...making Linux just a little more fun!

next -->

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.

AGP problem

Fri, 30 Sep 2005 07:37:55 -0700 (PDT)
Igor T. (tymarats from yahoo.com)

Hello, Gang!

I appologize in advance because I know this is going to be a long letter. The thing is, I've had this problem for more than a year now and I hope that you can help me solve it. Plus, I just have to tell somebody about it :)

The problem showed itself for the first time when I installed NVIDIA GLX drivers on my box. At the moment, I had a Riva TNT2 M64, and I wanted to see how far can I get with those drivers. Although most people hated NVIDIA drivers, I had no problem with them. Or, at least I thought so. Quake3 was fine, blender too, but as the time passed, I noticed that my box wasn't stable any more. Sometimes it happened that, out of no obvious reason, X hangs. I can move my mouse pointer, but computer doesn't respond to anything. Curiously enough, if I had xmms started, the songs would continue to play on normally, and everything seemed to work fine - except the fact that X is hung and I couldn't switch to console, either. Later, I found one way to reproduce the bug quickly, rather than to wait untill it happens (it was simply playing the "Virtual Machine" screensaver but, as you will shortly see, it seems that this wasn't of any importance) and I tried to log in from my laptop via telnet when the hang occurs. I succeeded, and everything seemed to work on the box side, I couldn't find ANY reason why X was hung, nor any indication that it actually happened. Furthermore, when I killed X (from telnet), box would have returned to console happily, and I was able to start X again! As soon as I returned to the original nv driver, system was steady as rock again. So, buggy NVIDIA driver/X version/kernel, I figured. But, as the time passed, I've changed a lot of those and so many combinations (XFree86, Xorg, kernel 2.4.x, 2.6.x, almost every NVIDIA driver...) that I finally had to admit that something was seriously wrong. Many nights without sleep and I finally found the way to workaround the problem -- setting "NvAGP" option to "0". As it shows, using either internal NVIDIA AGP support, or kernel agpgart hangs the X in a mentioned manner. I didn't know why, but now everything was accelerated, and system was stable again, so I forgot about the incident.

Since then, I've "upgraded" Riva to Radeon 9200. For a long time I was using only ati driver (FireGL drivers didn't support Xorg, and I didn't want to revert to XFree out of principle), but a month ago I had everything ready to actually try FireGL drivers. After some problems, I finally got it to work, and I was happy. But, after few hours -- same thing as with Riva. Hang of death. The old trick of reproducing the error doesn't work now, but I found another one -- dragging the mouse pointer rapidly while holding left mouse button (over empty KDE desktop). All other symptoms were exactly the same as with Riva (music plays, everything works...) except that now, when I kill the X from laptop, whole system hangs and nothing wors anymore (including networking -- telnet session terminates). So, it's almost certain that this is a hardware problem, but none of this happens under XP. So, I try to tweak various fgl parameters (disabling/enabling user page locking, internal/external AGPGART, ...), but the problem persists.

Is this a hardware problem (XP works fine, after all)? Do I have something misconfigured (I tried various AGP aperture sizes in BIOS, but it doesn't solve the problem)? Is there any way that I can "disable" usage of AGP as I did with NVIDIA driver? I don't have enough money to buy a new motherboard, so please don't tell me that it is the only solution :)

And, finally (finally!) techical stuff: 

Box: 	AMD Duron 750
	384MB RAM
	VIA CT-7AIA
	Radeon 9200 64MB DDR

(at the moment:)
	Slackware 10.2
	kernel 2.6.11-9

$ /sbin/lspci | grep -i agp
00:01.0 PCI bridge: VIA Technologies, Inc. VT8363/8365
[KT133/KM133 AGP]

I have via-agp module loaded before loading agpgart, because fglrx module doesn't work without it (fglrxinfo reports usage of Mesa, not ATI drivers, and 3D acceleration doesn't work). I appreciate your help!

igor

FIASCO (Fractal Image and Sequencing Codec)

Wed, 5 Oct 2005 08:58:04 -0700
Jim Harvey (harveyj from entrenet.com
Answer by Rick Moen (LG Contributing Editor)

I am trying to get a copy of the Fractal Image and Sequencing Codec (FIASCO) that was announced in the Sept 2000 issue of your journal.

You've actually written to the mailing list of the Answer Gang, which is a group of volunteers who collectively write column in the magazine that researches and answers people's technical Linux problems. In that spirit, let's see what we can turn up.

The item you mention appears to be at http://linuxgazette.net/issue57/lg_bytes57.html#software . (I mention that because it'll help the other volunteers if they wish to lend a hand. It's always a good idea to make it easy for others to assist you, if you're hoping for help. Thus, URLs in this case are useful.)

The kit's maintainer, Dr. Ullrich Hafner, also had a January 2001 Linux Journal article on the subject: http://www.linuxjournal.com/article/4367

All of the links returned by various search engines appear broke.

Some amount of googling later, I found that the FIASCO toolkit appears to have become unmaintained and is not found mirrored downstream in any of the usual places. That's unfortunately all too common among open source inventions that were precariously hosted only in one single location and never quite caught on. Ours is a rather carelessly Darwinian software ecology.

Depending on what you're looking for, you might find equivalents of particular tools (cjpeg, djpeg, etc.) in the libjpeg source (currently jpegsrc.v6b.tar.gz) published by the Independent JPEG Group at ftp://ftp.uu.net/graphics/jpeg

NNetpbm (http://netpbm.sourceforge.net) does include pnmtofiasco, which compresses the named pbm, pgm, or ppm image files, or Standard Input if no file is named, and produces a FIASCO file on Standard Output. "

Anyhow, back to Hafner's reference implementation of the FIASCO CODEC -- Internet Archive to the rescue: http://web.archive.org/web/20000829071003/ulli.linuxave.net/fiasco (the last mirrored version of Hafner's page) has mostly broken links -- but the source code RPM package, fiasco-1.1-1.src.rpm, is still retrievable. I have done so, pulled apart the RPM, and extracted its tarball for the benefit of you and posterity. You'll now find it at:

http://linuxmafia.com/pub/linux/utilities-general/fiasco-1.1.tar.gz

There is also a copy of the RPM SPEC file for that source tarball. I hope that helps!

Hi Rick:

Thanks for all the help. I fully understood that your groups mandate was to answer Linux technical problems. However, from past experience volunteer groups are by definition enthusiastic, challenge-oriented people. You have just confirmed this observation.

Thanks again.

Jim

I usually leave out sig quotes, but this delightfully silly and apropos item was attached to Rick Moen's signoff... -- Heather

Mark Moraes: "Usenet is not a right."
Edward Vielmetti: "Usenet is a right, a left, a jab,
and a sharp uppercut to the jaw.
The postman hits!  You have new mail."

(!) [Heather] The real "wanted" for this item - if you happen to be the creator of a nice little project that only lives in one place, see that a few copies are stashed elsewhere. If it has even a tiny following, encourage your pals to keep an extra of it. Otherwise whenever you need to move on, this is likely to be its fate... someone who desperately needs it, will feel terribly lucky if he manages to find it.
If open source is the way to protect some app's future, leaving that futire in only one cabinet won't do the job.

Fw: Vote for Rocard/Mueller for the "European of the Year"

Fri, 14 Oct 2005 23:12:58 +0200
Forwarded by Karl-Heinz Herrmann (The Answer Gang)
(!) [Heather] Normally this would go in News Bytes, but it's a timebound announcement, which warrants its inclusion in Help Wanted.

Hi TAGs,

something for the news maybe?

information missing in the blurb below: The poll opened on 22 September 2005 and will close on 11 November 2005. It is located on the Internet at http://www.ev50.com/poll

K.-H.

Begin forwarded message:

Date: Fri, 14 Oct 2005 19:14:55 +0200
From: Henrion Benjamin <bh from udev.org>
Subject: Vote for Rocard/Mueller for the "European of the Year" award

Dear Supporter of Noepatents.eu.org,

Let me thank you for your participation in our web demonstration in the spring. As you know, our efforts ultimately led to the rejection of the proposed software patent directive by the European Parliament.

We now have the next opportunity to demonstrate the unpopularity of software patents, and to draw attention to our cause. For the most prestigious awards in EU politics, the "EV50 Europeans of the Year Awards" by the European Voice (an EU-focused weekly newspaper), the jury has nominated two software patent critics, Michel Rocard MEP and NoSoftwarePatents.com founder Florian Mueller.

With the power of the Internet, we would like to ensure that our candidates win, and nominees who are in favor of software patents (such as EU commissioner McCreevy) lose. Please participate in the public Internet poll, and you can find voting recommendations and further instructions here: http://www.nosoftwarepatents.com/en/m/ev50/vote.html

Please display a banner on your website:
http://www.nosoftwarepatents.com/en/m/ev50/banners.html

There is a convenient way to inform your friends by email:
http://www.nosoftwarepatents.com/en/m/ev50/notify.html

Those materials are available in 12 languages. The language can be chosen from a pull-down menu at the top of the page, or in the list of languages at the bottom of each page.

Thank you for your continued support,

-- Henrion Benjamin <bh@udev.org> Noepatents.eu.org

Problems with accessing FTP site using windows pc on lan to a

Thu, 6 Oct 2005 15:39:23 +1000
Michael Harvey (cyberzork from gmail.com)

Hi There, Recently a Linux guy (Now moved away) setup a Linux Redhat Firewall server connected to a Cable Modem for our Windows PC's to share Internet etc.

Now the Browsing is ok and Email is ok, but we are having problems with FTP.

We can connect ok to the IP Address and the username/password verifies ok, but then we get a message saying eg:- 

500 I won't open a connection to 192.168.0.13 <http://192.168.0.13> (only to
60.226.143.244 <http://60.226.143.244>)
! Failed "port":
! Retrieve of folder listing failed (0)

Note:- I am on the Windows pc getting the IP Address 192.168.0.13<http://192.168.0.13>;.

Now i tried passive mode using the same ftp port 21 and same problem. The linux guy said we have to mod the IP Tables somewhere, but he is not exactly sure where to do it to allow us to FTP from local windows pc's.? I have included the full error message ftp log below. I found the IPTables File on the Linux box, but have no idea what to do there. Any help would be greatly appreciated. (Full FTP Log Error).

See attached ftp-errorlog.txt

Sun retiring the SISSL

Sat, 03 Sep 2005 12:08:57 +0100
Jimmy O'Regan (The LG Answer Gang)

http://blogs.sun.com/roller/page/webmink?entry=addressing_proliferation_deeds_not_just

Specifically, this means that OpenOffice.org is now LGPL only. They are also asking OSI to not recommend its use.

(!) [Heather] If your project uses SISSL currently, he also asks that you consider a different license for your next revision or release, so the sweep to remove this license from general use can happen quickly. That makes this a Help Wanted to project maintainers. Thanks!

MIT open sources LispM

Tue, 04 Oct 2005 09:36:29 +0100
Forwarded by Jimmy O'Regan (The LG Answer Gang)

As in, Lisp Machine: http://bc.tech.coop/blog/051002.html Code here: http://www.heeltoe.com/retro/mit/mit_cadr_lmss.html

RMS'll probably be happy about this: before he started GNU, he reproduced Symbolics' extensions to the Lisp Machine (http://www.gnu.org/gnu/rms-lisp.html). Much of that work is in this release.

(!) [Heather] It'd be fun to get an article from someone on the stresses of preparing a big project to become open source after many years under an odd or proprietary license.
GENERAL MAIL

[Fwd: Great Article!]

Tue, 04 Oct 2005 09:20:48 +0100
Jimmy O'Regan (The LG Answer Gang)

Wahey! Positive feedback (so far I've also gotten an angry mail on behalf of winlibre, a correction, and some additional information).

Date: Mon, 03 Oct 2005 14:59:56 -0700
From: Chris DiBona <cdibona from google.com>
Subject: Great Article!
To: jimregan@o2.ie

See subject. Well done. Makes me wantto throw you into our interview system :-)

Chris -- 

Open Source Program Manager, Google Inc.
Google's Open Source program can be found at http://code.google.com
Personal Weblog: http://dibona.com

Usually I leave sig blocks out, but I think if anyone thought it was to change Google's mind about "doing cool stuff" this would be the man who'd know.

You article about Google Summer of Code.

Tue, 04 Oct 2005 08:56:05 +0100
Noemi Tozjan (gjr from stcable.co.yu)
Answered by Jimmy O'Regan

Hello,

I find your article very offending, and I think you should modify it in order to reflect the reality, and here is why:

Before I even look at the rest, allow me to apologise in advance: the article has been published, and will not be changed.

Can I have your permission to forward your mail for publication next month?

Only if I have permission to publish your e-mail aswell.

Of course.

1) multiple people working on the installer/updater

Yes, it is true that WinLibre had more than just one people working

on that. But! If you look at the project listings from other mentoring organizations, you can see that the same thing is present in the following mentoring organizations too: Python, Nmap, Internet 2, Drupal and possibly more... Why does it bothers you only at the winlibre project?

The impression I got is that these were separate implementations of the same thing. With the other projects you mentioned, it was clear that these students were working on different aspects of the same project.

Well, it's partially true for the both cases. As for WinLibre, we had 2 students working on separate implementations of the same thing, while the others were working on different aspects. As for the others, I know (source SoC students themselves) that in some cases 2 or more students were working on separate implementations of the same thing aswell. The reason why it is ok, is that the mentoring organizations had no way to know if a student will finish his/her project or not. The number of finished projects is less than 90%, and even this percentage is way above the expectations (66.67%) on the beginning of SoC. So it was fairly reasonable for some mentoring organizations to have 2 students working on the different implementations of the same thing in case the thing was high priority for the mentoring organization.

If this installer is really a single installer with several people working on different aspects, as with the other projects, then please accept my unreserved apologies, and my promise that a correction will be published next month.

To compleatly clear it out, here is the correct list of the WinLibre projects:

Finished projects:

- installer (2 students) - updater (1 student) - control centre (1 student) - MacLibre (1 student) - 3 -original- games (3 students, 1 per game) - Final Touch /or Image Manipulation Tool/ (1 student)

Failed projects: - CDRoast GUI (1 student)

2) Michael Rybak's "Sneaky Snakes" Your text about this project is a complete missinformation. First: please take a look at it before you call it a "snake clone", because it has a totally unique concept, and it is not even similar to the snake game in any aspect. Also, you've used the words: "One project that has been singled out as a waste of funding". Ok, yes it was singled out, but by -one- single person, Uros Trebec, while all the other people on that thread in summer-discuss (which is undoubtable your source) supported the project. Also you should check out Uros's previous posts on summer-discuss, and you will find out that your only source is a person that was planning to cheat in SoC, but got rejected.

I will admit that that message was a source for that opinion, I will also admit the possibility that it was the source for the other places I saw that opinion voiced, but it was not the only source of that or similar opinions.

Can you -please- point out the other sources?

Hmm. Can't find anything at the moment, and I'm starting to suspect that there was only one other source, and it was in comments on Slashdot.

I'll get back to you if I find anything.

Have you actually downloaded it?

No.

Snake clone:

- you have one (or in some cases 2 or more) snake that collects pieces

on the playground and gets larger for each collected piece, if the snake touches itself (or the other snakes), the player loses.

Sneaky Snakes:

- 2 snakes are fighting each other by shooting their parts, they don't

collects pieces and if the snakes touches themselves or the other snake they don't lose.

That said, thank you for correcting me about the game: to my eyes, it looked like a snake clone -- a more modern snake clone, but a snake clone none the less.

Also, because it seems to have been unclear, I did not share that opinion. I thought that my contrasting the game with the installers made that clear, I'm sorry if it didn't.

It wasn't clear.

I would also like to inform you that every single SoC student I've

talked with finds your article very disappointing partially because of the lack of objectivity and because some of the projects were left out from the lists.

As for objectivity, well, the magazine is called Linux Gazette. I can't and won't apologise for giving a better write up to the projects I personally had interest in: the article simply would not have been written otherwise.

Hmm... Especially because it's -Linux- Gazette... You want Linux to be

more popular, or not? Do you want to see the Average Joe using Linux or not? Well, if you do, here is an interesting fact: The Average Joe uses Windows. If you want to attract him to opensource, you must show him that it's good. How can you show him that? You give him opensource for Windows as step one, as a step two, he'll be more openminded about Linux...

I was coming from the PoV that if you show people the cool things they can do with Linux that they can't do with Windows, they'll be more likely to switch. The truth, I suspect, is somewhere between these views.

Projects were only omitted simply because I could find no information about them. Please feel free to tell people to pass on information I missed, either to me personally, or to tag@lists.linuxgazette.net (from which we at Linux Gazette can publish them).

(!) [Heather] The irony that Jimmy had trouble finding some projects sponsored by Google! while others were easy to read about was not lost on us in the back room, but it was his article to write.
Our nature as a magazine is to be by and for ordinary people; this means that some things will be subjective by nature, while others will go to the other extreme and try to cover a subject broadly to avoid attaching themselves to one preference as if it's the only thing that exists - a laudable objective, as well as being objective.

I have already received mail from one of the LiveJournal participants, who provided me with information about several of those projects as well as his own. I am interested in providing a complete list of projects, without the article part. If you would like to help me to complete it, well, thanks. If you would like to complete it yourself, the article is under the terms of the OPL 1.0 -- you may freely modify it, etc. and make your own list.

Well, here is what people wrote on our internal SoC mailing list:

Pradeep Padala: Finally a list of projects http://linuxgazette.net/119/oregan.html

Anil Ramnanan: He missed my project, an Eclipse plugin for Apche Forrest. :(

Sorry about that. Apache was one of the most difficult organisations to find any information about.

Ivan Barrera A.: Mine too, but i didn't like his writting. I also think he was offensive without the need.

What was this project?

Lev Olkhovich: Codehaus also has 12 applicants with only two projects. It is especially strange because the link in the article points to a page with times more projects.

I tried to provide a link to the proposed projects for each mentoring organisation, but only to list the projects that were accepted. The projects listed were the only ones I could find information about.

Pradeep Padala: I actually didn't look at the article closely. It was forward to me by a friend. Looking closely, it shows SoC in a bad light, I agree.

Kai Blin: Also, I think the author didn't really do his homework. I can only talk about the wine projects, but he missed a couple of details there.

I realise that Kai's project has been added to CVS since I submitted the article, but beyond that I don't see what details I missed.

Pradeep Padala: However, I must say, not everything he said is not wrong either.

Best wishes, Noemi Tojzan

Thanks for writing, Jimmy O'Regan

Re: Your article about Google Summer of Code.

Wed, 05 Oct 2005 09:28:36 +0100

Hello,

I've forwarded your replies to the Google Summer of Code internal mailinglist, so I guess people will contact you with their projects if they want to. The attitude of most of them about the article is pretty negative tough, so don't expect miracle.

Well, I guess that must have something to do with the way in which it was presented, because the other mail I've received has been positive.

I'm not representing WinLibre in this issue, I'm writing only my toughts, but I think it would be fair from your side if you'd publicly apologize both to WinLibre and Google because of the unneeded negative criticism presented in your article, especially because you stated that the actual reason of this attack was the fact that the WinLibre is providing OSS for the Windows platform. (note: we do it for MacOS too now, and the 3 games are cross platform)

Erm... no, I don't see how you've come to that conclusion. That WinLibre is (was) a Windows-specific project is the reason I didn't take a very close look at the projects: hence, why I termed 'Sneaky Snakes' a snake clone. I admit that was inaccurate; I should also admit that I have very little interest in games in general: it's most likely that I would have come to the same conclusion no matter how hard I looked at it, simply for lack of a better comparison.

The criticism was aimed at the duplication of effort in the installers. That I called it a "waste of effort" was an error in tact -- I had intended on changing that sentence to read "unnecessary duplication of effort", but it slipped my mind. Even that, though less so, is still going to be offensive to some.

I apologise wholly and unreservedly for the unnecessarily offensive wording of that phrase; however, I will not apologise for having and expressing a negative opinion of the duplication of effort.

As a personal note: I don't understand why is it needed to have a negative attitude about Google Summer of Code in any aspect, when Google invested 2.000.000 USD in opensource. The entire OSS community should be thankfull for this, no matter if there were glitches or not. The only thing you can achieve by this attitude is to stop Google and potentially other companies to make programs like SoC in future. Is really that what you want?

I think you're overreacting here. I made two critical comments of my own in that article (the other was towards criticism made by an Ubuntu representative), the rest was merely a reflection of the criticisms made elsewhere: it would have been a bold-faced lie to say that everyone was entirely happy about how SoC went.

As for Google discontinuing SoC because of some criticism? You must realise how far-fetched that idea is. If Google was that thin-skinned as a company, they wouldn't have survived long enough to have the first.

Regards,
Noemi Tojzan

Take care, Jimmy

Correction for your Summer of Code article

Tue, 04 Oct 2005 10:06:58 +0100
Meredith L. Patterson (mlp from thesmartpolitenerd.com)
Forwarded by Jimmy O'Regan
Mirrors who poll least often, take note. -Heather

OK, I sent the actual correction to the editor gang already, but just in case:

<li><a href="http://austin.cs.uiowa.edu/charun">Charun</a>: more "natural" queries in PostgreSQL (<a href="http://archives.free.net.ph/message/20050601.205033.7e74a30c.en.html">details</a> ;)</li> </ul>

should be:

<li><a href="http://pgfoundry.org/projects/qbe">Query by Example</a>: more "natural" queries in PostgreSQL (<a href="http://archives.free.net.ph/message/20050601.205033.7e74a30c.en.html">details</a> ;)</li> </ul>

Subject: Re: Correction for your Summer of Code article
Date: Tue, 04 Oct 2005 09:27:41 +0100

Meredith L. Patterson wrote:

Hi Jimmy,

Just read your Linux Gazette article on the Summer of Code, and I'd like to submit a correction if at all possible. I was one of the students who worked for Google, and you've got my project listed as Charun, which isn't quite right. Charun was sort of a stunted proof-of-concept that I put together for my qualifying exams, and while it's certainly the intellectual ancestor of Query by Example (my actual project), QBE is (1) far more robust, and (2) actually integrated into PostgreSQL (as in, it modifies the SQL language itself, rather than being an after-the-fact add-on).

You can find QBE at http://pgfoundry.org/projects/qbe, with source code and (sparse) documentation. If it's possible to correct the link in the article, I'd really appreciate it.

I'll pass this on to my editor, who gets final say on what does and does not get corrected.

Can I also have your permission to pass this on to our mailbag for publication next month? (Some of our mirrors are slow to update, and the correction will go unnoticed by anyone who has already read the article).

Absolutely. Thanks very much!

Cheers, Meredith

[Ben] Done and repubbed.
GAZETTE MATTERS

Time to drop the Site Map?

Tue, 4 Oct 2005 10:58:50 -0700 (PDT)
Mike Orr (LG Contributing Editor)

We (LG's editors) are thinking about deleting or changing the Site Map page, and would like to get readers' input on it. The Site Map is basically a concatenation of the monthly TOCs, a kind of poor man's search engine. It was really useful when it was made, but now has become so huge (309 KB) that it's inconvenient to use, and will only get bigger. Also, our search engine at the time alternated between being broken, anemic, and nonexistent. Now we have a Google-powered search, and author pages linking to all articles by each author. So the only other use for the Site Map seems to be offline searching, for people who don't have immediate access to the Internet. Are people using the Site Map page? Does it have any other uses we've missed?

-- Mike Orr

[Ben] Slight correction, at least from my perspective: I'd like to get TAG's input on it. Readers, well... there's no way to get a reasonable-sized sample.
[Heather] The most obvious mechanism to get a reply on this from the readers and NOT the answer gang is to post a query to the readers at large in The Mailbag and/or encourage commentary back about it in the TAG blurb. At least then we know that any responses come from people who actually read the current issue without having the special interest of being contributors and staff.
(!) [Heather] At this point I snip a bit to leave out the collective of opinions so far, but point out the ideas that came of the thread...
[someone among the Gang] As for the Site Map, if we need to keep it (and I'd rather keep it as-is than repackage it into year-chunks or a dynamic page), how about moving the Site Map link off the main menu and changing the home page to:
Earlier issues are in the Archives. Or see the Index of All Issues.
[Ben] I'm curious - not for or against, but simply curious - about the reason that you don't like the idea of a dynamic page. For me, it has the benefit of de-bloating the tarball. Am I missing some way in which it would be a detriment?
[Mike] Whenever I use the site map, I use the version that came on Linux Format's DVD, rather than run up my phone bill... I'm sure we have a quite few off-line readers.
[Thomas] It wouldn't be a detriment --- if the reason you're doing this is to reduce the size of the tarball, then that's fine. I welcome it, as there's probably still a lot of people on dial-up trying to get each issue.
Thinking about it some more, I missed the reason (if there even was one) why the current Debian maintainer doesn't ship the sitemap for each LG deb (this to me was always a grave mistake on his part -- plenty of other debs have much larger amounts of crap in them).
[Heather] Probably the same philosophy that had him leaving out TWDT.
[Thomas] If we came up with a script that would do this for us (perl, awk, whatever -- a language that was certainly common on almost all Linux installations) then it would be a trivial matter to have it called as part of the post-install scripts. A similar process applies for RPMs. Distributing it as part of lg-base.tgz (as I had mentioned in the past) for those that download the raw tarballs from our FTP site would mean they too get the benefit.
That's how I see things. It's a good balance, because:
* You don't need any quantitative measures as to whether a certain amount of people want to keep or remove the TOC.
* It reduces the size of the tarball considerably.
* It means that it can be generated as and when, for those that want it.
Just my thoughts.
[Heather] ... a "limited fare" meta-TOC [only covering issues you have present] would not serve a purpose of spotting a specific past issue to glom. Though the KB does mention the right issue in the hotlink, since our hotlinks mention issue numbers.
[Thomas] Are we going to publish the thread in LG for next month and wait for reader's responses before a plan of action is made?
[Ben] AFAIC, it's a dead letter already. There was never an overwhelming reason for getting rid of it, nobody's adduced one, and I've seen several people jump in fairly quickly to say that they do indeed use it; ergo, it stays.
(!) [Heather] There you have it; we've thought about it, but we're not doing much to it, unless something sways us more solidly that some effort on this would make us a more appreciated 'zine.
Dear readers, if you've something to say, please let us know; mailing gazette@ (LG Editors) linuxgazette.net to reach us directly or tag@ (The Answer Gang) lists.linuxgazette.net to reach the Answer Gang as a whole (yes, we're a list, just a very open one) will be equally accepted. And thanks for keeping up with the Gazette :)

[Linux Gazette 118]Missing file?

Fri, 9 Sep 2005 20:27:04 +0000
Predrag Ivanovic (predivan from ptt.yu)
Here's what we were doing among the editor gang on 9/11 this year...
Offline mirrors take note :) -- Heather

Hello,Gang. I am reading Mailbag from current Gazette,and it seems that a file is missing from lg-118 tarball.It's lg/current/misc/wanted/philip_obrien.grub.conf.txt.

Pedja

...............

"....If Windows XP were the size of the Empire State Building, then the little barking Beagle virus - the size of a small dog - can come in through the front door, lift its leg, deliver its payload, and somehow cause the entire building to come crumbling down." -- theregister.co.uk

...............

[Jimmy] Well, I have the file, ready and waiting to go into SVN, but 'svn up' isn't doing anything.
[Ben] OK, file's been added.
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/

 

Copyright © 2005, . 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 120 of Linux Gazette, November 2005

Tux