help its readers sort out the good writings out of the vast amount there is available. This is exactly what strict editing does.

Keep up the great work,

Thomas Kappler

Thanks for the support.

Of course, LG is a monthly publication, not weekly. We've sometimes done it twice a month, and in the mythical future we'd like to see it published twice a month regularly, but that's not in the cards now.

-Mike Orr (aka. Sluggo)

Yes, that was just a mistake. Better monthly with high quality than twice a month, but not as good. And of course we want you to stay motivated rather than tired :)


A modest proposal

Thu, 18 Dec 2003 20:05:57 -0800
Felix Finch (felix from crowfix.com)
Original letter copied to webmasters at both linuxgazette.com and linuxgazette.net, to our editors, and to Phil Hughes

Seems to me this dispute over who should control the words:

Linux Gazette

is a real muddle. Each side's arguments seem pretty clear. It has been a monthly magazine for a long long time, and one side wants to continue that. The other side has supported it for a long long time, and wants to move into the future.

Both sides apparently have enough audience and crew to keep doing things their way.

Trimming his original down a bit... -- Heather

A modest proposal, how about both zine change their name, one to linux gazette monthly, one to linux gazette online. If you really want to get anal, make (the original sites) a CGI which chooses the order of the two links randomly, or a cron program to switch it every minute, or hour, or day. Then the contested shorter name can point to a simple page describing both, and people can bookmark either equally.

Our Gang was kind enough to leave SSC out of the reply stream. They can read it here like the rest of you.
Gentle readers, you really don't want to hear the heated portions of the responses, and some scufflings based on a few of the Gang being active among the editorial staff and so up on rather longer discussions of the topic. So if you can believe it, this is the summarized form...
Linuxgazette.org presently hosts a "portal" site, amount of connection with Linux unknown. Also note that the two-letter domains lg.com and lg.net have nothing to do with either "side" - or linux for that matter; they have their own owners and no awareness of us that I could find. Nor does that big electronics company with the L inside a big round G and a dot for their logo (LG Electronics, believe me they're far bigger than either group involved here). Any references here to these two "short" names refer to linuxgazette.com (SSC's site) or linuxgazette.net (our site) respectively. -- Heather
[Rick Moen] Felix, Mr. Hughes has made clear that he regards Linux Gazette magazine's use of any variation on "Linux Gazette" as a violation of his (alleged, phony) trademark. We've concluded that the proper way out of this is to disregard (and disarm as required) the bogus trademark claim, and just concentrate on publishing the magazine.
Thanks for your good thoughts.
[Mike Orr] A lower-tech alternative would be each site staying at their current URL (but modifying their popular name slightly) and linking to the other on the home page. We have repeatedly been willing to compromise throughout, but our e-mails are met with either intransigence or silence. If Phil changes his mind and wants to discuss further compromises, he knows where our mailbox is.
Robos spoke in favor of such a dualsite idea, assuming Phil goes for it. A few folks expressed this is... unlikely. -- Heather
[Jason Creighton] People working together for years without knowing that they disagreed on the most basic issue, that of control.
He suggested if such a "description only" site is created it may as well point at any other linux zines around, too. Which would imho be more of an almanac, or portal, than a "gazette". Now that could be confusing. -- Heather
[Tom Brown] Changing the name (and the web address) twice in so short a time can't be a good thing. Readers new and old need to find LG, not wander around looking for it in all the wrong places. A significant PR campaign would be needed to make it work. That's what makes the original proposal in this thread attractive: the original site gives people a choice. The trouble with the proposal is that egos are going to get in the way, no doubt about it. Get over that hurtle, and you have a reasonable solution. Of course, any agreement needs to be in writing so nobody has room to waffle in the future.
Musings among the Gang about conditions under which a name change might be forced, or acceptable, and likelihood of any of these, snipped. -- Heather
[Tom Brown] P.S. If this ever does go to court, maybe we can get Groklaw to cover our side of it. :-)
[Thomas Adam] ...(SSC) ...re-newed the lease for the domain... and may continue to use it... (haven't seen commercial effects on SSC's site yet)... just because he has utilised a CMS engine means nothing. All we have is inferences which one must be careful not to try and personify into 'evidence'.
[Phil] Again, we don't object to Phil operating a CMS -- or anything else, really.
[Thomas] He can do whatever he likes. lg.net is the official LG now.
[Jason] It would have been nice if it [*exactly* stating what SSC's role was to be] had been done seven years ago, in the same way that world peace would be nice.
[Thomas] It is Christmas afterall :)
Some tussles over whether verbal contract may or may not apply. Certainly it doesn't apply to most of us, many of whom have never met Phil, much less worked for him. Regarding what John Fisk passed on, perhaps; but that was strongly against becoming commercial in this sense -- he was kind enough to clarify the historical perspective when asked. -- Heather
[Jason] ...of course it was impossible to know that this would happen 7 years ago, and I don't blame John Fisk for not demanding that SSC's role be made clearer. There's simply no way he could have known.
[Mike] Not only that, but what were Fisk's alternatives? LG was on the verge of disappearing.
Interesting point, that; when we finally decided to take the zine and keep running it here, we thought that it was once more in danger of disappearing forever, at least in magazine format, and in fact, that only pressing within the space of one month would provide the continuity needed to preserve the magazine at all. -- Heather
[Rick] Phil absolutely did LG a huge and vital favour.
Honestly, I think their understanding about LG's non-commercial nature was perfectly clear back in 1996. It's just that, come 2003, Phil...
Speculation as to what he was really thinking or his motivation for choosing to apply trademark to this, snipped. -- Heather
[Rick] (As I've mentioned previously, computer geeks have a dismal record for running screaming in terror from even laughably unfounded demand letters.)
[Karl-Heinz Herrmann] Right now I'm all for making it a public issue. That way the new site and the issue gets some popularity. If LG.net is forced or at some point thinks its simply wise to move away we've some audience aware of the switch and it's not starting at a point zero with no known name or link. Right now the public opinion seems to be with us so we get the sympathy. If this changes and we come over as the stubborn ones without a case we should resolve the issue one sided by backing down -- again as publicly as possible.
[Ben Okopnik] ...although the idea has some appeal. ...As it is, we don't really need to play that game; we do have a quality 'zine, one that people obviously want to read, and the mechanism of that continues to work for us steadily, day in and day out. This is one of the major reasons that it behooves us ... just keep doing what we do... the status quo is our friend ... .
I agree that any name change by us would need to be attended by much fanfare, parades, dancing girls, and political rallies in all the major world capitals; however, at this point, I see no good reason for it and several reasons against it (including the political rallies; all the baby-kissing gets sticky.)
Ashwin M said he'd stand by what the core decides, but would prefer that we stick with producing the 'zine... -- Heather
[Ashwin M] If Phil is very persistent, just change the name to one that represents the spirit of the gazette and get on with life. Continue to deliver quality articles to the public and they won't give a s*** what the LG is named as.
...I just don't want to see LG.net becoming a turnoff to the readers in the petty quarrel between LG.com and LG.net.
flamage about what "very persistent" or "petty" means exactly, heavily doused with Halon... snipped. -- Heather
[K.-H.] ... lets get back to making LG.net. ... I would consider an ideal settlement if LG.com goes back to their non-monthly posting style they wanted at first -- then they could simply link us as a monthly edition and we can link the good articles on LG.com. Phil then will have his low work, uneditied, "anybody can post" gazette while we can edit and everybody oose what to read. Then both could even keep going as Linux gazette. As far as I can judge Phils reactions I don't have high hopes he will be agreeable to this, but who knows?

I have already seen calls for a boycott of SSC. That's stupid.

[Rick] I think I can speak for the entire staff in saying that we agree -- and have said so in numerous places including... [ SSC's forums ]. [Estimate regarding changing policy of SSC's version, snipped as speculative.]
It's our policy to do nothing at_all_ injurious to SSC's interests. We take defensive actions only, and regard SSC as our natural friends and allies. We heartily encourage others to take the same view.
In reply to Mike's comment that Phil knows how to mail us if he feels inclined to, Felix noted... -- Heather

I have had no response either, which is disappointing.

I hope my original email didn't sound like I had the answer to everything. I have no illusions about knowing more than those who are right in the middle of it. I was hoping that perhaps a suggestion from an outsider might be some good to keep both sides talking, since it was not an idea owned by the other side.

Thanks for not chewing my head off :-)

[Mike] It was an idea that nobody had proposed yet, so you can take credit for that.
Rick's for sticking to our guns 'til we turn blue; Ben and I fought for sticking with our name, and considering how long it took TLDP to decide what to do and continue to carry the magazine at all (which it is now doing), I'm for staying here for the long haul. A few expressed support for working with him much more happily if he stops claiming that his CMS is a magazine but supports one as a CMS is inclined to behave - crosslinking named threads there, etc. - possibly contingent on him renaming his site. Among the gang overall, desire to change our name for good was expressed in the form of an "if" plus some suggestions as to what names; as far as I can tell that means as a group, we'll do what we have to do... but right now, that's simply continuing to have the same great magazine at this site, linuxgazette.net.
We've gotten a lot of reader letters regarding the topic of our name and whatever its legal status might be, this month. Too many to publish, but we don't mind. Thanks, everyone, for your support - and for your barbs, your thoughts, and suggestions. Especially, thanks to those who sought us out through the twists and turns of the changeover and were happy to find us again. That'll be a lot easier now that TLDP.org points at both sites.
As for the heat that resulted among the Answer Gang... -- Heather
[K.-H.] *I'm* doing what I wanted and volunteered to do: wait for interesting questions on TAG and try to help people having more fun with linux -- I'm done with politics, feel free to decide whatever you want.
The editor gal considers putting up a sign in the TAG lounge stating "no fighting" -- but that would ruin the lighthearted banter we often see between Ben's dark glasses and the lot. Various other signs are considered, but how about the old standby:
"Making Linux just a little more fun."
Thanks, Karl, for I think you hit the nail on the head neatly. Shame the thing fell on the floor during that tussle over names. Maybe it'll stay nailed up this time.
For those of you who chimed in - heated or not - thank you for the encouragement, and your opinions, and for sticking around past any of the opinions you didn't like. We're not here because we're all the same; we're here because we all like Linux, but that covers quite a world of choices... and, for the most part, the fact that we'll have to make a few. Hopefully the kindest for a maximum number of people.
Happy New Year, everyone. -- Heather

Linux Gazette

Wed, 10 Dec 2003 09:30:08 -0500
John M. Fisk, N.D. (jf235 from email.med.yale.edu)
Question by Rick Moen (rick from linuxmafia.com)

Hi Rick, thanks for writing. I'll try to address your questions below.

Dear Dr. Fisk:

You've probably been too busy with real life to notice, but we at Linux Gazette magazine have been having some problems of late. To make a long story short, because SSC announced intentions in 3Q 2003 to effectively kill Linux Gazette by transforming it into a Slashdot-style Web discussion forum, and for several other reasons, the couple of dozen editors and staff unanimously voted to move the magazine away from SSC, to http://linuxgazette.net .

Phil Hughes at SSC, Inc. has unfortunately been extremely vindictive about this, and is retroactively asserting commercial trademark over our magazine's name, and based on that is attempting to seize our Internet domain.

It would help us a great deal if you could confirm our understanding of your intentions in August 1996. Based on your wording in issue #8, we believe your understanding was that SSC would continue to operate Linux Gazette as a free, entirely non-commercial magazine alongside its commercial offering, Linux Journal. Can you confirm this?

This is correct.

Bear with me for a moment and I'll try to provide a bit of background. I had started the Linux Gazette in early 1995, essentially as a means of learning HTML and to provide an educational/entertaining resource for other Linux enthusiasts. I had spent a bit of time lurking around various Linux related USENET groups and found that although there was a good deal of useful information there, the signal to noise ratio at times dipped pretty low.

I had no access to the internet, other than a 2400 baud dial up connection to the Vanderbilt University VAX machine. A physics grad student, Tim, offered to host it on a site that he was running, and that's where the LG got its start. Throughout the early life the LG, I was always beholden to others to the host the content.

My stated intention at that time was to start a monthly online magazine with a variety of article formats: in-depth articles, short tips-and-tricks, email correspondance, etc. It was always intended to be open (in the sense of open to all contributors), free (in the sense of beer AND speech), non-commercial, and "moderated" only in the sense that no flames or derogatory material were going to be allowed. At the time, the Linux community was pretty small (Patrick V. was just getting Slackware well established and Marc Ewing, Erik Troan, and Donny Barnes were gearing up for RedHat!) and collegial.

It was an almost instant success in terms of community interest and took very little time for several regular contributors to come forward and offer to provide monthly articles of good quality. Within a year, I was getting overwhelmed trying to provide the time to get new editions of the Linux Gazette ready and sent out to the various hosting sites.

In 1996, I received an unexpected call from Phil Hughes at SSC with an offer to take over the management of the Linux Gazette. He mentioned that he had the personnel to handle the editing and distribution aspects and that he would continue to make it available as it had been started - open, free, and non-commercial.

Phil was very pleasant and we entered into a "gentleman's agreement" that he would take over management of the Linux Gazette and I would be able to retire gracefully and continue to contribute as time allowed (which clearly, it has not...) No money was exchanged during this transaction and no documentation was created or signed -- we simply had an mutual verbal agreement.

They were good to their word and Marjorie Richardson did a wonderful job of handling the early efforts of keeping the LG going. On my part, I returned to residency in Pathology at Yale and am currently completing a fellowship in Transfusion Medicine. I was delighted to have someone interested in the LG and have greatly appreciated their years of dedication to it.

Can you also confirm that you made no agreement with SSC, Inc. to assign them any trademark?

That is correct: the verbal agreement that Phil Hughes and I entered into was that the Linux Gazette would continue as it had been started, the only change being that SSC would take over the day-to-day management of it.

I will stress that we bear absolutely no ill will towards SSC, Inc. or towards Mr. Hughes. We wish only to prevent his use of belated trademark claims to harrass the Gazette, after its departure from his site.

I'm deeply sorry to hear of this situation. Phil was a godsend and I appreciate his efforts in keeping the LG going. I will demur on comment as to whether moving to a "Slashdot Style" format is A Good Thing(tm) or not -- it certainly was not my original intention.

It's unfortunate that a complementary solution could not be amiably arrived at: a "two-site" Linux Gazette, as crazy as that sounds, wouldn't be entirely a bad idea. As long as the two sites linked to each other, I could envision having a Slashdot style site with daily chatter and such, and an ongoing monthly online magazine as a complementary site for more currated content. Just a thought...

(The situation is admittedly somewhat more confused than that, but I was going to omit the gory details.)

Again, sorry to hear about this mess. I hope that it can be resolved in an amiable fashion and that the community will support, rather than split from, whatever is decided.

I wish you the best. The grace and wisdom of God be with you.



John M Fisk, M.D.
Transfusion Medicine Fellow, Department of Laboratory Medicine
Yale University School of Medicine, New Haven, CT

For Rob and Thomas

Tue Dec 30 01:22:38 2003 -0800
Heather Stern (Linux Gazette Technical Editor)

Rob Tougher's family lost a good friend and companion this week; he would have helped us more this issue but for this. He's already done great by helping us with stylesheet improvements. Netscape and Phoenix users should have a better chance at reading LG now.

Thomas Adam, our Weekend Mechanic, has also been helping me out with preparing TAG; I'm sorry we didn't have an Answer Gang or Tips last month, for I was going to be out of town and could only prepare Mailbag. Thomas cheerfully offered to fill in for me, but it turns out that he has been ill for the last few months, and outstretched his poor health -- he was too ill to complete them in time, though he had hoped better of himself. He has pitched in such as he could across this month, but many of the Tips are last month's material. I am pleased to say that he is recovering now, but it looks like it may take awhile for him to be at his best again; the staff is encouraging him to take it easy.

Please join us in offering kind thoughts for Rob and Thomas.

This page edited and maintained by the Editors of Linux Gazette
HTML script maintained by Heather Stern of Starshine Technical Services, http://www.starshine.org/

Published in Issue 98 of Linux Gazette, January 2004

More 2 Cent Tips

See also: The Answer Gang's Knowledge Base and the LG Search Engine

About crypt function in Linux -- Easy Encryption

Tue, 7 May 2002 15:52:47 -0400
Marko Cehaja (internut from thetaworld.org)

(with minor editing by Ben Okopnik)

I have many files on my hard disk which are important and need to be encrypted. Once, I left my computer at a client's location while working on their database, and they tried to crack into it. They got in with a SuSE boot disk and installed FTPD so they could grab my files. The only reason they failed was because they didn't know how to crack a plain PostgreSQL database (which should be simple when you're logged in as root). Since then, I don't leave any important data un-encrypted in order to be ready for such incidents.

Since I often use Midnight Commander, I have automated this by creating a menu item in its menu file (~/.mc/menu) [ Note: This can also be done via MC's pull-down menu system - press "F9", select "Command", then "Menu File edit". ] The following entry will pack the entire current directory and encrypt the result with GnuPG.

See attached encryptdir-gpg.midnight-commander.txt

I've also added an entry for single file encryption:

See attached encrypt1file-gpg.midnight-commander.txt

What if the files are already encrypted? In order to save some time, I have put following into the extension file located at ~/.mc/bindings [ Note: 'F9', "Command", "Extension File edit" via the pull-down menu. ]

See attached view-gpg.midnight-commander.txt

Now, when you press F3 (view) on the file with extension *.gpg, it asks you for the passphrase and shows the contents of the file. Beware, however: if you are on the network or on a multi-user system, your memory and swap file (which contain the decrypted text) could be exposed to attackers.

One good use for this is to put all your passwords into a single file, and simply remember one passphrase. Be sure to remove the original plain text files with "wipe" or "shred" rather than "rm" to delete data in a secure manner.

Adding PHP to apache on Redhat 9.0

Sun, 30 Nov 2003 16:45:31 +0200
Raj (The LG Answer Gang)
Question by Francis Matsika (fmatsika from zdb.co.zw)

I have just installed redhat 9.0 and i am wondering how to configure or build php as an apache module

I installed apache and php using rpm's that come with redhat 9.0

[Raj] If you used RPMs it has already been installed. Create a file in your /var/www/html/ directory (say self.php), and put this lines into it
Point your browser to it, (usually by saying in your browser) and you will have everything you never wanted to know about your PHP module in front of your eyes.
[Thomas] You miss the point completely here. The question was not "how can I test php"; it was "how can I get apache to utilise it". The answer to that question is that the RPMs should have already sorted that out for you, Francis. You'll just have to enure that you:
apachectl restart
[Raj] Just another example why speed reading is bad for health :). Francis, as Thomas pointed out, if you have used RPMs for installation, then everything has already been taken care of. However, if you want to be sure that good old rpm has not cheated you out what is rightfully yours, then check for the following files.
The main workhorse for translating the php code into html: /etc/httpd/modules/libphp4.so
The file which controls how PHP behaves while interpreting php files: /etc/httpd/conf.d/php.conf
Controls quite a few other PHP features, like security, language options etc. Very well commented: /etc/php.ini

automatic backup in debian

Sat, 06 Dec 2003 13:53:41 +0530
Thomas Adam (The LG Weekend Mechanic)

I am a newbie debian user (switched from RH). I have faced a strange prob. in woody. whenever I modified any file the previous contents is backedup with a *~* sign and the modified one is saved with the actual name. like *test.kwd* will be *test.kwd* after modification, but there will be an extra *test.kwd~* also.

plz suggest me how to stop the generation of this second back-up file.

thanks in advanced

[Thomas] Well, yet again Joydeep, I am being poisoned with LOIS (Lack Of Information Syndrome) --
I am going to make an intelligent guess and assume that you are using vi (or in your case vim).
Vim will save backup files, using the caret (~) notation. You can add the following to your ~/.vimrc file:
:set nobackup
to disable it (in the configuration file the leading ':' is optional).
Of course, if you are not using vi[m] then that means we need to find an alternative solution -- something drastic I think, like:
cd / && find . -name '*~' -exec rm -f {} \;
You can then add this as a cron entry to run say every 20 minutes?
0-59/20 * * * * some_user  cd / && find . -name '*~' -exec rm -f {} \;

Bash Scripting

Sat, 7 Jun 2003 11:45:35 -0400
Faber Fedor (The LG Answer Gang)
Question by Robin Chhetri (robinchhetri from fastmail.fm)

Hi Gang,

I asked another question around one month ago and even though I could not come up with a answer I decided to come up here again.(Incidentally I tried it also in linuxquestions.org).

What would be the equivalent to the zsh script given below in bash

$echo ${${(z) $(whereis libcrypto)}[2]}

if the output of $(whereis libcrypto) is

libcrypto: libcrypto.so libcryto.a

it returns libcrypto.so only. Now I could come up with

$robin=($(whereis libcrypto)); echo ${robin[1]}

But can it be done in one go using some construct? I am not a shell guru so I wonder if it can be done!

[Faber] I must be missing something. If you simply want to print to STDOUT, try this:
$( whereis libcrypto | awk '{print $3}' )
which will print to STDOUT. If you simply must put it into a variable, then:
$robin=$(whereis libcrypto | awk '{print $3}') ; echo $robin
is the shortest way I know of, but that isn't saying much.

I need help, PLEASE!

31 May 2003 08:43:04 +0530
supreet (supreet from linux-delhi.org)
With comments by Heather Stern (the Answer Gang's Editor Gal)

Well if you need dos boot disks. visit www.bootdisk.com.

It has nice pointers to many flavors of boot or install disks for the various mswin, a notable floppy-linux or two, and the correct HOWTO out of TLDP for finding the major distros' floppy images. -- Heather

Debian's first installer CD can allow access to a prompt, as can several other distros.

Expect to need to press ALT-CTRL-F2 or another F key to leave the curses or GUI installer page and find a shell, then poke around mounting up your partition and chroot'ing in.
Some even allow starting a "rescue shell" tho beware this claimed feature does not work in all red hat flavors. -- Heather

Then you can load your boot-loader.


Help (Networking)

Wed, 9 Jul 2003 13:49:25 +0100 (BST)
Thomas Adam (The LG Weekend Mechanic)
Question by sivagnanam munusamy (searchsiva from rediffmail.com)

hello sir,

i am sivagnanam and i completed B.E.,(Computer science)in University of Madras at tamilnadu in India.

i did some projects in Linux platform. i would like to know about GRUB loader details ,NFS and DHCP.

please help me.

with regards,

Hello Sivgananam,

I've forwarded this reply to the Linux Gazette...

You have supplied me with VERY little information to go on, but I will try and help you...

GRUB (GRand Unified Bootloader), is a GNU bootloader. It offers an interactive shell in which to configure it. I actually dislike GRUB immensly and prefer the defacto "LILO" loader, although many distributions now use GRUB as their default, alas.

The FAQ can be found here:


I find the config file confusing, and terse (cf: "man tar" - perhaps the heaviest man page of them all).

NFS is Network FileSystem, and I use this extensively via PLIP (which is a Paralell Line IP), which allows you to mount another computer's FS over a network.

DHCP is a means of automagically having a machine assigned an IP address.

All of these can be found in the LG KB, at:


-- Thomas Adam

By far the most popular reason, in my opinion, is that GRUB takes nice high resolution pictures, while LILO only handles cruddy resolution of 640x480. Which probably means that its graphics work on more systems, but gosh, it'd be nice to have the option. If the code costs too much space though... maybe grub is bigger. -- Heather

Linux Hotplug for Dummies

Mon, 12 May 2003 18:57:17 -0400
Robos (the LG Answer Gang)
Question by dokhebi (dokhebi from earthlink.net)

I've tried reading all of the documentation for Linux Hotplug that I can find, and none of it seems to answer my questions; A lot of it referes to other documentation that goes off the main topic, and there is nothing that describes the format of the usb *mapfiles, or how to set up the auto-mounting of a specific device (in my case a SanDisk Cruzer) What I really want is a step-by-step configuration guide or a "Linux Hotplug for Dummies" type of book


Edward Hooper

Well, can't give you that but you reminded me that I wanted to look into this myself. Here is what I've found so far:

If you plug something into the usb port (that is what most of the time hotplug is for, in the future firewire and maybe pci hotplug will also come into the game) the kernel does something and then something else does something (in the flux, 2.5 is different than 2.4 and I really didn't get it since it didn't interest me) and then hotplug comes into the game. There are config files under /etc/hotplug (at least thats the place with debian): usbusb.distmap, usb.rc, usb.agent and usb.handmap. usb.agent and usb.rc are for hotplug itself and the interesting parts are distmap and handmap. Dunno for sure what handmap is for (devices where modprobe doesn't know what to do?), but in distmap are the devices and their name equivalents in the form idVendor idProduct and so forth. I have a usb memory stick which gets handled like this: scsi emulation needs to be there (sd_mod) usb-core needs to be there (obviously) usb-storage needs to be loaded (since the stick is storage) and maybe the right fs modules like vfat, msdos or whatever then the thing gets mounted like this mount -t auto /dev/sda1 /mnt/usbstick

OK, now to put this into the scripts hotplug can handle: I searched google and found some german LUG that had written some scripts already. I adapted this for my use. Here it is. Save the file as /etc/hotplug/usb/usb-storage:

See attached usb-storage.sh.txt

Now for a little (newbie-ish) explanation. The set >> thing writes the variables that get set by the kernel thingy to /tmp/usb-storage-variables. Take a look into that,




are important and maybe



The echo marks simply where one insert ends and the next starts. If the variable ACTION is add the next case is triggered which checks if the product id is c76/5/100. This is my usbstick. Since I want to have this script also for my girlfriends stick I need to destinguish between them. If it is my stick ($PRODUCT is c76/5/100) then it gets mounted.

If you unplug it the script should be run again if I understand it correctly, but this doesn't seem to happen. ACTION=remove never appears.

I want to make it like this later on: a usb-hub connected to our server (in my apartment-sharing community) where you plug in your usb stick upon entrance. Then this gets mounted and the other computer you turn on mounts that from the server and loads e.g. mozillas bookmarks from the stick. If you trigger a serial mouse button (very handy that they have 3 buttons mostly - 3 ppl in the community) the corresponding usb stick should then get unmounted and you can leave the house.

Not a book Edward, but I hope sufficient for the beginning.

Cheers Robos

He did write me this back. I keep it in the TAG for now, maybe someone knows something here (although I think most of the ppl here rather like doing things by hand) Quote Edward:

I've been able to scope this much out, but I don't like the idea of hard coding the device (/dev/sda1) because I can't be sure that it will always be that device. What if this is the second usb-storage device I plug in? What if I decide to add more SCSI devices (my main system has an Adaptec AHA-2940UW with 7 devices already) and I don't change the script? Is there a way to get the device node from the known info? In other words, is it encoded in /proc/bus/usb/xxx/yyy and I can extract it?

Well, it says in the docs (linux-hotplug.sf.net or something) that device info (in $DEVICE) is in 2.5....

Now that 2.6 has released, The Wonderful World of Linux 2.6 in this issue mentions how hotplug has been improved. -- Heather

I think the /proc/ stuff can be used for assigning to the right device, but sort of hard to do with only one stick. Later my flat mate comes back and I can test with his stick too, then I can tell you more.

I don't want to say this, but for all of Microsofts many faults, they can at least auto mount without having to go through all of this shit. I think it's time for the kernal hackers to wake up and smell the coffee, because Linux will not be ready for the desktop until it can do want M$ can do, but better, faster, and cheaper.

If you rely on the kernel hackers there you are sort of lost. They are already having fun hacking the kernel, they don't need anything more. And this is no longer a kernel-space thing: the kernel informs you already where and what (if we think /proc is the right thing) but there are simply not the userspace tools. The usb-stick phenomenon is just coming up really fast right now ...

Actually, this tip escaped proper mention a few months ago, and I dug it out of storage. However it seems he is correct; flash chips, memory sticks, and "thumb drives" have become amazingly popular. -- Heather

...and up to now this wasn't necessary. As for scsi hotplug, this is simply rare.

I will mail later if I figured something out.

Cheers Robos

Thanks for the info so far,

Edward Hooper

Covert HTML attachments to PDF at mail server

Wed, 03 Dec 2003 16:54:38 +0800
Thomas Adam (The LG Weekend Mechanic)
Question by Steven Lockhart (linuxoz from yahoo.com.au)

For security reasons I would like to convert all incoming mail attachments from htm/html to pdf. ( a procmail, ghostscript, ripmime, altermime ) or whatever it takes solution.

I am sure a lot off mail server sysops would love to do this.

Any ideas would be appreciated

[Thomas] For security reasons (and sanity reasons for that matter), you should really be sending all MIME-encoded e-mail to /dev/null. It is more likely that if you were to do a conversion that you'd only be translating spam into Yet Another Medium (tm), so why bother?
Doesn't mean we won't answer him, mind you. I chime in here to warn that just because HTML to PDF tranlators currently ditch javascript bugs and various browser-bug exploits, that someday the PDF internal scripting language may grow fancy enough to do some pretty strange things. Luckily at the moment, PDF's a pretty nice "destination = paper" kind of format. -- Heather
[Thomas] Yep -- again this depends on which filtering tool you would like to use. Personally, I find procmail the easiest, but that might not necessarily suit your needs.

I am sure a lot off mail server sysops would love to do this.

Nope, not me. Here's one possible solution (note: not tested).
# Is it html?
* ^Content-Type:.*html
| html2pdf /some_location/yum.pdf
Probably you want to use some tricks to create a reasonable filename that won't be overwritten constantly (if there's only one, message ID ought to do, with .pdf tacked on the end) and you might need to use a MIME decoder to pull off the portion which is the actual HTML portion. After that, the trick should be usable for other "readable" attachment types you feel like converting for sanity, e.g. quoted-printable to your local character set, etc. You'll need to handle each file format seperately, unless you have a converter that's smart enough to figure out the filetype on its own. -- Heather

Securing a dial in?

Sat, 28 Jun 2003 22:15:44 -0500 (COT)
John Karns (the LG Answer Gang)
Question by George Morgan (George_Morgan from sra.com)

Hello answer guy,

I need to be able to secure an external modem that has been connect to a Solaris box to protect against unauthorized calls.. What I mean is that I want to be able to allow people to connect to the box based purely on the phone number they are calling from. Is there a way on the modem to only allow certain calls to go through while rejecting all other calls?

See the "mgetty" open source pkg (http://alpha.greenie.net/mgetty). It offers this capability, provided that your modem line has caller id. The pkg includes pretty good documentation as well as good example cfg files.

linux printing

Mon, 15 Dec 2003 01:22:56 -0800 (PST)
Heather Stern (The Answer Gang's Editor Gal)
sanjoy patra (sanjoy_cal from yahoo.com)

i have one redhat 8.0 linux server and hp1200 laser printer is attached with this.i want print from other linux box on same network.printer is printing from server ok and also it is printing from windows m/c on same lan but it is not printing from any linux box

[Heather] First thing, thanks for writing to The Answer Gang.
Secondly, you may not be aware of the resource, but there's this great site about printing support for Linux called linuxprinting.org. You've gotten you printer working for at least a couple of machines though, so perhaps you've already been by there. If not, check it out :)
So that leaves the nitty gritty good stuff. I'm guessing the topology is like this, but if it's different, please describe how things are really connected:
 printer[hp1200] ==/parallel/== RH8 --------- other linux boxes
                                windows box
Usually in order to be a printer client, Linux boxes have to be running something small to make the connection. lpr with their only printer pointed at the big RH box as a remote printer, for example.
rlpr might be a fast thing to try on one of them.
Please advise if you're using CUPS or lprng. While the general answer (make sure the clients know who the printer is and have the protocol loaded) is the same, the details of the answer are quite different. And the docs at linuxprinting.org may be handy indeed.

Loads of Linux Links

Mon, 29 Dec 2003 11:41:25 -0800 (PST)
Barbara E. Irwin (beirwin from shaw.ca)

I am one of the contributors for the Loads of Linux Links project (http://loll.sourceforge.net/linux/links/index.html).

FYI, this is a GPLed database of 4000+ subject-classified, searchable, and important Linux and Open Source links for all levels of Linux users. It was originally a project started for the Victoria Linux Users' Group and is now hosted by SourceForge.


How can you find the correct /dev/hd?? for the C: drive

Thu, 04 Dec 2003 17:18:39 -0500
Neil Youngman (The LG Answer Gang)
Question by Michael Brusnahan (michael from commsoftware.net)


The install for mulinux cannot find the download files. After the boot command there is a message about /dev/hda?. My pc has two hard drives with 4 drive letters C, D, E, F C, E, F are 2 gig partitions on 1 drive D is a 4 gig drive

The boot command cannot find the C:\mulinux directory with the install files. It seems the C drive is not /dev/hda1. I have tried hda2, hdb1, hdb2 etc.

Would you know how I can find out the correct /dev/hd?? for the PC?


-- mb

[Neil] Normally it would be /dev/hda1, however if it is a scsi disk, it could be /dev/sda1.
You may also find an answer at
We don't have enough information to give a really helpful answer, see http://linuxgazette.net/tag/ask-the-gang.html for hints on how to help us help you.

Combining multiple PDFs into one

Thu, 5 Jun 2003 21:41:52 -0600
Jason Creighton (The LG Answer Gang)

#convert ps files to a pdf file
system $GS, $GS_ARGS, $filelist
and die "Problem combining files!\n";

This did not work no way, no how. I kept getting "/undefinedfilename" from GS no matter how I quoted it (and I used every method I found in the Perl Bookshelf).

Hm. I didn't try it, but -

perl -we'$a="ls"; $b="-l"; $c="Docs"; system $a, $b, $c and die "Fooey!\n"'

That works fine. I wonder what "gs"s hangup was. Oh, well - you got it going, anyway. I guess there's not much of a security issue in handing it to "sh -c" instead of execvp()ing it in this case: the perms will take care of all that.

The problem is probably that $GS_ARGS and $filelist are both strings with multiple arguments in them: Since you're calling system() with more that one arg, (bypassing /bin/sh) splitting the command line into words is your responsibility.

Oh, duh. Thanks, Jason; my brain musta been out on loan that day. You're right, of course.

# Convert string to list - TMTOWTDI, of course...
@a = split / /, "$GS -whatever -long -argument -string -follows";
system @a and die "Famine, sword, and fire! - $?\n";
If this isn't quite clear, some help from a handy script might work:
#! /usr/bin/env ruby

puts ARGV.inspect
(Actually, I could have witten this as "p ARGV", with does the same thing that I wrote here.)
Or, for those of you watching at home who haven't installed Ruby yet:
#! /usr/bin/env python

import sys
print sys.argv
Anyway, here's the script in action: (Actually, this isn't either of those scripts in action: It's a C version I wrote as a programming exercise. But it does the same thing.)
~$ putargs $(date)
[ "putargs", "Thu", "Jun", "5", "21:23:14", "MDT", "2003" ]
~$ putargs "$(date)"
[ "putargs", "Thu Jun  5 21:23:17 MDT 2003" ]
~$ putargs ls filename with spaces
[ "putargs", "ls", "filename", "with", "spaces" ]
~$ putargs ls 'filename with spaces'
[ "putargs", "ls", "filename with spaces" ]
~$ ls filename with spaces
ls: filename: No such file or directory
ls: with: No such file or directory
ls: spaces: No such file or directory
~$ ls 'filename with spaces'
ls: filename with spaces: No such file or directory
You see? The caller of a program does the splitting. And that means that if you want to call other programs, you have to do it too, or else your script will crash if somebody tries to use it on a filename with a character in it that the shell considers special, like a space or a star, because system() with one arg is calling /bin/sh behind your back.
So that's the main reason for bypassing /bin/sh: It gives you more flexibility with filenames: The "security risk" Ben keeps warning everyone about isn't, IMO, really the issue. It's just good style not to leave that sort of thing to the shell. Especially if your program is setuid/setgid.
Jason Creighton

what is "no"? (missing prog for binutils compile)

Thu, 29 May 2003 15:53:14 -0700
Faber Fedor, Karl-Heinz Herrmann (the LG Answer Gang)
Question by Jim Dennis (the LG Answer Guy)
with note by Alan Modra (binutils developer)

[K.-H.] Now -- it's an unbearable situation that my Linux doesn't know "no".... But I've no idea what it is and you can imagine that a google for "no" even with linux and some other keywords around are not very helpful.

Any ideas?

[Faber] Maybe it's called "nein" on your computer? <grvf>
I can't find a "no" on my Red Hat 8 box either.
[JimD] I think /usr/bin/no was (would be) a counterpart to the old /usr/bin/yes command:

See attached no.sh.txt

... so "no" could just be an alias or script that calls /usr/bin/yes with the "no" argument:
		 /usr/bin/yes no
I realize this sounds silly and stupid, and April 1st is long past for this year. But I'm not kidding. That Makefile (or whatever) seems to actually want to pipe an endless stream of "n" or "no" lines into some other process. (/usr/bin/yes was traditionally used in a pipeline with fsck to automate the process of repairing a filesystem that need lots of work -- then they just added the -y option to the GNU/Linux versions of fsck.

[K.-H.] I put the question up with bugreports for binutils and got:

[Alan Modra] /bin/sh: no: command not found
This is a result of binutils being stuck on using old buggy autoconf. Install a new version of GNU gettext, or configure with --disable-nls.
[K.-H.] Got a new gettext which includes some "no"'s
khh > find ./ -name "no*"
unfortunately with a new gettext (gettext-0.12.1.tar.gz) and nls enabled I get a linker error for some gettext symbol. The solution without nls works for getting binutils compiled.
He said he'd try it on the new kernels too, but we're not sure what his results were. -- Heather

Odd Behavior When First Launching Netscape Version 4.80

Sat, 7 Jun 2003 15:33:49 -0500
Chris Gianakopoulos (The LG Answer Gang)

Hello Gang,

I am now running the SuSE 8.2 Linux distribution on my main system. I use Netscape 4.80 for browsing my machine (locally) using Apache 1.3.27.

When first launching Netscape, followed by pressing the "Home" button, I would get a message indicating that a connection was being established to localhost. I would wait for a long time, cancel the connection attempt, try connecting to my host name (saturn in this case), and then connect to local host. Then the connection occurred immediately.

I searched the Web and found that something called asyncronous domain lookups are automatically enabled in Netscape 4.80 by default. Disabling this feature fixes my problem. Here is what I did.

edit the file, /usr/X11R6/bin/communicator, and place the following line in the file:


I changed the environment variable in the above file rather than my .bashrc because I wanted this to take effect systemwide (and via the various tool buttons associated with my window manager).

So, this might be taken as a 2 cent tip.

Chris Gianakopoulos

searching PDFs made from faxes

Tue, 01 Jul 2003 22:25:52 +0200 (CEST)
Karl-Heinz Herrmann (The LG Answer Gang)

Is anyone aware of a way to search PDF files that were created from faxes, e.g. tiff files?

I'm guessing that OCR has to be utilized here, right? I've come across things like pdftotext, but the fact that the PDF started life as a TIFF is, I think, a complication.

For the record, I'm putting together a fax server solution for a client. The ability to search the faxes for text strings would be killer.

your guess is quite right -- if the pdf contains only a large graphic and no actual text you would need ocr. gocr: http://jOCR.sourceforge.net

or claraocr:

...might come in handy (gocr seems already trained while clara ocr is a quite different method). gocr produced reasonable results for me already 1 or 2 years back. BUT: I had clean 300dpi scans. From a jagged looking Fax..... I guess you are facing serious problems.


PerlHoo rocks

Thu, 11 Dec 2003 15:07:49 -0800
Rick Moen (the LG Answer Gang)

Some folks will have noticed me referring people to flat ASCII files I've squirreled away over the years on my Web server, usually inside http://linuxmafia.com/~rick/linux-info . While useful, this collection has always been (1) butt-ugly and (2) disorganised.

I've long realised I needed some sort of proper Web framework for all that material, and Rob Tougher's work updating the Gazette's HTML showed me how much improvement the addition of cascading stylesheets (CSS) can bring with only modest effort[1]. All of these thoughts came together when I ran across PerlHoo, a Yahoo-like Web directory system implemented in two simple Perl CGI scripts.

Please see: Description by author Jonathan Eisenzopf <eisen@pobox.com> in his series of three articles at Mother of Perl, http://www.webreference.com/perl/tutorial (recommended reading).

PerlHoo is simple, malleable, lightweight, fast (up to some thousands of documents per directory), and can point to URLs on or off your system. Its design limitations are:

If you need those things, there's a follow-on called PHPhoo. Personally, neither wanted nor needed them, and PerlHoo's exactly right for my needs.

There were two minor problems with Eisenzopf's design, as I found it in his most-recent (v. 1.1) tarball:

  1. Sucky URLs. PerlHoo indexes show up at CGI-synthesised virtual directory locations, e.g., http://linuxmafia.com/cgi-bin/perlhoo.pl/Apps for the Apps directory of PerlHoo's document tree. Finding a way to substitute something shorter for the "cgi-bin/perloo.pl" portion of those URLs would fix several things at once:

Fixing this required use of Apache mod_rewrite to make the undesirable patch element disappear, and a tiny bit of surgery on PerlHoo itself.

  1. Outdated and somewhat broken HTML. Eisenzopf's CGI-generated pages lack SGML DTDs, closing "body" and "html" tags, and the required "ul" pair to go with its use of "li" elements. The page relies upon setting specific colours by their hexadecimal identities, rather than using CSS. It also incorrectly used a nested "p" and "h3" structure to attempt physical markup. I've fixed all of these things, so that pages generated by perlhoo.pl are now CSS-oriented and pass the W3C validator as HTML 4.01 Transitional.

Just so other people don't have to reinvent those particular wheels, I've posted my modified and documented version of PerlHoo at http://linuxmafia.com/pub/linux/apps/ . The tarball includes full instructions on how to configure Apache, including mod_write .

My PerlHoo instance, "Linuxmafia Knowledgebase", can now be found at http://linuxmafia.com/kb .

To answer the other obvious question: Why, yes, of course I've gotten Ben Okopnik hooked. I'm no dummy! Ben says he's hacked PerlHoo separately to support individual stylesheets for each directory of PerlHoo's index, but I've not yet seen the results.

[1] One difference being that Rob has graphical design talent. I'm certainly not trying to denigrate Rob's excellent work.

previewing and printing text files

Tue, 17 Jun 2003 20:59:45 -0400
Ben Okopnik (LG Answer Gang)

There are lots of times when I want to print a text file but don't want to do it in the Default Ugly Courier font, or want to shrink, say, a page and a half down to a single page - and word processors are just too much bother for a simple font/size tweak. Here's a script that lets you pick font/size combinations until you're satisfied with what you see, although the fonts are restricted to the AFM types that come with "enscript" (40 options, more can be added.) I find them to be more than adequate for my purposes.

See attached preview.bash.txt

Ben Okopnik

If you can scare your OS tell someone who can fix it

Sat, 20 Dec 2003 09:54:00 -0500
Ben Okopnik (the LG Answer Gang)

Here is a non emergency head scratcher for you.

I was wondering if anyone has any experience with the Sandisk 256mb USB flash drive and Linux. I got the drive to work once, and ever since it causes a kernel panic. This is not limited by distributions or hardware, as I have had it work once on a gateway laptop, then ever after it would crash any linux machine I plugged it into.

That's pretty impressive, in a bad way. :) As you know, it's usually pretty hard to crash Linux. On the other hand, USB is a special case - for my money, all of its related problems come from the fact that it brings user-space uncomfortably close to what was always a root-priv operation (hardware usage, control over ports, etc.) One of the major reasons for that separation has always been the security truism of "physical access equals root access" - and with USB, all our noses are being rubbed in that lesson. However, everybody seems to desperately want the stuff, so...

(USB is a nifty technical hack, and serial is way outdated - but giving the average user the ability to effectively fiddle with the guts of the machine is Not A Good Idea, IMO. Ditto PCMCIA - and note that there were a few years when it had pretty much the same problems that USB is going through now. Surprise, surprise...)

Not that I don't think it's possible to get it right, mind you; it's just going to take a bit of finer slicing and dicing than (in my opinion) we've managed so far... USB under Linux is still highly fragile, although it can mostly be made to work. Note that I'm talking about the stuff out on the bleeding edge, things that most people don't use: the more common run of things (printers, mice, keyboards, etc.) is pretty stable at this point.

However, there are people out there trying to resolve this fragility, and achieving measurable gains. The important part here is that they need your help - that is, the help of anyone who runs into a problem with USB. They particularly (I'm projecting, but with a reasonable degree of certainty) love to hear from folks like you, who have found what I call a "reliable problem" - this is the core of what makes Linux work, and why it is generally as bug-free as it has a reputation for. You have a problem, notify the maintainers about it, they fix the problem, you give them feedback - and that problem is gone. Repeat until the code is washed sparkling-clean, and that chunk of Linux is bullet-proof.

SO, to wrap up this long-winded expository, please report this problem to the appropriate maintainer. You can save a bit of time, and ease their work tremendously, by running "ksymoops" with the copy of "/var/log/kern.log" that contains the "Oops" from the crash and the current copy of "/var/log/ksymoops/[whatever].ksyms", and sending them the results - or at least including those two files in the problem report.

As I said, this is a general question, I recommend not using Sandisk to my Linux using companions.

I think it would be better all around if you were to help the maintainers resolve the problem instead. :)

This page edited and maintained by the Editors of Linux Gazette
HTML script maintained by Heather Stern of Starshine Technical Services, http://www.starshine.org/

Published in Issue 98 of Linux Gazette, January 2004

The Answer Gang


¶: Greetings From Heather Stern
(?)Apache LAN woes.
(?)Preventing Shutdown if NFS Shares Are Open
(?)SSH tunneling
(?)about windows and linux...
(?)checking hard-disk in debian
(?)question about env vars
(?)temporary environment changes

(¶) Greetings from Heather Stern

Greetings, gentle readers -- welcome to a new year here at Linux Gazette.

Congrats on finding the world of the Linux Gazette Answer Gang. If you never felt lost finding it this last month or three, thank you - ever so much! - for your perseverence.

For those among the Answer Gang whose names were lost when we couldn't retrieve the old list - Glad to have you back! The signup list is at:

If you're new to this magazine, welcome doubled. Have some hot chocolate and a few software packages. Pull up a chair. Hang out and share.

To catch everyone up to speed, the Peeve Of The Month refers to the most common reason, statistically, the querents did not get answered or didn't like the answer they got... expressed as whatever peeve of ours they crossed so's to make them lose their TAG lotto ticket.

It in so sense has much to do withas much to do with the toasty crispness we bring our marshmallows to while roasting our querent's ability to form a good question when OF COURSE :) they dunno the answer already...

At the moment, "statistically not getting an answer" and "peeving the gang" get different results. Statistically the biggest reason for not seeing your answer last month might be if you were still looking at the other site. Just to make it clear, if you like the style of the other site, visit both of us. Plenty of Linux to go around. But I suspect you'l find us... ahem a little more fun.

Statistically this month the reason went back to simply not providing enough information for us to figure out what your question was. With "not linux" being a close second. Honorable mention for the company who wants us to take over their "answering service" ... errr, we don't do general phones, linux based PBX or not.

For those whose question made their way to us - and it isn't as bad as we feared folks, we got 460 mails in November, and a little less this time but Christmas time is always light.

Now, I must apologize, Most of this is the Blurb I wanted you folks to read in December -- and I daresay the condition hasn't changed much. But I will top off with thoughts for the New Year, as well. (For why we missed December, please see the Mailbag.)

Now, we've got a new thing to annoy the heck out of us - after we start answering - people changing the subject line when the topic hasn't changed! One fellow not only did this almost every single message, but also was replying singly rather than to the group. We can't gang up on problems like that. No single one of us - even the grizzled among us - are experts at everything. (You want proof? see the SSH thread, and some of the Tips this month.) If you don't like the topic you picked at first, tell us inside the message. That's why we have an editorial staff, so we can do stuff like that to the message and make it easy to read. But make it easier for the folks who *have* decided to help to stay on your thread. *sigh*

Chanukah and Christmas both passed by and I've still mostly no idea what to get my geek friends that apt-get isn't already halfway to downloading. (Or urpmi, if they're Mandrake fans. Thanks to one of the Gang for that tip.) They buy parts for their computers faster than I do, anyway. Maybe they'd like some nice parchment editions of the GNU, artistic, perl, MIT, and a few other licenses to hang on their wall. Jim's mom found a great present though - a polo shirt with #! as its logo. Not only that, but I think that ThinkGeek has stopped offering them...

There's one they'll want to steer well clear of, except of course for the ones who love talking politics and law (and perhaps other things one doesn't wish to watch being made). But, if you want a good laugh - a good chuckling belly laugh - and maybe some better understanding of what's going on in the SCO case, you have got to read the Groklaw site. I laughed out loud just reading the "Why Groklaw" interview; who couldn't laugh at "SCO Falls Downstairs, Hitting Its Head on Every Stair" even just as a title. This is from someone who just has a lawyer friend with a blog; she claims no special talent in law, sysadmin tasks, nor coding. Just "the person in the small law firm who knows enough about computers" to get by. I know you won't believe me, but we all had to start somewhere. Hanging out with lawyers gives he an ear for hoping to translate it... and I agree with her - the hunger to actually undertand what the heck is going on with all these court cases is real. Specific to SCO, these threads are good too. Better yet they're not all silly, tho one of these is: http://www.groklaw.net/article.php?story=20031119041719640 http://www.groklaw.net/article.php?story=20031106164630915

And then there's what Netcraft had to say about it. Make sure your ribs are all in good order first - they're gonna ache from laughter - and set your mind to 7 bit ascii:

For balance, here are some more serious points to consider. I'm sure in the case of the GNU philosophy [http://www.gnu.org/philosophy/] we're singing to the choir, but once curious, may as well sate your thirst:

These do have pointers to other sites as well.

Ahhhh... philosophy. My December was a rather rocky time, full of both glad things and sad things, troubles and hope. So I think my lesson for the new year is about choices.

You have to make your own.

In the sense of Linux, there really are a great many. For at least a couple of years there have been more varieties than you can shake a stick at. And you know what? They're getting pretty good.

So before you go picking out a distribution, don't just look at what your geeky pal tells you is the best. Certainly he or she has spent some time discovering that for themself. Your needs, however - may vary. Maybe you write all your friends who don't have computers - then printing and its troubles will be important to you, maybe scanning too so they can see the silly things your cat is up to. Need to boot from almost anywhere but don't need much of a console? Maybe cramming a tiny distribution on one of those USB thumb drives would be the thing. Or whatever. Don't want to figure out all these scary things, just wanna surf? Well heck. Try Knoppix.

As a last note - the holiday season's a crazy time (at least here it is). Drive safe. Pay attention to people around you and what you're doing. If it's a time to be thinking of peace, think how best to keep that peace - and if the bricks fly, to defend it in a way still consistent with your own ethics.

Happy yuletide.

(?) Apache LAN woes.

From Dave Hope

Answered By: Jason Creighton, Benjamin Okopnik

Hello all,

Well, here goes, strange, I feel shy writing an e-mail, I suppose there's a first for everything... Anyway, I have a VERY basic LAN setup at home, so basic I should be ashamed to call it one.

(!) [Jason] Hey, that's why it's called a Local Area Network: It's local! If you have at least 2 computers talking to each other, you've got a LAN.

(?) Anyway I decided it was finally time to remove Apache from my desktop machine (which connects to the net) and put it on an old 500MHz machine of mine (Told you my LAN was small). Everything was, and to a certain degree, still is running fine. However, I decided it high time I made this webserver of mine accessable to the world. At the time, I thought it'd be a trivial task, how wrong I was.

(!) [Jason] Why did you do this? Not that there's anything wrong with it or anything, but if your desktop machine can handle the traffic without causing problems, I don't see any reason why you couldn't run your web server on it. But....

(?) Anyway, after asking on experts-exchange.com for some help with my iptables configuration and badgering various people in #hants on irc.blitzed.org I eventually got traffic forwarded to my webserver. However, when accessing the webserver from, not surprisingly the web, I get a lovely 403 (See Error Message :). I've just set LogLevel to overkill (more commonly known as Debug :D -- Thanks for the suggestion, Heather.) in Apache and have what seems to be useful information (See Access_Log: and Error_Log :). But, alas, I have no idea where to go from here, any advice would be more than welcome. (For information on my LAN and general other stuff, see Info :)

Server Distro: RedHat9
Desktop Distro: RedHat9
Apache Version: 2.0.40
	Diagram: (Yes, it IS that basic).

Error Message:
You were denied access because:
Access denied by access control list.
Access_log: - - [26/Nov/2003:17:26:08 +0000] "GET / HTTP/1.1" 200 2336 - - [26/Nov/2003:17:26:08 +0000] "GET / HTTP/1.1" 200 2336 - - [26/Nov/2003:17:26:08 +0000] "GET /favicon.ico HTTP/1.1" 404
[Wed Nov 26 17:26:08 2003] [error] [client] File does not exist:
[Wed Nov 26 17:26:08 2003] [error] [client] Syntax error in type
map, no ':' in /var/www/error/contact.html.var for header
[Wed Nov 26 17:26:08 2003] [error] [client] unable to include
"../contact.html.var" in parsed file /var/www/error/include/bottom.html

(?) Well, I'm now in an even worse situation. Having just moved from RedHat abck to SuSE, I cant get as far as I was before. I'm nbow using the following lines:

iptables -t nat -A POSTROUTING -o ppp0 -j MASQUERADE
iptables -t nat -A PREROUTING -i ppp0 -p tcp --dport 80 -j DNAT --to
iptables -t nat -A PREROUTING -i ppp0 -p tcp --dport 443 -j DNAT --to
iptables -A INPUT -i lo -j ACCEPT
iptables -A INPUT -i eth0 -d -j ACCEPT
iptables -A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT

(?) Now, when I try to access apache from my ppp0 ip, I don't get through, it

just doesn't seem to connect. Any clues as to why? (Ohh, and /proc/sys/net/ipv4/ip_forward is 1).

(!) [Ben] None, AFAIK; that would be why it's not happening. Here's me forwarding, both in and out (-s for source, -d for destination) for my iPaq:

# Flush iptables
iptables -F

# Masquerade any packets that go our from the specified address
iptables -t nat -I POSTROUTING -j MASQUERADE -s

# Forward any packets _for_ 202
iptables -I FORWARD -s -j ACCEPT
# Forward any packets _from_ 202
iptables -I FORWARD -d -j ACCEPT

echo 1 > /proc/sys/net/ipv4/ip_forward

(?) Preventing Shutdown if NFS Shares Are Open

From edal

Answered By: Thomas Adam, Jim Dennis

(!) [Heather] This thread followed us across the move from SSC, parts of it were on both editions of the answer gang's mailing list...

Hi there

Does anyone have any ideas ? Answers to edal@NOSPAM.freestart.hu please, remove NOSPAM for the address to work.


(?) I run a couple of machines at home, both setup with Fedora, a laptop and a server which also doubles as a second desktop machine. The laptop accesses a home directory on the server using an NFS share and the 'mount' command. All of this works just fine apart from one problem. When the server is shut down and I have an open NFS share on the server my KDE desktop hangs.

(!) [Thomas] Yep -- I can see how this might be. I run NFS on my LAN at home, and although I do not have the same problem as you (fvwm), I suspect the reason why KDE hangs is because "konqueror" is an integrated (highly integrated) part of KDE. It is not just a file/web manager, it is also the backbone. If that hangs, you've had it.

(?) I've done some playing around with the /etc/shutdown.allow file but all this does is list the people who are allowed to turn the server off. What I'm looking for is a way to prevent a server shutdown if someone else is using an NFS share on the machine.

(!) [Thomas] Hmm, you'd have to do the check before the "unmount -a" command is run on init 0. There is a file present in all Linux distro's called "/etc/halt.local" which gets run on init 0. The trick here though is to know the order in which it is run. Obviously, it'll be no good if it gets called before the "unmount -a" option. Luckily for you though -- it doesn't.
So, the steps you might do here is thus:
# touch /etc/halt.local
# vi /etc/halt.local
Add the following...

#halt.local -- ought to get read at init 0


[ $(mount | awk '/name_of_dir/ {print $5}') = "nfs" ] &&
   /sbin/shutdown -c
save the file.
But of course, if I had read your question, I'd have realised that actually, what you ought to have is something like this in your ~/.profile file (ignore everything previously -- I'm leaving it in for historical purposes):

See attached thomas.clientside_haltme.bash_profile.txt

Then run:
source ~/.profile
and try running:
haltme 0
Not tested it -- ought to work though.

(?) Perhaps it is my limited knowledge of English (I thought that posting in Hu ngarian would be a problem) but I do not think I got the question across co rrectly.

I do not want to do anything with the NFS client, I want the NFS SERVER to cease a shutdown if one of its NFS shares is in use. Is this what your file does ?

(!) [Thomas] No, it negates it the otherway around, and says that if the client NFS is mounted then do not shutdown the NFS client. Shrug -- OK, so we negate the problem onto the NFS server... This will be a little more trickier to do. I suppose you could utilise the /etc/exports file but even then, you'd have to have a way of testing it.
(!) [JimD] This symptom is the classic result of NFS hard mounts and without the "interruptable" option; which are the defaults under Linux.
Change the NFS line(s) in the /etc/fstab to list "intr" in the options field. Something like:
fserver:/usr/share      /mnt/nfs/fserver	auto	intr,ro               0 0
Feel free to read the fstab man page for details about what these fields mean; and the mount man page, particularly the section on NFS options.
Making it "interruptable" will allow process that attempt to access this export (share) to be killed. By default such processes will simply be blocked until the NFS share becomes available.
You could make it a "soft" mount --- which would be that the attempts to access such directories or files would eventually timeout. However, "soft" mounts are generally considered to be a bad idea. Most programs will abort and exit on some timeouts; however, some will just exhibit odd, unpredictable, behaviors on file/directory access timeouts.
When you mount filesystems you should make it a practice to unmount them when not in use and especially when shutting the NFS server down or disconnecting that machine from that network (in the case of laptops).
Keep in mind that NFS was not designed to support laptops, mobile use, and "occasional use" filesharing. It's built around a set of reliability assumptions and intended semantics that are not suited to situations where your fileserver might not be up or might be inaccessible. It's not suited to "browsers" and interactive file manager use where attempts to access a directory can result in a "soft" error.
NFS systems try to open a file or access a directory and they continue trying FOREVER until they are interrupted (if the intr option is enabled), the system is restarted or the server becomes available.
I've heard of an old case where a pair of UNIX systems were connected over NFS, where an unattended job was running on the NFS client while it's server was down. The server was replaced! The data was restored to the new server and, when it was brought up on the net the client's process' woke up and completed their job. (That was a month after the job started --- it just slept in the interim). I have personally had an NFS server fail, hard drives fail, brought it down, replaced the drives, restored from backups, and seen the clients just continue working on the newly restored system unaware of the change.
It's a different set of reliability semantics that harkens back to a batch processing computing model.
Eventually some form of AFS, Coda, Intermezzo or some other newer filesharing protocol (perhaps even NFSv4) may be more appropriate to your needs. For now, just add the intr option to your fstab and understand that processes that access those portions of the tree will block forever unless they implement their own non-blocking and timeout semantics.

(?) SSH tunneling

From Ben Okopnik

Ah... Ben. You just know it has to be juicy good stuff if it stumps one of the core Answer Gang like this. Enjoy! -- Heather

Answered By: Karl-Heinz Herrmann, Rick Moen, Robos, Heather Stern

Hi, all -

This week, I'm teaching at a facility in Baltimore where the admin has decided that a non-transparent proxy is The Flavor Of The Week. This, needless to say, is a Huge Pain. I have to define/undefine HTTP_PROXY and FTP_PROXY - and their lowercase equivalents - and log out and back in when I'm there, and reverse the process when I'm back in my hotel. Oh yeah, gotta do the proxy settings in Mozilla, too. Oh, and if I want to use Netscape to test something... Yecch.

<Ron Popeil mode>"But there's more!"</RPm> In order to do anything useful with files at LG, I have to tweak them locally, then upload them to the border router (Monsieur Admin saw fit, after much conversation, to give me SSH access to it), then shove them up to LG from there. This is annoying, to say the least.

So, my question is this: would it be possible for me to set up some sort of an SSH tunnel from my 'top through that border router? I saw something about tunneling in the MindTerm dialogs (I'm not really even sure why I'm playing with MT, except that I was curious about it :), but couldn't figure it out since I don't understand the basics behind the concept.

I've got "authorized_keys" on the router (which uses port 1022 - hey, might as well make it interesting, right?); I can download whatever software I need via HTTP or FTP. No "rsync", no SMTP, no POP, and no direct SSH access, though. :(

Any advice?

(!) [K.-H.] So you've ssh access on the router? then you can tunnel whatever you want, basically. In howfar things are getting more convenient is something else. Still you've the different setups inside and hotel.
Let's start with improving mail access :-)
from my ~/.ssh/config
(!) [Heather] With some tweaking to sanitize hostnames and make the examples consistent.

See attached kh-ssh_config.txt

(!) [K.-H.] One major drawback with ssh tunneling is:
You want to tunnel arbitrary connections like a http proxy, but for every target you have to setup a forwarded port as the information where you want to go is lost in the tunnel. Another problem might be that you need a target from where you can access everything you want. Having a proxy on that other end helps a lot for http and ftp.
Theres seem to be very recent ssh versions which can improve this situation, but I'm not quite sure how they handle this. My local version does not have anything in the man-pages. That might have come up on TAG -- or maybe somewhere else.
If you want to rsync LG files and this is a defined port you can set up a forwarding for that too of course.
forward a gateway port 9999 to target:rsyncport connect to gateway 9999 and tunnel to target:rsyncport
ftp passive should work too -- but http and ftp work via proxy anyway. ssh to a small set of targets is possbible via a set of forward rools, one each target. something like:
alias "ssh_target1"="ssh gateway:target1port"
might make it even convenient.

(?) OK... I'm missing something. I'm not sure if I've got this right, but here's a part of my ~/.ssh/config:

See attached ben_ssh_config.txt

I tried the above - "ssh -p 8022" - and got "Connection refused". :((( This is for the local machine (the laptop), right?

(!) [K.-H] All these hosts and ports are somewhat confusing. Or you might miss the "GatewayPorts yes" in the config.
You've two possibilities I think:

(?) Ah-ha. OK, this is starting to make sense - among all the permissions stuff, etc. I think that what you're saying is this:

  1. On the gateway, bind a listener to an unprivileged port and forward
  2. From the local host, connect to the forwarded port on the gateway

Man, that sounds too simple. :)

(!) [K-.H]
I explain the first in more details, I think, as this should be enough for e.g. mail access.
shell one
khh > ssh -f -N -L 8099:mod001.example.com:25   mod017@mod021.example.com

(?) This is being issued on the gateway, right? I understand the "port:host:port" syntax: 8099 is mod01:8099, which is being forwarded to mod021:25 (the remote machine).

(!) [K.-H]
shell two
khh > telnet localhost 8099
Trying ::1...
Connected to localhost.
Escape character is '^]'.
220 mod001.example.com
Sendmail 5.65v4.0 ( Tue, 9 Dec 2003 21:51:33
shell one reacted to the connection:
The following connections are open:
  #3 direct-tcpip: listening port 8099 for mod001.example.com
port 25, connect from ::1 port 33813 (t4 r2 i0/0 o0/0 fd 12/12)
mind the localhost as other interfaces are not "local".
khh > telnet khhlap 8099    ## khhlap is me too
telnet: connect to address Connection refused
What you do now is run one ssh from the lap to the gateway
telnet localhost 8025
do ?
point fetchmail (or MUA directly) to localhost port 8995 and you should be able to read mail instead of working ;-)
(!) [Rick] I'm tempted to suggest proxytunnel, corkscrew, or httptunnel, as mentioned in http://linuxmafia.com/~rick/linux-info/firewall-piercing .
http://proxytunnel.sourceforge.net http://www.agroman.net/corkscrew http://www.nocrew.org/software/httptunnel.html
Get in touch with your inner BOFH, Ben.

(?) I actually ran across "corkscrew" on a Debian list; however, its description (from "apt-cache show corkscrew") sounds exactly like the Perl script that Frodo sent in, requiring HTTPS and support for the "CONNECT" method. I hadn't run across either of the "*tunnel"s, and will check them out if Karl-Heinz' method (which sounds like it _should work!) doesn't pan out.

(!) [Robos] OK, I still have very little clue about networking, but here in my appartment my flat-pal set up a vtund (a tun) over which we pass everything when we go wireless. This is a tunnel over ssh. Ping, dhcp, http, ftp, everything goes through this. Isn't this what you need? Sorry if I misunderstood it.

(?) Hum. I just tried this on the laptop - I'm not at work anymore, so I'm a little restricted in my experiments.

ben@Fenrir:~$ ssh -fNL 8995:localhost:995 target.example.com

It did what I thought it should - backgrounded itself.

ben@Fenrir:~$ ps ax|grep "[s]sh "
  657 ?        S      0:00 ssh -fNL 8995:localhost:995 target.example.com

Then I tested it -

ben@Fenrir:~$ mail -sfoo ben@linuxgrrzette.net
ben@Fenrir:~$ fetchmail -vvv --ssl -uben -P8995 localhost
Enter password for ben@localhost:

Wow, cool. That worked. However... I'm still trying to figure out how it'll work with three machines. Would it be something like this?

# Issued on the gateway
ben@gateway:~$ ssh -fNL 8995:localhost:995 target.example.com

# Issued on the laptop
ben@Fenrir:~$ fetchmail --ssl -uben -P8995 gateway
(!) [K.-H] At least you got a working setup going. The ssh commandsequence I gave you was a sllightly different concept as the one you tried, that's why we still do not talk about the same thing.
I was trying to setup a connection like this:
lap runs a tunneling ssh to gateway. lap is 10.* so private, gateway is 10.* but should be able to route to outside, or it wouldn't be a gateway. So if you set up a ssh from lap to gateway at lap> ssh -L 8995:OUTSIDETARGET:995 gateway
you should then be able to connect to: at
lap> telnet localhost 8995
and reach OUTSIDETARGET 995
The other version is, as I tried to explain earlier (and what you tried successfully now):
You run the tunneling ssh from gateway to some place, OUTSIDETARGET
at gateway> ssh -L 8995:OneMoreTARGET:995 OUTSIDETARGET
You then can connect from lap to gateway 8995 and reach the OneMoreTarget 995
THIS needs GatewayPorts yes as you connect to the forwarded port on gateway from the lap, i.e. non local

(?) OK, I can do that (after disabling the forwarding in .ssh/config - otherwise I get "bind - Address already in use"):

on laptop> ssh -p 1022 -L 8995:target.example.com:995
on laptop> fetchmail -P 8995 -u ben --ssl localhost
Enter password for ben@localhost:

Rats. It didn't work.

Heh, "It didn't work". Might I suggest, gentle querent that you looky here:
That might help you with that phrase :) -- Thomas Adam

(?) I've been playing around with this forwarding thing all day, on and off (this course is a bit light on lecture and heavy on student exercise), so I've managed to try everything you folks here suggested. However, one item stands out: most of the suggestions (except those from Karl-Heinz) point to HTTP-type tunnels, all of which in turn rely on the HTTPS "CONNECT" method. One of the authors of "proxytunnel", Muppet, shows a test for it:

muppet@runabout:/home/muppet $ telnet some-proxy 8080
Connected to some-proxy.
Escape character is '^]'.
CONNECT www.verisign.com:443 HTTP/1.0

HTTP/1.0 200 Connection established
Proxy-agent: Netscape-Proxy/3.52

// ---> Tunnel and SSL session starts here
telnet> close
Connection closed.

My problem seems to be that I never get past the "CONNECT"; it just sits there. Which pretty much says none of the methods that rely on it are going to work.

I don't know what I can do at this point, since the admin here seems rather paranoid about touching the gateway setup... so I guess I'm stuck, unless someone comes up with another idea.

Thank you for trying, everyone.

(!) [K.-H]
This is getting more complicated If something on gateway interferes with ports. On the other hand I got out of the Indian research center which simply blocked everything in and everything but port 80 and 23 (and ftp) out. That required a sshd outside running on port 23. So don't despair yet... Oh -- but you said they block everything and offer only http proxy and ftp proxy.
I'm not 100% percent convinced it didn't. Ther was a connection to something. If fetchmail obeyed the -P 8995 it was not a pop3 running on laptop at port 8995 by accident. You would know.... for all fetchmail knows it*is* connected to localhost and you asked for user ben. Of course you have to supply users/password for target.example.com (secure pop3 on 995). Might the ssl stuff open other ports as well? Or just an afterthough while typing a reply below: Does fetchmail ask the passwd before it connects? Then it doesn't show anything of course.
On the other hand if supplying a password at that point didn't work and the user is ok.... hmmm....


If I try to enable GatewayPorts, I get "bind - Address already in use", which probably means some odd firewalling going on. The same thing happens with trying to forward 8022 to 22 on "target.example.com". Doesn't seem like this method is going to work.

(!) [K.-H]
Hm. You tried to switch on GatewayPorts where? For the above setup it would only make sense on Laptop (Fenrir) -- GatewayPorts allows non-local connections to the local forwarded port (i.e. the first number after -L to ssh).
Hmm... at this point lets assume they messed up the gateway so either the gateway sshd is not allowed to forward anything or or they just dump packets from inside which are not for the two proxy ports.
> at gateway> ssh -L 8995:localhost:995 target.example.com
(!) [K.-H] again looks ok


at laptop> fetchmail -P 8995 -u ben --ssl
Enter password for ben@

In the log file:

Dec 10 11:05:50 Fenrir fetchmail[2716]: POP3 connection to
failed: Connection refused
Dec 10 11:05:50 Fenrir fetchmail[2716]: Query status=2 (SOCKET)
(!) [K-H.] Hm.

(?) I've also tried it as

at gateway> ssh -L 8995:target.example.com:995 target.example.com
at laptop> fetchmail -P 8995 -u ben --ssl
(!) [K.-H] ok. good to make sure.

(?) Same error as above.

Just to test it, in a really simple manner:

at gateway> telnet target.example.com 25
(works fine)
(!) [K.-H] good. At least you do get out.


at gateway> ssh -L 8025:localhost:25 target.example.com at laptop> telnet 8025 Trying telnet: Unable to connect to remote host: Connection refused

(!) [K.-H] Hm. Might be firewall on gateway dumping/refusing your connection even if you've a nice open port.
Well at least I understand the next:

(?) at gateway> ssh -L 8025: target.example.com # My IP

(!) [K.-H] if it's on gateway (and only there you can see target.example) you've got the port on gateway. You are forwarding to a private IP -- whatever that in context of target.example might be.

(?) Tried it both enabled and disabled (on the gateway machine, that is); no luck.

(!) [K.-H] That would be the proper place (gateway).
I just wanted to admit defeat, but can't you connect from the back form the gateway to lap with -R? Where is the manpage....
ok, one last try:
- you connect (ssh) to gateway - on gateway run: ssh -R 8995:target.example.com:995 laptop
- now on laptop your fetchmail sequence
- try again with (on gateway)
ssh -R 8025:target.example.com:25 laptop on laptop: telnet localhost 8025
This is cutting the gateway sshd out of the chain -- but they still might have non overrideable ssh client configs prohibiting -L entirely. "-v" to ssh does not give any errors/warnings?
If that fails too -- I think it's possible to run a ppp line over a terminal (telnet) connection. I don't know how to setup a pppd over terminal but I think I know how to setup the terminal tunnel:
on lap:
pipe here |  ssh -e none gateway ssh -e none target.example.com | pipe here

sprinkle freely with -f -n -N
(!) [Heather] I know we have a number of tunneling toys on LNX-BBC; I wonder if it has something that we haven't mentioned. If not, it would be awful fun to chase that on down.
My normal solution is to put an ssh service on a port that people, um, think means something the firewall says is ok. After that it's all a pipe... a port's a port.

Ben then cheerfully reported back the following....

(!) [Ben] WOO-HOO! Karl-Heinz, you're The Man! :) It works fine. I can get my email... Can't send it yet, though. I've done the following:
gateway> ssh -p 22 -R 25:target.example.com:25 root@laptop
which gets me genetikayos:25 sitting at laptop:25... but I still don't have name resolution on localhost:
delivering message 1AUVAe-0002gK-00
  == ben@linuxgrrzette.net R=lookuphost defer (-1): host lookup did not complete
Almost there, though!
I ran out of time before I had a chance to try that out (I'm sure it would have worked fine) - this class usually wraps up around 1 or 2pm Friday, and then I'm out of there and looking for the fastest way home. However, it looks like I might be teaching there again soon (the students gave me perfect ratings, and the facility manager was very_ happy), so I'll probably get another shot at it.
Thanks for all your help - it's been a terrific education in SSH capabilities!

(?) about windows and linux...

From Viper9435

Answered By: Heather Stern, Thomas Adam, Tom Brown

Im currently using Xoblite, and do you know how i can make my windows xp look more like linux?

Please, Please, please send your e-mails in plain/text. HTML is evil and just wraps useless meta-data around the precious text. Both Heather and I have been mentioning this in past months...don't do it again, gentle readers. -- Thomas Adam
(!) [Heather]
  1. There are alternative window managers for Windows; you could switch to
  2. I once saw a package called "enlightenment for Windows" and what it
(!) [TomB] For the command line part, you shouldn't forget Cygwin. It does a good job of giving you a Linux CLI, and it's free.
(!) [Thomas] I am going to have to agree here, and also mitigate this question by asking why would you want to play a game of 'cloak and daggers' with your windows machine -- dressing it up all you like to try and make it look like Linux won't change the operational fact that underneath all the superfluous style remains IMHO, an unstable, unreliable operating system. If you ask me, if you have to make Windows look like Linux, don't. Instead, just install Linux and be had with you.
(!) [TomB] But, if you're looking to change the appearance of XP, there are several solutions. None are free that I know of. The best is from Stardock, in their Object Desktop collection of utilities. The whole thing costs about $50, and has a ton of great stuff in it. Or, you can buy just one piece of it for about $20: Window Blinds. Window Blinds allows you to change the entire GUI using "themes". For example, someone wrote a "Blue Curve" theme that looks exactly like Red Hat's GUI. Someone else has ported the Blue Curve icons, which you can install using Object Desktop's Icon Packager. There are utilities that allow you to change the logon screen -- and again, someone's created a Red Hat logon screen. Look at some of the screen shots on www.wincustomize.com to see the themes available before you buy anything. The Object Desktop collection even includes a tool to design your own Window Blinds theme, if you don't see anything you like on the web.
(!) [Thomas] There is also now a port of fluxbox to windows. Unfortunately I don't remember the URL, but this'll give you, the gentle readers, a chance to re-aquaint yourselves with http://www.google.com/linux

(?) checking hard-disk in debian

From Joydeep Bakshi

Answered By: Colin Charles, Thomas Adam

Hi list,

Here is a typical problem in debian. after particular days my debian show during booting * /dev/hda6 mounted 31 times without checking, check forcde* and it starts fsck.

now my question is that ; has debian programmed to check hard disk after 31 times mounting the disk ? if so how to change this so that it will check hard disk whenever find a problem like red-hat ?

thanks in advance.

(!) [Thomas] This is not a 'problem' but a design descision. When you originally created the partitions during the debian install, debian does tell you that this feature can be changed via the tune2fs program
(!) [Colin] I find using the option:
shutdown -fh now
where the -f switch skips fsck on the next reboot a rather helpful thing to avoid getting fsck started up at all.
Yes (but I'm not certain with regards to 31 times, it could be higher). To make Red Hat do the same thing (it does, but after a much higher mount count), use the tune2fs tool.
(!) [Thomas] Perhaps you are confused, Colin? tune2fs will either check the drive after a certain number of mounts have been had, or it will check it after or uptil a certain date -- whichever one comes first.
(!) [Thomas] I have mentioned tune2fs countless times over the years, however...
tune2fs -c 100 -C 1 /dev/hdxx
where hdxx is your device, will mean that after every 100 successive mounts, your drive will be checked.
(!) [Colin] If you shutdown incorrectly (instead of issuing shutdown/halt, you hit the power switch), Red Hat or Debian will run fsck upon the next reboot since there could be "problems".
(!) [Thomas] This is only due to the fact that mount did not umount the drives correctly. Again, this can be had with tune2fs. The process by which init goes through to shut your machine down is usually pretty good. Unless one is still using ext2, the process is usually quick since if one is using ext3, the journal will only check the superblock for the last changes made.
As an aside, one tip I always give people is that when one is creating new partitions, for '/boot' I make that ext2, since as it is mounted ro (read-only) it doesn't require a journal.

(?) question about env vars

From - EJ -

Answered By: Thomas Adam, Karl-Heinz Herrmann, Jim Dennis

Again, this thread has followed us across both "TAG" mailing lists to the new site. For readers keeping up on both, be advised that very few if any of the LinuxGazette.Net answer gang hang out on SSC's version of the list at all anymore; this may be the last month that the older list sees any answers. Some of the Gang left the old list more because of spam overload via that source than the changeover per se but there you go. The correct place to reach The Answer Gang now is tag@lists.linuxgazette.net. -- Heather

Could someone please help me setting env vars within a scrpt but will remain with my interactive environment. Please note I am trying to do this with ksh and bash; however, I am not getting success. The env vars set in the script, I can echo them, but they disappear after the script has completed. How can I have the env vars remain after the script is completed similar to .profile?

Thanks in advance.

(!) [Thomas] You have sent several e-mails to this list before...PLEASE please send in PLAIN-TEXT only.
You have to "export" them, like so:
export MY_ENV_VAR="my value"
Then when the script exits, you can do:
echo $MY_ENV_VAR
from the CLI, and you will see the value stored therein.
(!) [K.-H] This might be a problem with subshells.
khh > ./test.sh
khh > echo $TEST_VAR
khh > cat test.sh
export TEST_VAR="test"
echo $TEST_VAR
The script runs in its own shell and CAN NOT change the environment of the parent (your shell in which you are typing).
run the sccript with source:
khh > source test.sh
khh > echo $TEST_VAR
a shortcut often is ".":
> . test.sh
(!) [JimD] It can't be done. You are suffering from a fundamental misunderstanding of how Linux (and UNIX) works.
Variables set in your shell are part of your process. Environment Variables are set in your shell and moved (exported) to a region of memory that is preserved through exec*() system calls.
When you run an external command (binary or shell script) it runs in a subprocess. You subprocess inherits A COPY its parent's environment. I can modify that. However, at the end of the process then the COPY is reclaimed (freed).
So, if you have a script that set variables for you; you can't execute it in the normal way. That is to say you can't invoke it as a program. So you have to "source" it. This is done using the . (dot) command.
Let me give an example:
... contains a set of lines like:
	export FOO BAZ
If you invoke it:
... then your shell runs mysettings.sh in a subprocess; which dutifully sets those variables and exports them; and then promptly FORGETS them as it dies (exits). (Right after the end of the script; there's an implicit exit to the subprocess).
If you source it:
 	. ./mysettings.sh
For those of you playing along at home the "." is a synonym for 'source' -- Thomas Adam
(!) [JimD] ... then your shell reads each line of the file and evaluates each one as if you'd typed it in yourself. Any settings made IN THIS WAY will persist for the life of that process (your interactive login shell for this example).
This is, by far, one of the most confusing and most often misunderstood facets of shell programming and based UNIX usage.
Some day I'm going to have Heather create an animated web picture, and slide show, perhaps even a little "flash" file depicting this process of variable assignment, export, sub-process creation (fork()ing), program execution (exec*()ing), process termination (exit()ing), sub-process exit status harvesting (or reaping, using wait()), and signal handling (SIGCHLD).
It's a big part of my basic Linux classes.

(?) temporary environment changes

From Ben Okopnik

Answered By: Jason Creighton, Thomas Adam, Karl-Heinz Herrmann

Recently, I spent a week at a client's location which required setting several environment variables in order to use their proxy server. Something that made it quite annoying was the necessity of un-setting these variables when I went back to my hotel room and connected via dial-up. Setting and unsetting the variables and logging in and out twice every day did not appeal to me, so I modified my "~/.bashrc" file by adding the following lines to it while logged in and running X --

[ -f ~/PROXY ] && {
	export HTTP_PROXY=
	export FTP_PROXY=
	export http_proxy=
	export ftp_proxy=

I then created a file called "PROXY" in my home directory. Proceeding from this point was a simple matter: when I needed the above variables to be unset, I moved "PROXY" to "NOPROXY" (any other name would do as well, but I wanted it to be an obvious reminder) and closed all the open xterms. Any xterms I opened from that point on would not have these variables set. Reversing it was just as obvious - a matter of renaming the file back to the original name and closing all xterms again.

Mozilla isn't really amenable to this kind of thing and would have required manual changes every time, so I just used Dillo and w3m when away from the office.

(!) [Jason] Seems like there should be a way to do this automatically. If there's a network share at that client's location, you could make PROXY a symlink to it, thus rendering it broken when you don't have the share mounted, causing it to fail the existence test.
(!) [Thomas] Indeed, Jason -- something which I do all the time, i.e.:
[ ! -e $(ls -l $HOME | awk '/PROXY/ {print $11}') &&
   # hmm, you must be joking, right?
   exit 1;
} || {
  # so it is there, and working, continue with the exports....
If I was really worried, I might also just prefix a test for PROXY to make sure that it actually is a symbolic link (test -L).
(!) [Jason] Or you could look at the network address of the interface that you're using (Ethernet? Or some cool wireless dealy?) to see if it matches a certain pattern. (Presumably the IPs are handed out by DHCP)
(!) [Thomas] If it were DHCP, I wouldn't bother with this idea, since the IP would change each time.
(!) [Jason] Or you could just stick with what you've got, but that wouldn't be as much fun.

(?) Maybe not - but it _would allow me to work at different clients' locations, with different network shares, IP patterns, etc. - that being the point of leaving this gadget in place rather than just deleting it once I was done. ISTR running into this in at least one other client center... maybe more, but I can't recall.

(!) [K.-H.] There are programs out there which determine the network you are in and run scripts for you (e.g. link different resolv.conf and hosts in place and set a proxy).
One I've used for some time is divine (seems unsupported by now and a recompile just didn't want to work the last time I tried). Another I've found but not yet tested is intuitively (intuitively_0.1.5-1.tar.gz). That would automate the change of the basic network config based on IP's found in the neighbourhood (divine sends arp requests).

(?) Wouldn't "divine" require knowing a given network's specifics in the first place?

(!) [K.-H.] Yes -- you would have to put a line in the divine.conf with an IP to be found on the network to identify it. Some other details as well. Once done it's fully automatic.

(?) The problem is that I don't, until I get to the specific site. It seems that the centers where I teach are set up based on the local sysadmin's preferences. However, I do use a self-modifying script that "memorizes" the IPs I give it; after running it once in a location, set up for the rest of the week is a matter of running it and hitting "Enter" four times. I've just rewritten in in Perl (it used to be a shell script with Perl one-liners in it...) Note that it does have to be run as root - or it could be modified to use "sudo".

See attached memorize-network.perl.txt

I'll admit that the experience _is interesting - at this point, I can fit my laptop into just about any network environment that these folks have been able to think up, which is a point of pride. Of some sort, anyway.

(!) [K.-H.] That way of modifying the script itself is interesting. I would have thought of input files only. I know you get into deep trouble if you overwrite a shell-script which is running, with perl this should work as perl is compiled at the beginning.

Published in Issue 98 of Linux Gazette, January 2004

News Bytes

By Michael Conry

News Bytes


Selected and formatted by Michael Conry

Submitters, send your News Bytes items in PLAIN TEXT format. Other formats may be rejected without reading. You have been warned! A one- or two-paragraph summary plus URL gets you a better announcement than an entire press release. Submit items to bytes@lists.linuxgazette.net

Legislation and More Legislation

 Jon Johansen

The trial in Norway of Jon Johansen regarding his involvement in the development of software to circumvent the content encryption used on DVD movies has ended ... again ... for now. As was reported before, Jon was already acquitted of the charges, but an appeal against the verdict was quickly filed by state prosecutors. Happily for Johansen, the appeal court has upheld the earlier acquittal. Less happily, there is the possibility of further appeals to higher Norwegian courts. A detailed report on the story is available at Aftenposten.

Linux Links

Linux Weekly News has released a timeline for the past year.

RMS has written a personal account of his experiences at and opinions of the recent World Summit on the Information Society

Helping to introduce charities to free software [NewsForge]

Using a Linux desktop for graphics and media production

How the WorldVista software is to be used to affordably help medical systems in the less-developed world [NewsForge].

Linux Mobile System's GNU/Linux system boots from a USB pendrive

NewsVac highligted an interesting article at OSNews about the state of video editing on Linux/Unix.

VNUnet reports on the use of Linux in the earthside systems behind the (possibly ill-fated, it now appears) Beagle Mars lander.

An interesting article from O'Reilly.com Myths Open Source Developers Tell Ourselves

The Economist on regional Linux variants

Upcoming Linux conferences and events are listed on Linux Journal's Events page.

News in General

 Linux Kernel 2.6 Released

The Linux kernel has a new stable version. Version 2.6.0 was released on the 18th of December. The Changelog can be read here, while a more conversational announcement can be read here. LG has an article in this issue about the changes.

 Open Source in Israel

The Register reported earlier this month that the Israeli Treasury was following earlier moves by the Department of Commerce and investigating the possibility of using open source software in place of Microsoft offerings. As noted by Zuri Dar in Haaretz, this could well be no more than a ploy to gouge a better price out of Microsoft for their software. However, there does appear to be some real momentum behind the open source initiatives as the government has instigated plans to distribute free OpenOffice.org CDs to the Israeli public.

Distro News


As plugged at NewsVac the CollegeLinux project has released version 2.5 ObiWan


Debian Weekly News highlighted the publication of the Debian Timeline for 2003.


OSNews recently reviewed Lorma linux 4.0. This distribution is based on Fedora (itself related closely to Red Hat), and uses a trimmed down set of packages such that it ships on a single CD. The focus is largely on desktop applications.

Software and Product News


Version 5.0.0 of MySQL has been released.

 EMC acquires VMware

NewsForge has reported that EMC has acquired VMware

Mick is LG's News Bytes Editor.

[Picture] Originally hailing from Ireland, Michael is currently living in Baden, Switzerland. There he works with ABB Corporate Research as a Marie-Curie fellow, developing software for the simulation and design of electrical power-systems equipment.

Before this, Michael worked as a lecturer in the Department of Mechanical Engineering, University College Dublin; the same institution that awarded him his PhD. The topic of this PhD research was the use of Lamb waves in nondestructive testing. GNU/Linux has been very useful in his past work, and Michael has a strong interest in applying free software solutions to other problems in engineering.

Copyright © 2004, Michael Conry. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004

Constructive Paranoia at the End of 2003

By Rick Moen

Some weeks ago, I was spending some time assisting my mother-in-law, who's working on her Ph.D in computer network security, do a survey of a half-dozen or so Linux users at local user group, CABAL, about our security practices — with results like these:

Do we use anti-viral software? (No, except where we handle files or mail destined for Microsoft-OS machines.) Do we study our networks' security exposure using vulnerability-scanning software such as nmap and snort? (Many of us do, yes.) Do we run log-analysis security utilities such as logcheck? (Ditto.) Do we run Intrusion Detection System (IDS) suites such as Tripwire? (Almost never.) What measures do we take to eliminate security holes as they arise in a timely fashion? (Various.) Do we use kernel-level IP-traffic filtering scripts ("firewalls")? (Some do. Several more-experienced users operating all-Unix networks do not.)

These questions kept haunting me as I answered questions from new Linux administrators. Sometimes, those were very astute: Q: "How can you be sure that your system hasn't been compromised by hostile parties?" A: "Excellent question. You can't know, absolutely. A truly subtle and competent intruder manifests those abilities in part by being difficult to spot, and covering his tracks. But intruders (or their automated attack tools) generally break into a system to do some significant unauthorised activity, leaving clues that will be spotted by alert and capable admins who know their systems well enough to notice peculiar goings-on." That answer tends to leave questioners slightly uneasy — as intended.

Two Classes of Attack

I always told new admins that there two levels of threat to systems: from outside and from inside. Conventional thinking worries mostly about the former, e.g., the sort of perimeter survey you get by running

nmap -vv -sT -sR -O -n -oN tcpscan.log
nmap -vv -sU -sR -O -n -oN udpscan.log
nmap -vv -sA -sR -O -n -oN ackscan.log

...from the far side of your LAN (a different machine) against your IP= Linux host, to guesstimate how hard the latter's (figurative) exterior shell is — in the event of attack from elsewhere.

But the latter sort of threat — from the inside — is both more worrisome and more interesting. That is, if there's a entrance method (ssh, whatever) from the outside into your machine, and someone steals the password or other token required to use it, then you have unwelcome guests, who can then subvert your system's security from its own command prompt. It's well known that the latter step's often successful, but the real news is how easily and frequently passwords get stolen.

Consider inbound ssh access to your machine(s), a convenience much used and cherished by Unix users. Do you ever ssh in from machine you don't personally administer and have absolute confidence in? Even if you don't, do you always carry your ~/.ssh/known_hosts2 file with you, so you can be sure the remote host you reach is really yours? If you're ever lax on any of those matters, and you're even slightly unlucky, the bad guys will steal your access tokens and enter masquerading as you, later.

Even if you never do any of those things (making you a rare paranoic, indeed), can you say the same of all the friends you gave shell accounts to? Nobody ever used a cybercafe or university computer, or used PuTTY on a family Windows box teeming with spyware (or maybe even a keystroke-recording dongle connected to the keyboard)? Thought not. And there's your problem.

The November Surprise

Which brings us to November's security incidents. A timeline should help us set the scene:

You'll notice how quickly the Debian and Gentoo people realised their problem, and corrected it — a point I'll come back to. But the first point to note is how the bad guy entered to begin with — a necessary first step before he could use the kernel flaw.

It turns out that one of the 1000+ Debian developers had been the victim of security-token theft. He used an ssh client on some machine, somewhere, that happened to have already been subverted. The ssh software was "trojaned" and privately logged his Debian-server login credentials, later conveying those to the attacker — who was then able to waltz in as if he were the developer. Only then did he use the kernel bug to escalate privilege to root-user access, something he might equally have done by finding an un-patched flaw in any other piece of security-sensitive software. The main point is: The intruder got in despite everyone (probably) being reasonably cautious and prudent, because one of his password-grabbing processes got lucky somewhere.

IDSes and Their Discontents

Two things put an immediate halt to this malarkey at the debian.org and gentoo.org sites. One was the presence of alert sysadmins, who, in debian.org's case, noticed the pattern of kernel "oopses" on two machines simultaneously, judged that far too great a coincidence, and thus were tipped off. (Similar alarm bells probably went through the gentoo.org admins' heads, but far fewer incident details have emerged from them.)

The other was a much-lauded but little-used type of software called host-based Intrusion Detection Systems, the classic example of which is Tripwire, invented by Gene Spafford and Eugene Kim at the fabled COAST security laboratory at Purdue University from 1992 through 1994. For most of its history, it was proprietary (with source code available for inspection but no right to independently develop it), offered for sale to business, and with a "free for non-commercial use" edition called Tripwire Academic Source Release (ASR) available for download.

Over time, Tripwire underwent a complete rewrite that unfortunately did nothing about the program's nagging usability issues (about which, more below), and then, under pressure from open-source alternatives and with help from VA Linux Systems, its sponsoring firm (Tripwire, Inc.) re-released Tripwire in October 2000 as open source software under the GNU GPL.

CTO Eugene Kim has, since then, professed indignation at the sparse participation in response by open source community coders — but the firm's pride and joy turns out to be non-portable C++ with no autoconf support. (Gosh, Gene, maybe those antediluvian coding standards, bizarre choice of language, and your firm's turning to open source only after mindshare had already fled to more-open alternatives have something to do with it?)

I remember Tripwire ASR; like most sysadmins, not at all fondly — having attempted to start living with it in 1994 and deciding it wasn't worth the hassle. It was and is an absolute horror to set up. In theory, you write a description of what files and directories (and what aspects of them) to check for unauthorised changes, have it take a snapshot of the current, non-compromised system state, and commit to disk all that information in a cryptographically verifiable state. Nice theory; teeth-grating execution.

Unfortunately, the tools and configuration syntax are impenetrably obscure, every operation runs incredibly slowly, and its system-integrity-checking mode churns out long and mostly meaningless reports to the root user, which must be studied and then used to further refine Tripwire's human-hostile ruleset to gradually refocus its attention on system changes that actually matter and cease reporting trivia. Because of the heavy use of encryption, each of those steps tends to be dog-slow, and the process must be run through iteratively, many times, using expert knowledge of one's system, before the results start to be useful and not just verbose babble.

Information overload, horrific configuration language, slow and performance-sapping operation, twisted administrative interface... argh! Save me from this! I quit wrestling with it, within a week or two.

Not at all coincidentally, starting one year before Tripwire went open source, it started getting serious open-source competition, starting with AIDE, a package by Rami Lehti and Pablo Virolainen in Finland. AIDE has lately been joined by similar designs starting in 2001: Ed L. Kashin's Integrit, Rainer Wichmann's Samhain, Yoann Vandoorselaere's Prelude IDS, and no doubt others.

I considered using AIDE, when it emerged in August 1999, and played with it a bit. Where Tripwire was slow and system-clogging, AIDE was fast and light. Where Tripwire was obscure and prone to breakage with puzzling errors, AIDE was easy to understand and debug. It had one big problem: The system-snapshot database, program binary, and configuration file weren't stored with cryptographic verification (as Tripwire does). The authors urged, instead, that those all be stored on write-protected media and updated only as needed.

Keeping the AIDE files on floppy or CDR is a major nuisance. The alternative, of just using them on the system's own hard drive, is easier but tends to give a false sense of security. That is, if/when the bad guy comes in and subverts your system, isn't he going to subvert the IDS, too? So, when an IDS tells you all is well, how do you know the bad guy isn't pulling its puppet strings? Tripwire has an answer to that objection; AIDE and friends do not.

That sort of false reassurance is the same one often encountered among users of RPM-based systems reassured by the results of running "rpm -Va" to "verify" the md5sum signatures of installed files: The values are "verified" against a simple Berkeley DB record in /var/lib/rpm — which of course a competent intruder will update to match his changes.

So, in the end, I didn't run AIDE routinely. The Debian Project developer boxes did, and it paid off — the intruder having been sophisticated enough to leverage a previously-unknown Linux kernel exploit, but not enough to notice AIDE and sandbag it before it could inform on him.

The Best of Both Worlds

Following November's security incidents, and my mother-in-law's raised eyebrow over us Linux old-fogies not running host-based IDSes, I felt I had to revisit the matter, and explore options. To my great fortune, the last piece serendipitously arrived in a post to the debian-security mailing list by Lupe Christoph:

"We don't use AIDE exclusively at a client site, but in combination with Tripwire. We think Tripwire is a little more secure because it uses signed databases. So, we protect aide.db with Tripwire. AIDE is used for the parts Tripwire can't do because of its limited configurability...."

Um... yeah. Why didn't I think of that? Whacking Tripwire's configuration down to just the few minimum items it's best suited to handle, including AIDE's own otherwise-unchecked files, means the usual pain of using Tripwire fades into background noise, and makes its operations run in less than geologic time. Meanwhile, AIDE picks up the rest — and I don't have to worry that I'm fooling myself into complacency like an overconfident rpm user. It works a treat.

Concluding Sermon

My Web site's "Lexicon" page includes Moen's Laws, such as:

Moen's First Law of Security: "It's easier to break in from the inside." E.g., many Internet break-ins result from masquerading as a legitimate user to gain user-level access, e.g., with sniffed passwords. The attacker then has a dramatically wider field of system weak points he can attack, compared to penetrating the system from outside.
Moen's Second Law of Security: "A system can be only as secure as the dumbest action it permits its dumbest user to perform." Your users are often your weakest link; smart bad guys will attack you there (e.g., via social engineering). Smart admins will try to compensate for this tendency, e.g., by using multi-factor authentication instead of just passwords.

Between the two of those, one could have predicted the sort of small calamity that overcame the Debian, Gentoo, and Savannah projects in November. Given the considerable likelihood of security tokens being stolen, especially on machines used by many people, it's a wonder it didn't happen sooner. The small miracle of that was that two of the three detected and fixed the break-in immediately — courtesy of host-based IDSes.

Detection is great, and better than a kick in the head (or living in a fool's paradise), but what about prevention? One way is to run an ssh daemon on an additional, non-standard port (maybe 2222 instead of 22) that requires OPIE or S/Key one-time passwords instead of regular, stealable ssh authentication. More precisely, one-time passwords can certainly be stolen, but then are useless because they've already been used up by the authorised user.

One-time passwords are a nuisance to manage: You generate a password "seed" and convey them somehow to your user. He either carries around a printout in very small type of the resulting series of 500 or so one-time passwords, crossing them off as they're used up, or puts the seed in a PalmPilot and generates those passwords from it using PalmKey, Strip, or pilOTP for PalmOS.

I may not use such a setup every time I'm away from home and tempted to cut corners — nor require my users to — but it might be nice to have that option the next time I'm in a cybercafe or some malware-infested bank of public Windows machines at a trade show.

Some Protective Measures to Ponder:

Wichert Akkerman's page of information on the Debian.org compromise includes some intriguing recommendations to add to that, including some behavioural ones:

The first of these is interesting and subtle: How many times have you ssh'd to someone else's machine, and then scp'd a file back to yourself ("pushing" it back to yourself)? Well, don't do that. Instead, scp it in "pull" mode from your own machine's command line:

$ scp username@remotehost:/tmp/somefile .

...rather than this form on remotehost's command line:

$ scp /tmp/somefile username@myhost:

Why? This gets back to the problem of stolen tokens, again: When you initiate the scp from remotehost to "push" the file back to where you are on myhost, you have to provide a stealable security token on a machine you don't control and have no reason to trust. "Pulling" the file from myhost poses no such risk.

Hardly anyone follows Wichert's second recommendation (unique passwords) because good passwords are too difficult to remember. The human brain isn't wired to support that sort of data retention. However, if you care enough about the problem, you can use my solution of Keyring for PalmOS, an "electronic wallet" for security tokens that stores them all in a 3DES-encrypted database, unlockable with a single password, so you need remember only that one.

I would add to Wichert's recommendations:

The debian.org admins, as it turned out, didn't strictly need an IDS to know their machines had been compromised: They noticed the suspicious pattern of kernel "oopses", did a small amount of checking, and immediately drew the right conclusion. The nightly report from AIDE served mainly to confirm what they already knew. In general, an alert sysadmin is by far your best protection.

Security in general is a tough problem. Screw-ups and people shooting you in the foot are endemic, and meaningful improvement comes at a cost in inconvenience. I've barely scratched the surface of threat models that should be of concern — and there are other checking tools such as chkrootkit that are worth using. But I hope I've outlined some of the low-hanging fruit that yields the biggest improvements in areas that matter.


Christophe Lupe's post about synergy between AIDE and Tripwire:

Moen's Laws and other lexicon items:

Wichert Akkerman's Debian.org Compromise 2003 pages:

Nmap, a free open source utility for network exploration or security auditing:

Snort, an open-source IDS of sorts (but networked, not host-based):

Logcheck, a script to detect anomalous logged events and mail the sysadmin:

Tripwire, the original, classic host-based IDS. Notice that, although Tripwire is self-checking, it has the problem in common with all other host-based IDSes that intruders may disable or tamper it, to sabotage its protection. However, because every part of it, right down to the nightly reports, is cryptographically signed, it has the advantage of being extremely tamper-evident: If you ever fail to receive it nightly report, or get one that fails to validate as genuine, then you immediately know something's up. Having it check all files of your other IDS(es) further extends this advantage to those.

Tripwire note:
If running this verification regime on a suspect host strikes you as precarious, you're probably correct — and Tripwire, Inc. recommends that, at a minimum, you verify Tripwire files using the siggen utility provided for that purpose, and preferably store them on read-only media. Adjust to suit your level of paranoia (e.g., recompiling components using static linking, etc.).

AIDE, the younger challenger:

Integrit, a similar newcomer:

Samhain, a similar newcomer that's said to be exceptionally good. A truly careful admin would run two lightweight IDSes, such as AIDE and Samhain, and have Tripwire check them both, in order to avoid having one IDS's flaws be a single point of failure:

Prelude-IDS, another newcomer:

OPIE (One-time Password In Everything) and OpenSSH, via pam_opie module:

S/Key and OpenSSH:
As with OPIE, you may need to recompile OpenSSH to ensure support:



pilOTP (proprietary):

Keyring for PalmOS:

Chkrootkit examines your system for common, known software toolkits used to conceal an intruder's presence after break-in ("rootkits"). As such, it gives only negative reassurance of "No, I don't see any of the signs I believe indicative of rootkits my designer taught me to look for", and in that sense is similar to a virus checker. Inherently, it cannot actually rule out the presence of rootkits it doesn't know about, let alone the intruders themselves.

Jim Dennis's Security Tips page has many further ideas:

Linuxmafia Knowledgebase (my PerlHoo documentation tree) also has further resources:

[BIO] Rick has run freely-redistributable Unixen since 1992, having been roped in by first 386BSD, then Linux. Having found that either one sucked less, he blew away his last non-Unix box (OS/2 Warp) in 1996. He specialises in clue acquisition and delivery (documentation & training), system administration, security, WAN/LAN design and administration, and support. He helped plan the LINC Expo (which evolved into the first LinuxWorld Conference and Expo, in San Jose), Windows Refund Day, and several other rabble-rousing Linux community events in the San Francisco Bay Area. He's written and edited for IDG/LinuxWorld, SSC, and the USENIX Association; and spoken at LinuxWorld Conference and Expo and numerous user groups.

His first computer was his dad's slide rule, followed by visitor access to a card-walloping IBM mainframe at Stanford (1969). A glutton for punishment, he then moved on (during high school, 1970s) to early HP timeshared systems, People's Computer Company's PDP8s, and various of those they'll-never-fly-Orville microcomputers at the storied Homebrew Computer Club -- then more Big Blue computing horrors at college alleviated by bits of primeval BSD during UC Berkeley summer sessions, and so on. He's thus better qualified than most, to know just how much better off we are now.

When not playing Silicon Valley dot-com roulette, he enjoys long-distance bicycling, helping run science fiction conventions, and concentrating on becoming an uncarved block.

Copyright © 2004, Rick Moen. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004

Python Simplicity

By Mike Orr (Sluggo)

This article was inspired by Brian Dorsey, who hosted a SeaPIG meeting last month. (SeaPIG is the Seattle Python Interest Group.) As I was perusing the bookshelves at his house, I saw that his books on the Simplicity Movement (see below) match his fanaticism for simplicity in programming. Brian is a big-shot database administrator. (At one of Paul Allen's companies, boo, hiss.) He knows more SQL than most people I know. You'd think that means he loves complexity. Don't all database administrators love complexity? (It keeps them employed, after all.) But Brian stunned me by revealing his enthusiasm for trying out all the little Python database modules that are uncomplex. He follows the 80/20 rule: sometimes 20% of the features solve the problem for 80% of the population. Brian has also been demonstrating several other simple modules over our last several Python meetings, so I'd like to share with you a few of those.

In most articles, the author walks through code he's worked with extensively. (Or pretends he has worked with extensively.) In this article, I'm discussing modules I have not used. The point is not to say these modules are the best thing since sliced bread (although some of them are nifty). The point is that these modules demonstrate simplicity, either in their code or in their use.

By in their code, I mean the module itself is short. Less code means less chance for things to go wrong, so more reliability. That's the KISS principle: "Keep it Simple, Stupid!" In their use refers to the user's perspective: it takes only a few lines to activate their features. Some modules are simple in one way or the other, while others are simple in both.

Doc XML-RPC Server

Doc XML-RPC Server has got to be one of the easiest ways ever to offer services on the Internet. It's one of those inventions that makes you bonk your head and think, "Why didn't somebody think of this sooner?" Say you've written your services as methods of a certain class. You want to offer these services on the Internet or on an intranet. It takes just six lines of code:

from DocXMLRPCServer import DocXMLRPCServer
server = DocXMLRPCServer(('', 8000), logRequests=0)
'SimpleShareServer' is a class we created. We start a server on port 8000, register an instance of our class, register some optional services that come with DocXMLRPCServer ("introspection functions"; e.g., 'help'), and away we go. Here's the services we're offering:
import time
class SimpleShareServer:
     def message(self, msg):
         """message('Print me!') => True 
         Log everything passed to this function"""
         print time.asctime(), msg
         return True
     def wait(self, seconds):
         """wait(5) => 5 
         Wait for a certain number of seconds before returning.
         Returns the same number passed in."""
         print time.asctime(), "Waiting %s seconds" % seconds
         print time.asctime(), "Finished waiting %s seconds" % seconds
         return seconds
Of course, a local routine can just instantiate the class and call the methods directly. To do the same remotely, you used to have to either write a custom protocol implementation yourself, or read a long reference books to configure an off-the-shelf server or library. But a remote user can access these services with just a couple lines of code:
import xmlrpclib
s = xmlrpclib.ServerProxy('http://localhost:8000')
s.message("Hello, simple world!")
result = s.wait(15)
After these lines have executed, 'result' is 15, and "Hello, simple world!" appears on the server's console (standard output). Note that the arguments and return value were passed seamlessly between client and server, just like invoking a local method. The server proxy object "stands in" for the remote instance. Note that the client is using a generic XML-RPC library; it doesn't have to use a library specific to DocXMLRPCServer.

RPC (Remote Procedure Call) has been around on Unix systems for decades. NFS uses it, for example. But plain RPC (so I'm told) cannot cross programming languages. If the server is Python, the client has to be Python too, or something that knows how to encode/decode Python argument types. XML-RPC removes the language restriction. The arguments are converted to language-neutral XML, and so is the return value. This has some limitations:

In exchange for these limitations, Perl and other clients can access your server -- just like any web browser can access an HTTP server.

Speaking of HTTP, that's the niftiest part of DocXMKLRPCServer. If a client sends an HTTP POST request to the same port, the server recognizes it and translates it to the corresponding method call. This could be used to collect form submissions for a survey, for instance. If a client sends an HTTP GET request, the server responds with an HTML page documenting itself. You've heard of emacs, the Self-Documenting Editor? Here's the self-documenting arbitrary server. Your server class can define three extra methods to customize the documentation output:

set_server_title(STRING)         # For the <TITLE> tag.
set_server_name(STRING)          # For the <H2> header.
set_server_description(STRING)   # The documentation, in HTML format.

DocXMLRPCServer is built on top of SimpleXMLRPCServer, which provides everything except the HTTP ability.

Somebody might object, "But it's using XML, and XML is decidedly non-simple." This is true. XML is a horrible bastard beast that should never have seen the light of day. In theory, it's wonderful. In practice, most of the DTD's are are so unnecessarily complex and the namespaces so nitpickily detailed that it looks like something only a bureaucracy could have designed -- the union of all attributes lobbied by every single special-interest group. You have to trust that the expat parser or whatever it's using under the hood won't blow up someday. So DocXMLRPCServer isn't simple in the code it depends on. But it's simple to use. Did you see any XML above? I didn't. I'm all for using XML if you don't have to look at it. Like the way Elvis impersonation bands are fun to attend as long as you keep your back to the stage, so you can enjoy the music without having to look at the tacky 50s kitsch. But I digress....

The SimpleShareServer above is based on a server Brian demonstrated at a SeaPIG meeting, which he describes on a wiki page.


db_row is a short module to wrap a SQL result set. The DB API database modules (MySQLDb, several PosgreSQL modules, Oracle and others) return a query row as a tuple of column values. Let's fool it with our own tuple and see what it does.

tup = (1, 2, 3)     # E.g., "SELECT a, b, a+b FROM SomeTable WHERE id=456;"
R = IMetaRow(['a', 'b', 'sum'])
                    # Create a custom class that names the rows in order.
                    # IMetaRow is a "class factory": it creates a class.
r = R(tup)          # Instantiate our custom class.
print r[0], r['b'], r.fields.sum
                    # Prints "1 2 3".  Access values by subscript, key or
                    # attribute.  (The "I" in IMetaRow means case-insensitive.)
print r.keys()      # Look ma, dictionary methods!
print r.dict()      # Just give me a real dictionary, please.

To convert an entire multi-row result set to a list of such jobbies, use a list comprehension:

lis = [ R(row) for row in cursor.fetchall() ]
print lis[0]['a'], "+", lis[0]['b']
print lis[0].fields.sum
Wrap the list comprehension in a function, and you only have to see it once.

Why do I like this module? It's short. You don't have to wait for it to be incorporated into your favorite DB API module; it works with all of them already. It works with non-SQL and ad-hoc result sets too. It solves a common problem in a simple way. (It's not all that simple. It uses Python slots, for instance. But we'll ignore that and hope Python's obscure slots feature has had most of its early bugs ironed out.) It claims to use less memory than a list of dictionaries.

But db_row's simplicity does come at the cost of certain disadvantages. It has no knowledge of the database field names or data types. You can lie to IMetaRow() and rename the fields anything you want. That may be convenient in some situations, but in an application with more than a few tables, it can get out of hand pretty quickly. Confusing yourself (and future maintainers) with inconsistent field names is a decidedly unsimple idea. Or doubleplus ungood as Orwell would say.

(The example above was inspired by db_row's docstring.)

SQLite and pysqlite

SQLite is an entire SQL server encoded in a little C library. pysqlite is a Python wrapper (DB API compatible). Brian calls this combination, "80% of what you'll ever need a database for in a single 270K executable (or Python module)."

The "80% you need" is ACID-compliant transactions, basic data types (strings, numbers, BLOBs, DateTimes), auto-increment fields, NULLs, temporary tables ("CREATE TEMPORARY TABLE"), a command-line utility (à la mysql and psql), dumping a database to SQL statements (PosgreSQL compatible), and huge databases (2 terabytes). There's even support for concurrent access of the database file in multiple processes, which I was pretty amazed at. You can't store strings that contain null characters (0 decimal) though. The database schema is stored in a table called 'sqlite_master'. Security is done by file permissions.

Another quirk of SQLite is typelessness. You can put letters into a numeric field, whatever that means. Actually, it means that a database is meant to store data, not to impose its will on the data. The SQLite developers call the rigid type system in the SQL standard and in most implementations a misfeature. (See the Datatypes page in the SQLite documentation for the full justification.) Fields can be created with all the usual SQL type specifiers, but those are just hints to the user, not rules SQLite enforces. Actually, SQLite does honor the types to some degree: they influence the sort order and whether two values are identical. There is one exception to SQLite's permissivism: auto-increment fields ("INTEGER PRIMARY KEY") have to be integers.

Other modules

There are a few object-oriented wrappers for SQL access, including SQLObject and DBO.

ctypes is a way to call C libraries directly from Python, which is supposedly easier than SWIG.

Python built-in features

Python 2.2 introduced three features that took people a bit of time to get their heads around, but they turned out to be incredibly useful: iterators, generators and properies. Iterators let you have a for-loop without having to pregenerate the entire sequence of values and keep them all in memory simultaneously. Generators allow an easy way for a function to iterate: it dispenses with the "topmost for-loop", leaving you more horizontal screen space and less clutter. Properties allow you to define "smart" attributes: those that trigger an action when they're get or set. Properties are controversial to some purists, but they avoid the clumsiness of accessor methods (aka parenthesesitis, which is a serious disease among C/Java-phobics).

Python 2.3 continues the trend with more features that simplify your programs. Sets are like dictionaries without the values ("just the keys, please"). If you're using a dictionary only to weed out duplicates, why define "values" you're not going to use? There's a logging module and a simple DateTime object. But the thing I use most is enumerate():

>>> lis = ['vanilla', 'chocolate', 'strawberry']
>>> for i, element in enumerate(lis)::
...    print "Element %d is %s." % (i, element)
Element 0 is vanilla.
Element 1 is chocolate.
Element 2 is strawberry.
This is a long-requested feature that avoids the equivalent but clumsier:
>>> for i in range(len(lis)):
...     element = lis[i]
...     print "Element %d is %s." % (i, element)
Element 0 is vanilla.
Element 1 is chocolate.
Element 2 is strawberry.

The simplicity movement

"And now for something completely different..."

The simplicity movement, championed by authors like Amy Daczyczyn (author of The Tightwad Gazette, a paper zine), Joe Dominguez & Vicki Robin, Cecile Andrews, Elaine St James and others, is about deciding what you really want from life and which material posessions really matter to you. Keep the stuff you need or want (e.g., for a hobby), and get rid of the stuff that's not a priority so it's not a distraction. This may not seem like it has much to do with programming, but we'll see that it does. Here's a few gems in the theories:

There are two ways toward a higher standard of living: earn $100 more per month, or cut your expenses by $100 per month. Buth achieve exactly the same thing: $100 more in your pocket. Most people adopt the former strategy, but that means depending on somebody else: you have to convince them to give you the money. In contrast, cutting expenses is entirely under your control. Having both spouses working means more expenses for transportation, clothing, food, daycare and unwinding; are you sure your net income is really higher than it would be without that second job? What about the lost opportunity for the second spouse to pursue a hobby or be a full-time volunteer? I love my freedom in not having a car; it gives me enough money to travel a couple times a year. Sure it limits where I can live and work, but those are the places I want to be anyway.

Then there's the question of technology. The Amish may be a bit too luddite for most people's taste, but they have a good point: accept new technology carefully, and only when it's proven its worth. I love my cell phone, but my stereo looks like it came from 1987 (which it did).

This feeds right into environmental sustainability, and the theory of waste. Why pay for stuff you don't want (and nobody wants)? Did you buy the applesauce for the applesauce itself or for the aluminum can it came in? Did you buy it because it has an extra plastic seal at the top? Did you buy it because of the energy used and effluent spent to produce the can? I can't discuss all this properly here, but there's a book, Natural Capitalism (Lovins, entire text online at www.natcap.org), that's easily the most important book of the 21st century so far. It looks at the question of waste from the individual's, businessman's, and policymaker's perspective, and how the (US) accounting and tax system allows companies to externalize the cost of environmental cleanup, which falsely skews their profit/loss statements and stock prices. But it takes only a change in business model to begin eliminating waste, work with the environment rather than against it, and turn a greater profit at the same time. Good stuff, Maynard.

What does all this have to do with simplicity in programming? The principles are the same. Decide what you really want, and look for a tool that does that. Maybe SQL is the cat's meow, but do you really need all the features of MySQL or PosgreSQL? Maybe you do, but it's reassuring to have thought out exactly which features you need and why you need them. (Especially when Postgres segfaults and you're wondering, why did I choose this?) Or maybe SQL isn't the cat's meow, and an object database like ZODB, or something even lighter weight like DBM or pickle/shelve might do the job.

Brian's Marklarizing webproxy

To conclude this article, I have to mention Brian's funniest invention. Here's the wiki entry describing it:
Several of my friends have a running Marklar joke spawned from a South Park episode. (Example. For Marklar for the episode review click here and search for 'Marklar'.) Short version: There are aliens who use the word marklar for every noun - and no, it's not confusing. :) After a long weekend of Marklar overdose, I happened to run into a free word list which included parts of speech (Greg Ward's Moby). Something clicked in my brain and I decided I had to make a Marklarizing web proxy, so that the entire internet could be seen as a Marklar would see it. Eight hours and some pretty horrible code later (mostly on the proxy & HTML parsing side, but the libraries I used and the program I wrote are only a couple pages each), I had something that mostly worked. Anyway, I demo'd it at the meeting, and this was our favorite page. It's an article from a newspaper.

Two arrested for running marklar-end marklar marklar 
By Marklar Ko 
Marklar Marklars staff marklar 
Marklar County marklar's detectives have broken up a large marklar marklar  
in the Marklar marklar, and they said they've recovered a "black book" with  
the names of hundreds of marklars, including men who work for marklar marklars  
headquartered in Marklar. 
The two marklars of the marklar, a 49-marklar-old woman who lives in Marklar,  
and her 31-marklar-old marklar, who lives in Marklar, were arrested. They have  
not been charged. 
The two marklar used the Internet to advertise a marklar-end escort company  
called the "Marklar of Marklar." The Web site had pictures of available  
companions, a calendar of when they were available and marklars. Marklars  
could be made online. 
Some of the escorts were brought in from out of state "Las Marklars, New  
Marklar and Los Marklars" to work marklar, according to the Marklar's Marklar. 
Marklars were carefully screened, said marklar's Sgt. Marklar Marklar. 
For example, potential marklars had to leave a work number, and someone inside  
the marklar marklar would marklar the number, marklar it was to confirm a  
dental marklar, he said. 
This was done to make sure the marklar wasn't a police officer, Marklar said. 
The Marklar's Marklar marklar not release the names of the marklars the marklars  
worked for, to avoid tainting the companies, Marklar said. 
The men in the book, however, marklar likely be contacted soon, detectives said.  
They could face marklar charges of patronizing a prostitute. 
In the 1990s, the two marklar were involved in another marklar marklar in the  
marklar called "Affluent Marklars." 
An marklar tipped off the Marklar's Marklar about the Marklar of Marklar.

picture Mike is a Contributing Editor at Linux Gazette. He has been a Linux enthusiast since 1991, a Debian user since 1995, and now Gentoo. His favorite tool for programming is Python. Non-computer interests include martial arts, wrestling, ska and oi! and ambient music, and the international language Esperanto. He's been known to listen to Dvorak, Schubert, Mendelssohn, and Khachaturian too.

Copyright © 2004, Mike Orr (Sluggo). Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004

Mathematical Explorations with Scilab/Linux

By Pramode C.E.

Little would Jean Baptiste Joseph Fourier, the 18th century French mathematician and revolutionary, have imagined that the analytical techniques he had invented to study the behaviour of mathematical functions would someday become one of the most powerful tools in the hands of scientists and engineers working in disciplines as diverse as neurophysiology and digital communication.

As I was fast sliding into the depths of mathematical ignorance, I thought maybe I would refresh some high school memories by trying to understand a bit of Fourier's math. Much of what I read flew far above my head - my only consolation was that I discovered Linux to be an ideal platform not only for Operating System hacking but also for mathematical recreation and research.

I came upon a great tool called Scilab and also a nice little tutorial on Fourier Math by Chris Meyers which demonstrated some interesting sine-wave combination/analysis stuff using Python code. This article demonstrates a few simple Scilab tricks and reimplements Chris's code in Scilab's native scripting language. Readers looking for mathematical wisdom are warned not to rely too much on what I say here!

What is Scilab?

Scilab is a powerful, free environment for mathematical computation. It provides an extensible framework for general matrix manipulation and `toolboxes' for doing stuff like control system design, digital signal processing etc. The C/Fortran source code is available for download from the project home page - I had absolutely no difficulty in building the system - the standard `configure; make; make install' magic worked perfectly.

Here is a screen shot of Scilab running on my Linux box:

Simple math

Let's get started by doing a few simple matrix manipulations. A 3-by-3 matrix is created by simply typing, at the Scilab prompt:

-->a = [1,10,20; 5,6,7; 12,11,45]
 a  =
!   1.     10.    20. !
!   5.     6.     7.  !
!   12.    11.    45. !
It's easy to get the transposed matrix:
ans  =
!   1.     5.    12. !
!   10.    6.    11. !
!   20.    7.    45. !

A few other functions:

-->sum(a, 'c')
 ans  =
!   31. !
!   18. !
!   68. !
-->sum(a, 'r')
 ans  =
!   18.    27.    72. !
 ans  =
!   1.  !
!   6.  !
!   45. !

Elements can be extracted from matrices in many different ways - the simplest is the standard indexing procedure. Writing a(1,2) would yield the element at row 1 and column 2 (note that the index starts at 1). Indexing a matrix beyond its bound will result in an error. Writing to a non-existent index will result in the matrix growing dynamically.

-->a(3,4) = 3
 a  =
!   1.     10.    20.    0. !
!   5.     6.     7.     0. !
!   12.    11.    45.    3. !

The 'colon' operator

The 'colon' is a cute little operator. We can create a vector of numbers 1,2,3 ... 10 by just writing:

-->a = 1:10
 a  =
!   1.    2.    3.    4.    5.    6.    7.    8.    9.    10. !

Many other tricks are possible:

 b  =
!   1.    2.    3. !
!   4.    5.    6. !
!   7.    8.    9. !

ans  =
!   2.    3. !
!   5.    6. !
!   8.    9. !
 ans  =
!   1.    3.    5.    7.    9. !

Note that 1:2:10 means create a vector starting from 1, each successive element being computed by adding 2, until the value becomes greater than 10.

Simple plotting

Let's look at an example of a simple sine wave plot. We want one full cycle of the sine curve (from 0 to 2*PI) - let's take 240 points in between, so each division would be 2*PI/240. First we create a vector containing all the angle values in this range and then we plot it (%pi is a constant standing for the value of PI):

--> = 0:(2*%pi)/240:2*%pi
 x  =
         column 1 to 5
!   0.    0.0261799    0.0523599    0.0785398    0.1047198 !
         column 6 to 9
!   0.1308997    0.1570796    0.1832596    0.2094395 !
         column 10 to 13
!   0.2356194    0.2617994    0.2879793    0.3141593 !
         column 14 to 17
!   0.3403392    0.3665191    0.3926991    0.4188790 !
[More (y or n ) ?] 

Now, we use a simple plot function:

-->plot(x, sin(x))

Writing Scilab scripts

Writing Scilab scripts is simple. Here is an example of a 'for' loop which can be entered at the Scilab prompt itself:

-->s = 0         
 s  =
-->for i=1:3:10
-->  s = s + i
 s  =
 s  =
 s  =
 s  =
[More (y or n ) ?] 

Defining functions

The function definition syntax is a wee bit tricky. Here is a simple example:

-->function [r] = my_sqr(x)
-->  r = x * x
 ans  =

After the keyword 'function', we supply a list of `output values'. Any value written to an `output' value will be `returned' by the function. The argument 'x' is of course the input argument to the function. The function returns the value 'r' which is the square of 'x'.

The question obviously is what if we want to return two values. We try the following at the Scilab prompt:

-->function [r1, r2] = foo (x, y)
-->  r1 = x + y
-->  r2 = x - y
-->[p, q] = foo(10, 20)
 q  =
  - 10.  
 p  =

Note the special way we call the function. The value of r1 will get transferred to 'p' and value of r2 to 'q'.

The following invocations of 'foo' demonstrates the fact that the language is dynamically typed.

-->[p, q] = foo([1,2], 1)
 q  =
!   0.    1. !
 p  =
!   2.    3. !
-->[p, q] = foo([1,2], [3,4,5])
 !--error     8 
inconsistent addition
at line       2 of function foo   called by :  
[p, q] = foo([1,2], [3,4,5])

It is possible to store function definitions in a file and load them at a later time. Suppose the above function definition is stored in a file called 'fun.sci'. We need to simply invoke, at the Scilab prompt:


Enter Fourier!

We encounter 'signals' everywhere. The PC speaker generates sound by converting electrical signals to vibrations. We see objects around us because these objects bounce back light signals to our eyes. Our TV and radio receive electromagnetic signals. We are immersed in a 'sea of signals' ! Analysis of signals is therefore of central importance in most branches of science and engineering.

The basic Unix philosophy is `Keep it Simple, stupid'. Physicists (and most other scientists and engineers) often can't stick to this dictum when they start analysing stuff, simply because the phenomena they are studying have awesome complexity. But it seems that most complex things in this world can be explained on the basis of simpler things. Joseph Fourier's insight was that complex time varying signals can be expressed as a combination of simple sin/cos curves of varying frequency and amplitude. We will verify this assumption by plotting a few simple equations with the help of Scilab.

Let's start with a simple sum of two 'sin' signals.

-->delta = (2*%pi)/240
 delta  =

-->x = 0:delta:2*%pi

-->a = sin(x) - (1/2)*sin(2*x)
-->plot(x, a)
Here is the plot:

There is very little indication here that something interesting is going to happen. Next, we try plotting.

b = sin(x) - (1/2)*sin(2*x) + (1/3)*sin(3*x)
We keep on adding terms to the series, the next term would be -(1/4)*sin(4*x), the next one +(1/5)*sin(5*x) and so on. Here is what I got when I plotted this series with 200 terms in it (you will have to write a function to do this for you):

Seems like magic! The sin curve has vanished completely and we have a brand new signal! How exactly Mr.Fourier 'knew' such a series would ultimately give us something totally different from the sum of its parts would be more appropriately dealt with in a mathematics class(Do I hear you yawn? Do we have a case for a more `practical' math education with students being given access to Linux boxes running Scilab, Python(Numeric), and a whole lot of other free, educational tools?)

Determining the components of a signal

We have seen that adding together sines of different frequency and amplitude gives us signals which look totally different. Now the question is, given some numbers which represent a complex waveform, will we be able to say what combination of sine's (frequency and amplitude) gave rise to that particular signal? Let's try.

Let's first write a function which performs simple numerical 'integration' over the range 0 to 2*PI. We divide the area under our curve into tiny strips, each of width say 2*PI/240. The area of a strip at point 'x' (0 < x < 2*PI) will be its height multiplied by the width, which will be sin(x) * (2*PI/240). This is the idea behind the integration function, which can be typed at the Scilab prompt. The argument to integrate is a vector of sin values in the range 0 to 2*PI-delta where delta is (2*PI)/240. The difference between two successive values in the vector is 'delta'.

-->function [r] = integrate(a)
-->  r = sum(a)*(2*%pi)/240

Let's try integrating the simple sin function, sin(x).

-->x = 0:delta:(2*%pi-delta)

 ans  =
We see that the integral is zero. The sin curve has equal area above and below the zero-point.

Let's try plotting sin(x).*(-sin(x)) (Note that the .* operator performs memberwise multiplication of two vectors):

We see that the function has been shifted completely below the zero-point. It should now definitely have a non-zero area.

 ans  =
  - 3.1415927  

Scilab tells us it is -PI. Let's now try plotting sin(2*x).*(-sin(x))

The graph tells us that the integral should be zero. We verify this:

 ans  =
We are now beginning to get a 'feel' of the idea we would employ to separate out the components of our complex signal. Multiplying a sine with negative of a sine of a different frequency gives us zero - only when the frequencies match do we get non zero results. Say our complex signal is:
sin(x) - (1/2)*sin(2*x) + (1/3)*sin(3*x) - (1/5)*sin(5*x)
If we multiply this with -sin(x), what we get is:
sin(x).*(-sin(x)) - (1/2)*sin(2*x).*(-sin(x)) + 
(1/3)*sin(3*x).*(-sin(x)) - (1/5)*sin(5*x).*(-sin(x))
The first term gives us -PI, all other terms become zero. The fact that we are getting a non zero value tells us that sin(x) is present in the signal. Now we multiply the signal with -sin(2*x). If we get a non-zero result, that means that sin(2*x) is present in the signal. We repeat this process as many times as we wish.

How do we get the amplitude of each component? Let's try out another experiment:

-->b = sin(x) - (1/2)*sin(2*x) + (1/3)*sin(3*x) - (1/4)*sin(4*x)
 ans  =
  - 3.1415927  
 ans  =
 ans  =
  - 1.0471976  
 ans  =
We see that dividing each result by -PI gives us the amplitude of each component of the signal.


Very high quality proprietary tools exist for doing numeric/symbolic math - but they are sometimes priced beyond the reach of the student or the hobbyist. I hope this article has convinced you that Free Software alternatives do exist. Kindly let me know about any inaccuracies you find in this document. I can be contacted via my home page at pramode.net.


Thanks to the Scilab team for creating such a wonderful tool and also documenting it thoroughly. This article would not have been written without the help of Chris Meyers document explaining Fourier's math - Chris has also written some other very interesting Python programs which you are sure to enjoy. A big Thank You to him!

[BIO] As a student, I am constantly on the lookout for fun and exciting things to do with my GNU/Linux machine. As a teacher, I try to convey the joy of experimentation, exploration, and discovery to my students. You can read about my adventures with teaching and learning here.

Copyright © 2004, Pramode C.E.. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004

The Wonderful World of Linux 2.6

By Joe Pranevich

[Reprinted from http://kniggit.net/wwol26.htm with the kind permission of the author.]

Although it seems like only yesterday that we were booting up our first Linux 2.4 systems, time has ticked by and the kernel development team has just released the 2.6 kernel to the public. This document is intended as a general overview of the features in the new kernel release, with a heavy bias toward i386 Linux. Please also be aware that some of the "new" features discussed here may have been back-ported to Linux 2.4 after first appearing in Linux 2.6, either officially or by a distribution vendor. I have also included information on a handful of cases where a new feature originated during the maintenance cycle of Linux 2.4, and those will be marked as appropriate in the text.

At present, this document has been translated into ten languages. Please see the "Translations" section at the very bottom for more information.

The Story So Far...

The Linux kernel project was started in 1991 by Linus Torvalds as a Minix-like Operating System for his 386. (Linus had originally wanted to name the project Freax, but the now-familiar name is the one that stuck.) The first official release of Linux 1.0 was in March 1994, but it supported only single-processor i386 machines. Just a year later, Linux 1.2 was released (March 1995) and was the first version with support for different hardware platforms (specifically: Alpha, Sparc, and Mips), but still only single-processor models. Linux 2.0 arrived in June of 1996 and also included support for a number of new architectures, but more importantly brought Linux into the world of multi-processor machines (SMP). After 2.0, subsequent major releases have been somewhat slower in coming (Linux 2.2 in January 1999 and 2.4 in January 2001), each revision expanding Linux's support for new hardware and system types as well as boosting scalability. (Linux 2.4 was also notable in being the release that really broke Linux into the desktop space with kernel support for ISA Plug-and-Play, USB, PC Card support, and other additions.) Linux 2.6, released 12/17/03,  stands not only to build on these features, but also to be another "major leap" with improved support for both significantly larger systems and significantly smaller ones (PDAs and other devices.)

Core Hardware Support

One of the most important strengths of Linux-powered operating systems is their flexibility and their ability to support a wide range of hardware platforms. While this document is geared specifically to the uses of Linux on PC-derived hardware types, the Linux 2.6 kernel has made some remarkable improvements in this area that deserve to be pointed out.

Scaling Down -- Linux for Embedded Systems

One of the two most fundamental changes to Linux in 2.6 comes through the acceptance and merging of much of the uClinux project into the mainstream kernel. The uClinux project (possibly pronounced "you-see-Linux", but more properly spelled with the Greek character "mu") is the Linux for Microcontrollers project. This variant of Linux has already been a major driver of support for Linux in the embedded market, and its inclusion in the official release should encourage further development in this space. Unlike the "normal" Linux ports that we are generally accustomed to, embedded ports do not have all the features that we associate with the kernel, due to hardware limitations. The primary difference is that these ports feature processors that do not feature an MMU. ("memory management unit" - what makes a protected-mode OS "protected") While these are generally true multitasking Linux systems, they are missing memory protection and other related features. (Without memory protection, it is possible for a wayward process to read the data of, or even crash, other processes on the system.) This may make them unusable for a multi-user system, but an excellent choice for a low-cost PDA or dedicated device. It is difficult to over-emphasize this architecture shift in Linux 2.6; all versions of Linux up to this point were derived (however indirectly) from the limitations inherent with Linus' initial work on his Intel 80386.

There are several new lines of embedded processors supported by Linux 2.6, including Hitachi's H8/300 series, the NEC v850 processor, and Motorola's line of embedded m68k processors. Motorola's offerings are the most familiar to the average Linux user as they are the guts underneath Palm Pilots starting with the first (the Palm 1000), up until the Palm III. Other models go by names such as Dragonball and ColdFire and are included on systems and evaluation boards manufactured by Motorola, Lineo, Arcturus, and others. Sadly, support for other, older m68k processors without MMUs (such as the 68000s used in early Macintoshes) is not yet covered by the new release but it is highly possible that "hobbyist" projects of the future will seek to support Linux on these and other antique system.

Although not a part of the uClinux merge, the new revision of Linux also include support for Axis Communications' ETRAX CRIS ("Code Reduced Instruction Set") processors. (Actual support for this processor arrived as a feature during the 2.4 kernel's maintenance cycle -- well after the 2.4.0 release-- so it deserves a brief mention.) These are embedded processor, but with MMUs, that is primarily used in network hardware. Related support for MMU-less variants of these processors has not yet been accepted into the kernel, but are being worked on by outside projects.

In addition to pure hardware support, there have been a number of other wins through the integration of the embedded work into the mainline kernel. While most of these changes are under the hood, changes such as ability to build a system completely without swap support add to the overall robustness of the OS.

Scaling Up -- NUMA and Bigger Iron

The second of the two most fundamental changes in Linux 2.6 happens to work in the other direction: to make Linux a more acceptable kernel on larger and larger servers. (Some of these larger servers will be i386 based, and some not.) The big change in this respect is Linux's new support for NUMA servers. NUMA (or "Non-Uniform Memory Access") is a step beyond SMP in the multi-processing world and is a major leap forward for efficiency on systems that have many processors. Current multiprocessing systems were designed with many of the same limitations as their uniprocessor counterparts, especially as only a single pool of memory is expected to serve all processors. On a many-processor system, there is a major performance bottleneck due to the extremely high contention rate between the multiple cpus onto the single memory bus. NUMA servers get around that difficulty by introducing the concept that, for a specific processor, some memory is closer than others. One easy way (and not terribly technically incorrect) to imagine this is that you have a system built with separate cards, each containing CPUs, memory, and possibly other components (I/O, etc.) There are many of these cards in a system and while they can all talk to each other, it's pretty clear that the CPUs will have the easiest time talking to the local memory (the memory on the cpu card rather than on a separate card.)You can imagine the new NUMA architecture being somewhat similar to a very tight-knit cluster at the lowest levels of hardware.

To properly support these new NUMA machines, Linux had to adapt in several respects to make the new model efficient. To start with, an internal topology API was created to actually let the kernel internals understand one processor or one memory pool's relations to I/O devices and each other. Derived from that, the Linux process scheduler now is capable of understanding these relationships and will attempt to optimize tasks for best use of local resources. Additionally, many NUMA machines are built in such a way that they have "holes" in the linear memory space "between" nodes. The new kernel is able to deal with those discontiguous cases in a reasonable way. There are many other internal changes which were made to allow Linux to support these new high-end machines, and this is definitely an area of growth for the kernel as a whole. However, this is an area where Linux is very rapidly growing and maturing and much work remains to be done to make the most efficient use of resources possible. Over the course of the next year, we can expect to see many more improvements in Linux's support for these very high-end systems.

Subarchitecture Support

While not quite as core as the two previous changes, the new revision of the kernel also includes a new concept called a "subarchitecture" which further extends Linux's reach into new hardware types. Previously, Linux often had the underlying assumption that processor types and hardware types went hand in hand. That is, that i386-descendant processors are only used on PC/AT-descendant servers. In Linux 2.4, this assumption was broken for i386 with the addition of support for SGI's Visual Workstation, a "legacy-less" platform running with an Intel chip. (And in fact, it was broken long before on many other architectures. For example, m68k has long supported Amigas, Macintoshes, and other platforms.) The big change in Linux 2.6 is that this feature and concept was standardized so that all architectures handle this in a similar and saner way that allows for more clear separation of the components that need to be separated.

With this standardization comes two new platforms to support for i386. The first is NCR's Voyager architecture. This is a SMP system (developed before the now-standard Intel MP specification) supporting 486-686 processors in up to 32x configurations. The actual number of configurations that were sold with this architecture is relatively small, and not all machines are supported yet. (The oldest ones are unsupported.) The second architecture supported is the more widespread PC-9800 platform developed by NEC into the (almost) dominant PC platform in Japan until relatively recently. The original PC-9800 machines shipped with an 8086 processor and the line eventually evolved and matured (in parallel with the AT-descendants) until they featured Pentium-class processors and SMP support. (Of course, the support for Linux is limited to 386 or better.) Although completely unknown in the US, versions of Microsoft products up until Windows 95 were ported to run on this hardware. The line has been officially discontinued by the manufacturer in favor of more "standard" PCs.

By formalizing Linux's support for these "slightly different" hardware types, this will more easily allow the kernel to be ported to other systems, such as dedicated storage hardware and other components that use industry-dominant processor types. To be absolutely clear though, one should not take this subdivision too far. These subarchitecture have been separated because very low-level components of the system (such as IRQ routing) are slightly or radically different. This is quite different than running Linux on an X-Box, for example, where relatively little other than hardware drivers and some quirks separate the system from being a "generic" i386 system. Support for the X-Box would not be a subarchitecture.


Another major hardware advancement supported under Linux 2.6 is hyperthreading. This is the ability, currently only built into modern Pentium 4 processors but applicable elsewhere, allows a single physical processor to masquerade (at the hardware level) as two or more processors. This allows for performance boosts in some circumstances, but also adds scheduling complexity and other issues. Key in the kernel's improved support for this feature is that the scheduler now knows how to recognize and optimize processor loads across both real and virtual processors within a machine. In previous versions of Linux, it was quite possible to overwork a single processor because it was not possible to factor in the workload as a whole. One of the great things to note about this feature is that Linux was ahead of the market curve on supporting this new hardware feature transparently and intelligently. (Windows 2000 servers can see the additional faux-processors, but does not recognize them as virtual. Thus, you also require additional CPU licenses to take advantage of the feature. It was not until the Windows XP release that Microsoft completely supported this feature.)

Linux Internals

Scalability Improvements

In addition to the previously described generic features such as NUMA and hyperthreading, Linux 2.6 also has other changes for Intel servers at the top of the food chain. First and foremost is improved support for other new Intel hardware features including Intel's PAE ("Physical Address Extension") which allows most newer 32-bit x86 systems to access up to 64GB of RAM, but in a paged mode. In addition, IRQ balancing has been significantly improved on multiprocessor systems through major improvements to Linux's APIC support.

In addition to just supporting new hardware features, internal limits have been also increased when possible. For example, the number of unique users and groups on a Linux system has been bumped from 65,000 to over 4 billion. (16-bit to 32-bit), making Linux more practical on large file and authentication servers. Similarly, The number of PIDs (Process IDs) before wraparound has been bumped up from 32,000 to 1 billion, improving application starting performance on very busy or very long-lived systems. Although the maximum number of open files has not been increased, Linux with the 2.6 kernel will no longer require you to set what the limit is in advance; this number will self-scale. And finally, Linux 2.6 will include improved 64-bit support on block devices that support it, even on 32-bit platforms such as i386. This allows for filesystems up to 16TB on common hardware.

Another major scalability improvement in Linux 2.6 is that the kernel itself can now not only support more types of devices, but also support more devices of a single type. Under all iterations of Linux (and indeed, most UNIX-derived operating systems), users and applications running on a system communicate with the attached hardware using numbered device nodes. (The entries in the "/dev" directory.) These device nodes were limited to 255 "major" devices (generally, one type of device gets one or more device nodes) and 255 "minor" numbers (generally, specific devices of that type.) For example, the "/dev/sda2" device (the second partition on the first detected SCSI disk), gets a major number of 8, common for all SCSI devices, and a minor number of 2 to indicate the second partition. Different device types allocate their major and minor numbers differently, so it can't be said with assurance how many devices you can have on a Linux system. Unfortunately, this system breaks down badly on large systems where it would be possible, for example, to have many more than 255 of any specific device in a system. (Think large storage arrays, print farms, etc.) Under Linux 2.6, these limitations have been eased to allow for 4095 major device types and a more than a million subdevices per type. This increase should be more than adequate to support high-end systems for the time being.

Interactivity and Responsiveness

In addition to just scaling up, another priority with the new release has been to make the system more responsive. This is useful not only for the general desktop user (who always likes to see things respond quickly), but also to more time-critical applications where absolute preciseness is required to achieve the desired effect. Despite these changes, Linux 2.6 will not be a "hard" Real Time OS, which has very strict requirements for absolutely ensuring that actions happen predictably, but the overall responsiveness improvements should appeal to all classes of Linux users. (That said, there are external projects which have unofficial patches to provide RTOS functionality. Those projects could conceivably be made official in the next major release.)

One of the key improvements in Linux 2.6, is that the kernel is finally preemptible. In all previous versions of Linux, the kernel itself cannot be interrupted while it is processing. (On a system with multiple processors, this was true on a per-CPU basis.) Under Linux 2.6, the kernel now can be interrupted mid-task, so that other applications can continue to run even when something low-level and complicated is going on in the background. Of course, there are still times when the kernel cannot be interrupted in its processing. In reality, most users never saw these delays, which are rarely over small fractions of a second. Despite that, many users may notice an improvement in interactive performance with this feature enabled; things like user input will "feel" faster, even when the system is bogged down.

Linux's Input/Output (I/O) subsystems has also undergone major changes to allow them to be more responsive under all sorts of workloads. These changes include a complete rewrite of the I/O scheduler, the code of the kernel that determines what processes get to read from devices and when. The newly rewritten layer is now better capable of ensuring that no processes get stuck waiting in line for too long, while still allowing for the older optimizations which made sure that reading data still happens in the most efficient way for the underlying hardware.

On the application software side, another change that will help make Linux programs more responsive (if they use the feature) is support for new "futexes" (or "Fast User-Space Mutexes") Futexes are a way in which multiple processes or threads can serialize events so that they don't trample on each other (a "race condition"). Unlike the traditional mutex operations that most threading libraries support, this concept is partially kernel based (but only in the contention case) and it also supports setting priorities to allow applications or threads of higher priority access to the contested resource first. By allowing a program to prioritize waiting tasks, applications can be made to be more responsive in timing-critical areas.

In addition to all of the above, there have been a number of other smaller changes which will improve interactivity and performance in many cases. These include more removals of the "Big Kernel Lock" (non-fine-grained locks which were used in the early days' of Linux's support for multiple processors), optimizations of filesystem readahead, writeback, and manipulating small files, and other similar changes.

Other Improvements

Linux, like the Open Source movement in general, has always been a flag-bearer for the benefits of open standards. Another major change in the 2.6 release, is that the kernel's internal threading infrastructure has been rewritten to allow the Native POSIX Thread Library (NPTL) to run on top of it. This can be a major performance boost for Pentium Pro and better processors in heavily threaded applications, and many of the top players in the "enterprise" space have been clamoring for it. (In fact, Red Hat has already backported the support to Linux 2.4 and includes it starting with Red Hat 9 and Advanced Server 3.0) This change includes new concepts to the Linux thread space including thread groups, local memory for individual threads, POSIX-style signals, and other changes. One of the major drawbacks is that applications (such as some versions of Sun Java) not written to spec that rely on old Linux-isms will break with the new support enabled. As the benefits overwhelm the cost (and with so many large players in the game), it's clear that most important applications will support the changes before too long after the new kernel is released.

Module Subsystem and the Unified Device Model

Increasingly in modern operating systems, the device handling subsystems have taken on new prominence as they are forced to deal with a myriad of internal and external bus types and more devices by more vendors than you can shake a stick at. It should come as no surprise then, that the upcoming upgrade to the Linux kernel will include improved support both in its module loader, but also in its internal understanding of the hardware itself. These changes range from the purely cosmetic (driver modules now use a ".ko" extension, for "kernel object", instead of just ".o") to a complete overhaul of the unified device model. Throughout all of these changes is an emphasis on stability and better grasp of the limitations of the previous revision.

Strictly in the module (driver) subsystem, there are a handful of major changes to improve stability. The process for unloading modules have been changed somewhat to reduce cases where it is possible for modules to be used while they are still being unloaded, often causing a crash. For systems where this problem cannot be risked, it is now even possible to disable unloading of modules altogether. Additionally, there has been extensive effort to standardize the process by which modules determine and announce what hardware they support. Under previous versions of Linux, the module would "know" what devices it supported, but this information was not generally available outside of the module itself. This change will allow hardware management software, such as Red Hat's "kudzu", to make intelligent choices even on hardware that would not otherwise recognize. Of course, in the event that you know better than the current version of the driver what it supports, it is still possible to force a driver to try to work on a specific device.

Outside of just module loading, the device model itself has undergone significant changes in the updated kernel release. Unlike the module loader, which just has to concern itself with detecting the resource requirements of incoming hardware, the device model is a deeper concept which must be completely responsible for all of the hardware in the system. Linux versions 2.2 and earlier had only the barest support for a unified device model, preferring instead to leave almost all knowledge of the hardware solely at the module level. This worked fine, but in order to use all of the features of modern hardware (especially ACPI), a system needs to know more than just what resources a device uses: it needs to know things like what bus it is connected to, what subdevices it has, what its power state is, whether it can be reconfigured to use other resources in the event of contention, and even to know about devices that haven't had modules loaded for them yet. Linux 2.4 expanded on this foundation to become the first edition to unify the interfaces for PCI, PC Card, and ISA Plug-and-Play buses into a single device structure with a common interface. Linux 2.6, through its new kernel object ("kobject") subsystem, takes this support to a new level by not only expanding to know about all devices in a system, but also to provide a centralized interface for the important little details like reference counting, power management, and exports to user-space.

Now that an extensive amount of hardware information is available within the kernel, this has allowed Linux to better support modern laptop and desktop features that require a much more in-depth knowledge of hardware. The most readily apparent application of this is the increasing proliferation of so called "hot plug" devices like PC Cards, USB and Firewire devices, and hot-plug PCI. While it's hard to think back that far now, Linux didn't offer true support for any of these devices until the 2.2 kernel. Given that hot-plugging is the rule these days and not the exception, it is fitting that the new device infrastructure essentially eliminates the differences between a hot-plug and a legacy device. Since the kernel subsystem does not directly differentiate between a device discovered at boot time from one discovered later, much of the infrastructure for dealing with pluggable devices has been simplified. A second up and coming driver of this newly rewritten subsystem is for improved support of modern power management. The new power management standard in recent years, called ACPI for "Advanced Configuration and Power Interface", was first supported in rough form for the previous version of the kernel. Unlike old-fashioned APM ("Advanced Power Management"), OSes run on systems with this new interface are required to individually tell all compatible devices that they need to change their power states. Without a centralized understanding of hardware, it would be impossible for the kernel to know what devices it needs to coordinate with and in what order. Although these are just two obvious examples, there are clearly other areas (such as hardware auditing and monitoring) that will benefit from a centralized picture of the world.

The final, but possibly the most obvious, ramification of the new centralized infrastructure is the creation of a new "system" filesystem (to join 'proc' for processes, 'devfs' for devices, and 'devpts' for UNIX98 pseudo-terminals) called 'sysfs'. This filesystem (intended to be mounted on '/sys') is a visible representation of the device tree as the kernel sees it (with some exceptions). This representation generally includes a number of known attributes of the detected devices, including the name of the device, its IRQ and DMA resources, power status, and that sort of thing. However, one aspect of this change that may be confusing on the short term is that many of the device-specific uses of the "/proc/sys" directory may be moved into this new filesystem. This change has not (yet) been applied consistently, so there may continue to be an adjustment period.

System Hardware Support

As Linux has moved forward over the years and into the mainstream, each new iteration of the kernel appeared to be leaps and bounds better than the previous in terms of what types of devices it could support-- both in terms of emerging technologies (USB in 2.4) and older "legacy" technologies (MCA in 2.2). As we arrive at the 2.6 however, the number of major devices that Linux does not support is relatively small. There are few, if any, major branches of the PC hardware universe yet to conquer. It is for that reason that most (but certainly not all) of improvements in i386 hardware support have been to add robustness rather than new features.

Internal Devices

Arguably as important as the processor type, the underling bus(es) in a system are the glue that holds things together. The PC world has been blessed with no shortness of these bus technologies, from the oldest ISA (found in the original IBM PC) to modern external serial and wireless buses. Linux has always been quick to adapt to a new bus and device type as they have become popular with consumer devices, but significantly less quick adapting to technologies that get relatively little use.

Improvements in Linux's support for internal devices are really spread across the board. One specific example where Linux is playing "catch up" is support for the old ISA ("Industry Standard Architecture") Plug-and-Play extensions. Linux didn't offer any built-in support for PnP at all until the 2.4 release. This support has been rounded out with the upcoming kernel to include full PnP BIOS support, a device name database, and other compatibility changes. The sum of all of those modifications, is that now Linux is now a "true" Plug-and-Play OS and may be set as such in a compatible machine's BIOS. Other legacy buses such as MCA ("Microchannel Architecture") and EISA ("Extended ISA") have also been wrapped into the new device model and feature device naming databases. On a more modern front Linux 2.6 brings to the table a number of incremental improvements to its PCI ("Peripheral Component Interconnect") subsystem including improved Hot-Plug PCI and power management, support for multiple AGPs ("accelerated graphics ports" -- a separate high-speed extension to the PCI bus), and other changes. And finally, in addition to all of the "real" buses, Linux 2.6 has internally added the concept of a "legacy" bus that is specific to each architecture and contains all of the assumed devices that you would expect to find. On a PC, for example, this may include on-board serial, parallel, and PS/2 ports-- devices that exist but are not enumerated by any real buses on the system. This support may require more complicated work (such as querying firmware) on some platforms, but in general this is just a wrapper to ensure that all devices are handled in a standard way in the new driver paradigm.

External Devices

While it is true that the older-style internal device buses have not seen many new features during the most recent development cycle, the same cannot be said for hot new external hardware. Possibly the most important change in this space is the new support for USB 2.0 devices. These devices, commonly referred to as "high speed" USB devices, support device bandwidth of up to 480 megabits per second, compared to 12 Mbit/sec of current USB. A related new standard, USB On-the-Go (or USB OTG), a point-to-point variant on the USB protocol for connecting devices directly together (for example, to connect a digital camera to a printer without having a PC in the middle) is not currently supported in Linux 2.6. (Patches for this feature are available, but not yet rolled into the official release.) In addition to device support, much of the way USB devices have been internally enumerated has been revised so that it is now possible to have many more devices of the same type all accessible from within Linux. In addition to the large changes, there has been an emphasis placed in this development cycle on simplification, stability, and compatibility that should improve the support of USB devices for all Linux users.

On the complete opposite end of the field, Linux 2.6 for the first time includes support that allows a Linux-powered machine to be a USB device, rather than a USB host. This would allow, for example, your Linux-powered PDA to be plugged into your PC and to have both ends of the line speaking the proper protocol. Much of this support is new, but this is an essential direction for Linux to move into for embedded devices.

Wireless Devices

Wireless technology has really taken off within the public in the past several years. It often seems as if cords (except power... maybe?) will be a thing of the past within a handful of years. Wireless devices encompass both networking devices (the most common currently) and also more generic devices such as PDAs, etc.

In the wireless networking space, devices can generally be divided into long range (for example, AX.25 over amateur radio devices) and short range (usually 802.11, but some older protocols exist.) Support for both of these has been a hallmark of Linux since the early days (v1.2) and both of these subsystems have been updated during development of 2.6. The largest change here is that major components of the short range subsystems for the various supported cards and protocols has been merged into a single "wireless" subsystem and API. This merge resolves a number of minor incompatibilities in the way different devices have been handled and strengthens Linux's support for the subsystem by making a central set of userspace tools that will work with all supported devices. In addition to just standardization, Linux 2.6 introduces a number of overall improvements including better capability to notify in the event of a state change (such as a device that has a "roaming" state) and a change to TCP to better handle periodic delay spikes which occur with wireless devices. Due to the immediate desire to better support wireless devices in the current Linux 2.4 kernel, many of these changes have already been back-ported and are available for use.

In the generic wireless devices space, there have been similar major advancements. IrDA (the infrared protocol named for the Infrared Data Associates group) has received some advancements since the last major release such as power management and integration into the new kernel driver model. The real advancements however have been made in providing Linux support for Bluetooth devices. Bluetooth is a new wireless protocol that is designed to be short range and low on power consumption, but does not have the line of sight limitations that IrDA has. Bluetooth as a protocol is designed to go "anywhere" and has been implemented in devices like PDAs, cell phones, printers, and more bizarre things such as automotive equipment. The protocol itself is made up of two different data link types: SCO, or "Synchronous Connection Oriented", for lossy audio applications; and L2CAP, or "Logical Link Control and Adaptation Protocol", for a more robust connection supporting retransmits, etc. The L2CAP protocol further supports various sub-protocols (including RFCOMM for point-to-point networking and BNEP for Ethernet-like networking.) Linux's support for the things that Bluetooth can do continues to grow and we can expect this to mature significantly once more devices are in the hands of the consumers. It should also be mentioned that initial support for Bluetooth has been integrated into later editions of the 2.4 kernel.

Block Device Support

Storage buses

Dedicated storage buses, such as IDE/ATA ("Integrated Drive Electronics/Advanced Technology Attachment") and SCSI ("Small Computer System Interface"), have also received a major update during the 2.6 cycle. The most major changes are centered around the IDE subsystem which has been rewritten (and rewritten again) during the development of the new kernel, resolving many scalability problems and other limitations. For example, IDE CD/RW drives can now be written to directly through the real IDE disk driver, a much cleaner implementation than before. (Previously, it was required to also use a special SCSI-emulating driver which was confusing and often difficult.) In addition, new support has been added for high-speed Serial ATA (S-ATA) devices, which have transfer rates exceeding 150 MB/sec. On the SCSI side, there have also been many small improvements scattered around the system both for wider support and scalability. One specific improvement for older systems is that Linux now supports SCSI-2 multipath devices that have more than 2 LUNs on a device. (SCSI-2 is the previous version of the SCSI device standard, circa 1994.) Another important change is that Linux can now fall back to test media changing like Microsoft Windows does, to be more compatible with devices that do not completely follow the specification. As these technologies have stabilized over time, so too has Linux's support for them.

Although not a storage bus in itself, Linux now includes support for accessing a newer machine's EDD ("Enhanced Disk Device") BIOS directly to see how the server views its own disk devices. The EDD BIOS includes information on all of the storage buses which are attached to the system that the BIOS knows about (including both IDE and SCSI.) In addition to just getting configuration and other information out of the attached devices, this provides several other advantages. For example, this new interface allows Linux to know what disk device the system was booted from, which is useful on newer systems where it is often not obvious. This allows intelligent installation programs to consider that information when trying to determine where to put the Linux boot loader, for example.

In addition to all of these changes, it should be stressed again that all of the bus device types (hardware, wireless, and storage) have been integrated into Linux's new device model subsystem. In some cases, these changes are purely cosmetic. In other cases, there are more significant changes involved (in some cases for example, even logic for how devices are detected needed to be modified.)


The most obvious use of a block device on a Linux (or any other) system is by creating a filesystem on it, and Linux's support for filesystems have been vastly improved since Linux 2.4 in a number of respects. Key among these changes include support for extended attributes and POSIX-style access controls.

When dealing strictly with conventional Linux filesystems, the extended filesystems (either "ext2" or "ext3") are the systems most associated with a core Linux system. (ReiserFS is the third most common option.) As these are the filesystems that users care about the most, they have also been the most improved during the development of Linux 2.6. Principal among these changes is support for "extended attributes", or metadata that can be embedded inside the filesystem itself for a specific file. Some of these extended attributes will be used by the system and readable and writable by root only. Many other operating systems, such as Windows and the MacOS, already make heavy use of these kinds of attributes. Unfortunately, the UNIX legacy of operating systems have not generally included good support for these attributes and many user-space utilities (such as 'tar') will need to be updated before they will save and restore this additional information. The first real use of the new extended attribute subsystem is to implement POSIX access control lists, a superset of standard UNIX permissions that allows for more fine-grained control. In addition to these changes for ext3, there are several other smaller changes: the journal commit time for the filesystem can now be tuned to be more suited for laptop users (which might have to spin up the drive if it were in a power save mode.), default mount options can now also be stored within the filesystem itself (so that you don't need to pass them at mount time), and you can now mark a directory as "indexed" to speed up searches of files in the directory.

In addition to the classic Linux filesystems, the new kernel offers full support for the new (on Linux) XFS filesystem. This filesystem is derived from and is block-level compatible with the XFS filesystem used by default on Irix systems. Like the extended filesystems and Reiser, it can be used as a root-disk filesystem and even supports the newer features such as extended attributes and ACLs. Many distributions are beginning to offer support for this filesystem on their Linux 2.4-based distributions, but it remains to be seen yet what place this filesystem will have in the already crowded pantheon of UNIX-style filesystems under Linux.

Outside of those, Linux has also made a number of improvements both inside and outside the filesystem layer to improve compatibility with the dominant PC operating systems. To begin with, Linux 2.6 now supports Windows' Logical Disk Manager (aka "Dynamic Disks"). This is the new partition table scheme that Windows 2000 and later have adopted to allow for easier resizing and creation of multiple partitions. (Of course, it is not likely that Linux systems will be using this scheme for new installations anytime soon.) Linux 2.6 also features improved (and rewritten) support for the NTFS filesystem and it is now possible to mount a NTFS volume read/write. (Writing support is still experimental and is gradually being improved; it may or may not be enabled for the final kernel release.) And finally, Linux's support for FAT12 (the DOS filesystem used on really old systems and floppy disks) has been improved to work around bugs present in some MP3 players which use that format. Although not as dominant in the marketplace, Linux has also improved compatibility with OS/2 by adding extended attribute support into the HPFS filesystem. Like previous releases, the new additions to Linux 2.6 demonstrate the importance of playing well with others and reinforces Linux's position as a "Swiss Army Knife" operating system.

In addition to these changes, there have been a large number of more scattered changes in Linux's filesystem support. Quota support has been rewritten to allow for the larger number of users supported on a system. Individual directories can now be marked as synchronous so that all changes (additional files, etc.) will be atomic. (This is most useful for mail systems and directory-based databases, in addition to slightly better recovery in the event of a disk failure.) Transparent compression (a Linux-only extension) has been added to the ISO9660 filesystem (the filesystem used on CD-ROMs.) And finally, a new memory-based filesystem ("hugetlbfs") has been created exclusively to better support shared memory databases.

Input / Output Support

On the more "external" side of any computer system is the input and output devices, the bits that never quite seem as important as they are. These include the obvious things like mice and keyboards, sound and video cards, and less obvious things like joysticks and accessibility devices. Many of Linux's end-user subsystems have been expanded during the 2.6 development cycle, but support for most of the common devices were already pretty mature. Largely, Linux 2.6's improved support for these devices are derived directly from the more general improvments with external bus support, such as the ability to use Bluetooth wireless keyboards and similar. There are however a number of areas where Linux has made larger improvements.

Human Interface Devices

One major internal change in Linux 2.6 is the reworking of much of the human interface layer. The human interface layer is the center of the user experience of a Linux system, including the video output, mice, and keyboards. In the new version of the kernel, this layer has been reworked and modularized to a much greater extent than ever before. It is now possible to create a completely "headless" Linux system without any included support for a display or anything. The primary benefit of this modularity may be for embedded developers making devices that can only be administrated over the network or serial, but end-users benefit as many of the underlying assumptions about devices and architectures has been modularized out. For example, it was previously always assumed that if you had a PC that you would need support for a standard AT (i8042) keyboard controller; the new version of Linux removes this requirement so that unnecessary code can be kept out of legacy-less systems.

Support of Linux's handling of monitor output has also received a number of changes, although most of these are useful only in configurations that make use of the kernel's internal framebuffer console subsystem. (Most Intel Linux boxes are not configured this way, but that is not the case for many other architectures.) In my personal opinion, the best feature is that the boot logo (a cute penguin, if you've never seen it) now supports resolutions up to 24bpp. That aside, other new features for the console include resizing and rotating (for PDAs and similar) and expanded acceleration support for more hardware. And finally, Linux has now included kernel support for querying VESA ("Video Electronics Standard Association") monitors for capability information, although XFree86 and most distributions installation systems already have covered this detail in user-space.

In addition to the big changes, Linux 2.6 also includes a number of smaller changes for human interaction. Touch screens, for example, are now supported. The mouse and keyboard drivers have also been updated and standardized to only export a single device node ("/dev/input/mouse0", for example) regardless of the underlying hardware or protocol. Bizarre mice (with multiple scroll wheels, for example) are now also supported. PC keyboard key mappings have also been updated to follow the Windows "standard" for extended keys. Joystick support has also been improved thanks not only to the addition of many new drivers (including the X Box gamepad), but also to include newer features such as force-feedback. And finally (but not least important), the new release also includes support for the Tieman Voyager braille TTY device to allow blind users better access to Linux. (This feature is important enough that it has been back-ported to Linux 2.4 already.)

As a side note, Linux has also changed the "system request" interface to better support systems without a local keyboard. The system request ("sysrq") interface is a method for systems administrators at the local console to get debugging information, force a system reboot, remount filesystems read-only, and do other wizardly things. Since Linux 2.6 now supports a completely headless system, it is now also possible to trigger these events using the /proc filesystem. (Of course, if your system hangs and you need to force it to do things, this may not be of much help to you.)

Audio & Multimedia

One of the most anticipated new features of Linux 2.6 for desktop users is the inclusion of ALSA (the "Advanced Linux Sound Architecture") in lieu of the older sound system. The older system, known as OSS for "Open Sound System", has served Linux since the early days but had many architectural limitations. The first major improvement with the new system is that it has been designed from the start to be completely thread and SMP-safe, fixing problems with many of the old drivers where they would not work properly outside the expected "desktop-means-single-cpu paradigm." More importantly, the drivers have been designed to be modular from the start (users of older versions of Linux will remember that modularity was retro-fitted onto the sound system around Linux 2.2), and that this allows for improved support for systems with multiple sound cards, including multiple types of sound cards. Regardless of how pretty the internals are, the system would not be an improvement for users if it did not have neat new whiz-bang features, and the new sound system has many of those. Key among them are support for newer hardware (including USB audio and MIDI devices), full-duplex playback and recording, hardware and non-interleaved mixing, support for "merging" sound devices, and other things. Whether you are an audiophile or just someone that likes to play MP3s, Linux's improved sound support should be a welcome step forward.

Beyond simple audio these days, what users want is support for the really fancy hardware like webcams, radio and TV adapters, and digital video recorders. In all three cases, Linux's support has been improved with the 2.6 release. While Linux has supported (to a greater or lesser extent) radio cards (often through userspace) for many iterations, support for television tuners and video cameras was only added within the last one or two major revisions. That subsystem, known as Video4Linux (V4L), has received a major upgrade during the work on the new edition of the kernel including both an API cleanup and support for more functionality on the cards. The new API is not compatible with the previous one and applications supporting it will need to upgrade with the kernel. And on a completely new track, Linux 2.6 includes the first built-in support for Digital Video Broadcasting (DVB) hardware. This type of hardware, common in set-top boxes, can be used to make a Linux server into a Tivo-like device, with the appropriate software.

Software Improvements


Leading-edge networking infrastructure has always been one of Linux's prime assets. Linux as an OS already supports most of the world's dominant network protocols including TCP/IP (v4 and v6), AppleTalk, IPX, and others. (The only unsupported one that comes to mind is IBM/Microsoft's obsolete and tangled NetBEUI protocol.) Like many of the changes in the other subsystems, most networking hardware changes with Linux 2.6 are under the hood and not immediately obvious. This includes low-level changes to take advantage of the device model and updates to many of the device drivers. For example, Linux now includes a separate MII (Media Independent Interface, or IEEE 802.3u) subsystem which is used by a number of the network device drivers. This new subsystem replaces many instances where each driver was handling that device's MII support in slightly different ways and with duplicated code and effort. Other changes include major ISDN updates and other things.

On the software side, one of the most major changes is Linux's new support for the IPsec protocols. IPsec, or IP Security, is a collection of protocols for IPv4 ("normal" IP) and IPv6 that allow for cryptographic security at the network protocol level. And since the security is at the protocol level, applications do not have to be explicitly aware of it. This is similar to SSL and other tunneling/security protocols, but at a much lower level. Currently supported in-kernel encryption includes various flavors of SHA ("secure hash algorithm"), DES ("data encryption standard"), and others.

Elsewhere on the protocol side, Linux has improved its support for multicast networking. Multicast networks are networks where a single sent packet is intended to be received by multiple computers. (Compare to traditional point-to-point networks where you are only speaking to one at a time.) Primarily, this functionality is used by messaging systems (such as Tibco) and audio/video conferencing software. Linux 2.6 improves on this by now supporting several new SSM (Source Specific Multicast) protocols, including MLDv2 (Multicast Listener Discovery) and IGMPv3 (Internet Group Messaging Protocol.) These are standard protocols that are supported by most high-end networking hardware vendors, such as Cisco.

Linux 2.6 also has broken out a separate LLC stack. LLC, or Logical Link Control protocol (IEEE 802.2), is a low-level protocol that is used beneath several common higher-level network protocols such as Microsoft's NetBeui, IPX, and AppleTalk. As part of this change-over, the IPX, AppleTalk, and Token Ring drivers have been rewritten to take advantage of the new common subsystem. In addition, an outside source has put together a working NetBEUI stack and it remains to be seen whether it will ever be integrated into the stock kernel.

In addition to these changes, there have been a number of smaller changes. IPv6 has received some major changes and it can now also run on Token Ring networks. Linux's NAT/masquerading support has been extended to better handle protocols that require multiple connections (H.323, PPTP, etc.) On the Linux-as-a-router front, support for configuring VLANs on Linux has been made no longer "experimental".

Network Filesystems

Overlaid on top of Linux's robust support for network protocols is Linux's equally robust support for network filesystems. Mounting (and sometimes exporting) a network filesystem is one of the very few high-level network operations that the kernel cares about directly. (The most obvious other, the "network block device", did not receive many changes for 2.6 and is generally used in specialized applications where you end up doing something filesystem-like with it anyway.) All other network operations are content to be relegated to user-space and outside the domain of the kernel developers.

In the Linux and UNIX-clone world, the most common of the network filesystems is the aptly named Network File System, or NFS. NFS is a complicated file sharing protocol that has deep roots in UNIX (and especially Sun Solaris' excellent implementation). The primary transport protocol can utilize either TCP or UDP, but several additional sub-protocols are also required, each of which also run on top of the separate RPC ("remote procedure call") protocol. These include the separate "mount" protocol for authentication and NLM ("network lock manager") for file locking. (The common implementation is also tied closely to other common RPC-based protocols, including NIS-- "network information service"-- for authentication. NIS and its progeny are not commonly used for authentication on Linux machines due to fundamental insecurities.) It is perhaps because of this complexity that NFS has not been widely adapted as an "Internet" protocol.

In Linux 2.6, this core Linux filesystem received many updated and improvements. The largest of these improvements is that Linux now experimentally supports the new and not widely adopted NFSv4 protocol version for both its client and server implementations. (Previous versions of Linux included support for only the v2 and v3 versions of the protocol.) The new version supports stronger and more secure authentication (with cryptography), more intelligent locking, support for pseudo-filesystems, and other changes. Not all of the new NFSv4 features have been implemented in Linux yet, but the support is relatively stable and could be used for some production applications. In addition, Linux's NFS server implementation has been improved to be more scalable (up to 64 times as many concurrent users and a larger request queues), to be more complete (by supporting serving over TCP, in addition to UDP), to be more robust (individual filesystems drivers can adapt the way files on those systems are exported to suit their particularities), and more easily maintainable (management though a new 'nfsd' filesystem, instead of system calls.) There have also been may other under the hood changes, including separating lockd and nfsd, and support for zero-copy networking on supported interfaces. NFS has also been made somewhat easier to secure by allowing the kernel lockd port numbers to be assigned by the administrator. The NFS client side has also benefited from a number of improvements to the implementation of the underlying RPC protocol including a caching infrastructure, connection control over UDP, and other improvements for TCP. Linux's support for using NFS-shared volumes as the root filesystem (for disk-less systems) has also been improved as the kernel now supports NFS over TCP for that purpose.

In addition to improving support for the UNIX-style network filesystems, Linux 2.6 also delivers many improvements to Windows-style network filesystems. The standard shared filesystem for Windows servers (as well as OS/2 and other operating systems) has been the SMB ("server message block") protocol and the Linux kernel has had excellent client support of the SMB protocol for many revisions. Windows 2000 however standardized on an upgraded superset of the SMB protocol, known as CIFS ("common internet filesystem.") The intention of this major update was to streamline and refine certain aspects of SMB which had at that point become a complete mess. (The protocol itself was loosely defined and often extended to the point that there were cases even where the Win95/98/ME version was incompatible with the WinNT/Win2k version.) CIFS delivered on that intention and added UNICODE support, improved file locking, hard linking, eliminated the last vestiges of NetBIOS dependencies, and added a few other features for Windows users. Since Linux users do not like to be kept in the dark for long, Linux 2.6 now includes completely rewritten support for mounting CIFS filesystems natively. Linux 2.6 also now includes support for the SMB-UNIX extensions to the SMB and CIFS protocols which allows Linux to access non-Windows file types (such as device nodes and symbolic links) on SMB servers which support it (such as Samba.) Although not as commonly seen today, Linux has not completely forgotten about the Novell NetWare users. Linux 2.6 now allows Linux clients to mount up to the maximum of 256 shares on a single NetWare volume using its built in NCP ("NetWare Core Protocol") filesystem driver.

Linux 2.6 also includes improved support for the relatively new domain of distributed network filesystems, systems where files on a single logical volume can be scattered across multiple nodes. In addition to the CODA filesystem introduced in Linux 2.4, Linux now includes some support for two other distributed filesystems: AFS and InterMezzo. AFS, the Andrew filesystem (so named because it was originally developed at CMU), is presently very limited and restricted to read-only operations. (A more feature complete version of AFS is available outside the kernel-proper.) The second newly supported filesystem, InterMezzo (also developed at CMU), is also newly supported under Linux 2.6 and it allows for more advanced features such as disconnect operation (so you work on locally cached files) and is suitable for high-availability applications where you need to guarantee that storage is never unavailable (or faked, when down). It also has applications for keeping data in sync between multiple computers, such as a laptop or PDA and a desktop computer. Many of the projects providing support for these new types of filesystems are initially developed on Linux, putting Linux well ahead of the curve in support for these new features.

Miscellaneous Features


Another of the big changes in Linux 2.6 that does not receive enough attention is the wealth of new security-related changes. Most fundamentally, the entirety of kernel-based security (powers of the super user under a UNIX-like operating system) has been modularized out to be one out of a potential number of alternate security modules. (At present, however, the only offered security model is the default one and an example how to make your own.) As part of this change, all parts of the kernel have now been updated to use "capabilities" as the basis of fine-grained user access, rather than the old "superuser" system. Nearly all Linux systems will continue to have a "root" account which has complete access, but this allows for a Linux-like system to be created which does not have this underlying assumption. Another security-related change is that binary modules (for example, drivers shipped by a hardware manufacturer) can no longer "overload" system calls with their own and can no longer see and modify the system call table. This significantly restricts the amount of access that non-open source modules can do in the kernel and possibly closes some legal loopholes around the GPL. The final change that is somewhat security-related is that Linux with the new kernel is now able to use hardware random number generators (such as those present in some new processors), rather than relying on a (admittedly quite good) entropy pool based on random hardware fluctuations.

Virtualizing Linux

One of the most interesting new features in Linux 2.6 is its inclusion of a "user-mode" architecture. This is essentially a port (like to a different hardware family) of Linux to itself, allowing for a completely virtualized Linux-on-Linux environment to be run. The new instance of Linux runs as if it was a normal application. "Inside" the application, you can configure fake network interfaces, filesystems, and other devices through special drivers which communicate up to the host copy of Linux in a secure way. This has proved quite useful, both for development purposes (profiling, etc.) as well as for security analysis and honeypots. While most users will never need this kind of support, it is an incredibly "cool" feature to have running on your box. (Impress your friends!)


In addition to all of the other general purpose support described above (improved APM and ACPI, wireless support improvements, etc.) Linux also includes two other hard-to-classify features that will best assist laptop users. The first is that the new edition of the kernel now supports software-suspend-to-disk functionality for the Linux user on the go. This feature still has some bugs to iron out, but is looking solid for many configurations. The new version also supports the ability of modern mobile processors to change speed (and, in effect, power requirements) based on whether your system is plugged in or not.

Configuration Management

Linux 2.6 includes another feature which might seem minor to some, but will both greatly assist developers' abilities to debug kernel problems of end-users as well as make it easier for individual administators to know configuration details about multiple systems. In short, the kernel now supports adding full configuration information into the kernel file itself. This information would include details such as what configuration options were selected, what compiler was used, and other details which would help someone reproduce a similar kernel if the need arose. This information would also be exposed to users via the /proc interface.

Legacy Support

Although Linux 2.6 is a major upgrade, the difference to user-mode applications will be nearly non-existent. The one major exception to this rule appears to be threading: some applications may do things that worked under 2.4 or 2.2 but are no longer allowed. Those applications should be the exception to the rule however. Of course, low-level applications such as module utilities will definitely not work. Additionally, some of the files and formats in the /proc and /dev directories have changed and any applications that have dependencies on this may not function correctly. (This is especially true as more things shift over to the new "/sys" virtual filesystem. In the "/dev" case, backwards-compatible device names can easily be configured.)

In addition to those standard disclaimers, there are a number of other smaller changes which may affect some environments. First, very old swap files (from Linux 2.0 or earlier) will need to be reformatted before they can be used with 2.6. (Since swap files do not contain any permanent data, that should not be a problem for any user.) The kHTTPd daemon which allowed the kernel to serve web pages directly has also been removed as most of the performance bottlenecks that prevented Apache, Zeus, et. al. from reaching kernel speeds have been resolved. Autodetection of DOS/Windows "disk managers" such as OnTrack and EzDrive for large harddisk support with older BIOSes has been removed. And finally, support for using a special kernel-included boot sector for booting off of a floppy disk has also been removed; you now need to use SysLinux instead.

Stuff At The Bottom

This document was assembled primarily from long hours looking at BitKeeper changelogs, looking at and playing with the source, reading mailing list posts, and lots and lots of Google and Lycos searches for documentation about this and that. As such, there are likely places where something could have been missed or misunderstood, and places where something could have been backed out after the fact. (I have been especially careful of the two versions of IDE support that were worked on during this time period, but there are other examples.) As a bit of my research was done by looking at the web pages of various kernel projects, I have had to be careful that the independent projects weren't farther ahead with features than were accepted into the mainline Linux code. If you see any errors in this document or want to email me to ask me how my day is going, you can do so at jpranevich <at> kniggit.net.

The newest version of this document can always be found at http://kniggit.net/wwol26.html.


Not an English speaker? This document (or an older revision) has been translated into a handful of other languages.

Bulgarian - http://kniggit.net/wwol26bg.html (Ivan Dimov)
Chinese - http://www-900.ibm.com/developerWorks/cn/linux/kernel/l-kernel26/index.shtml (Stone Wang, et. al.)
Czech - http://www.linuxzone.cz/index.phtml?ids=10&idc=782 (David Haring)
French - http://dsoulayrol.free.fr/articles/wonderful_2.6.html (David Soulayrol)
Hungarian - http://free.srv.hu/b/e/behun/pn/modules.php?op=modload&name=News&file=index&catid=&topic=12 (Ervin Novak) (Not yet completed.)
Italian - http://www.opensp.org/tutorial/vedi.php?appartenenza=42&pagine=1 (Giulio Ciuffi Vampa)
Portuguese (BR) - http://geocities.yahoo.com.br/cesarakg/wwol26-ptBR.html (Cesar A. K. Grossmann)
Russian - http://www.opennet.ru/base/sys/linux26_intro.txt.html (Sergey Prokopenko)
Spanish - http://www.escomposlinux.org/wwol26/wwol26.html (Alex Fernández)

An abridged version also appeared in German in the 09/2003 issue of LanLine magazine. I believe that an unabridged edition may be floating around also, but I am uncertain of the link. If you know of any additional translations to add to this list, please let me know.

Author's note: please send me a copy of any off-line reprints of this article.

[BIO] I'm just this guy, y'know? More info about me is on my web page.

Copyright © 2004, Joe Pranevich. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004


By Javier Malonda

The Ecol comic strip is written for escomposlinux.org (ECOL), the web site that supports es.comp.os.linux, the Spanish USENET newsgroup for Linux. The strips are drawn in Spanish and then translated to English by the author.

These images are scaled down to minimize horizontal scrolling. To see a panel in all its clarity, click on it.


All Ecol cartoons are at tira.escomposlinux.org (Spanish), comic.escomposlinux.org (English) and http://tira.puntbarra.com/ (Catalan). The Catalan version is translated by the people who run the site; only a few episodes are currently available.

These cartoons are copyright Javier Malonda. They may be copied, linked or distributed by any means. However, you may not distribute modifications. If you link to a cartoon, please notify Javier, who would appreciate hearing from you.

Copyright © 2004, Javier Malonda. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004


By Jon "Sir Flakey" Harsem

These images are scaled down to minimize horizontal scrolling. To see a panel in all its clarity, click on it.


Bonus! A HelpDex cartoon by Shane Collinge.


All Qubism cartoons are here at the CORE web site.

[BIO] Jon is the creator of the Qubism cartoon strip and current Editor-in-Chief of the CORE News Site. Somewhere along the early stages of his life he picked up a pencil and started drawing on the wallpaper. Now his cartoons appear 5 days a week on-line, go figure. He confesses to owning a Mac but swears it is for "personal use".

Copyright © 2004, Jon "Sir Flakey" Harsem. Released under the Open Publication license unless otherwise noted in the body of the article. Linux Gazette is not produced, sponsored, or endorsed by its prior host, SSC, Inc.

Published in Issue 98 of Linux Gazette, January 2004