Social Contract Amendment May Bump Sarge To 2005

An anonymous reader submits "Debian's Release Manager Anthony Towns announced that after the Grand Resolution to amend the Social Contract has been successful (it does not only apply to software any more), vital parts to modern Linux systems, such as important documentation, firmware needed for proper hardware support will have to be removed from the distribution before the next release. Moreover, the upcoming installer will need to be changed. He goes on to say that he does not expect this to happen by the end of this year which means that Sarge will not be released in 2004."

Comments

Why can't they

[Anonymous Coward]

Release one version with the new contract next year, and one without it sooner? Call it sarge- and sarge+.

Re:Why can't they

[Openstandards.net]

That's what I was thinking. Why not call it the Social Contract of 2005?

Or amend the social contract to promise hardware support and then prioritize the goals to hardware support takes a priority until a "free" option is created or becomes available?

It appears as though Debian is going to take a big step backwards if something isn't done. The goals are clearly good, yet the real world has always required a compromise between the ideal and the real. Don't the Debian developers actually work in IT for a living?

I'm really concerned about this, because I was highly considering Debian for the next OS to try since RH is discontinuing free security updates, and I'm not sure at all how Fedora is supposed to address it. The last thing I need, though, is a hardware problem, particularly with a network card.

Re:Why can't they

[Wyzard]

Nobody's saying that your proprietary hardware will cease to work in Debian. The packages will still exist; they'll just be in the "non-free" section, separated out so that people who don't want any non-free software can omit that section from their sources.list file. Non-free packages are technically not part of Debian, but if you have a non-free line in your sources.list, there's no difference whatsoever in how you use them.

Re:Why can't they

[FuegoFuerte]

It's going to be hard to say this without people getting their panties in a knot, but I absolutely hate trying to 1) get debian systems running how I want and 2) keep them running how I want.

Maybe it's just me, but it seems like if some of the Debian folk spent as much time fixing their distribution as they do ranting about the philosophy behind their system, it could just about literally jump into my computer, read my mind, and magically do everything I wanted without me touching a keyboard. As it is though, I'm forever saying to myself "now where the hell did they put THAT file, since it's not in its standard location..." and "what version is this package really? It looks like version 3.1 from 2 years ago.... no wait, that's 3.1-15... wtf is the -15? It has features that weren't released until 3.9? Huh?!??" and similar.

I once made the mistake of trying to figure out what flags were being used to compile a Debian package... after jumping around through about 7 different intertwined and slightly obfuscated shell scripts for about an hour, I gave up.

Unfortunately, I'm still stuck using Debian on one server (the owner doesn't want to change OSs), but I've gone to Slackware on all my systems. Much simpler system to deal with overall, IMNSHO.

Re:Why can't they

[Wyzard]

Standard location? In my mind, the "standard" location for a file is where Debian puts it, and I get confused when it's located somewhere else in another distribution. :-)

I get the sense that you're used to installing things via "configure; make; make install". It's good to have a simple method like that available, but when I talk about maintainability of a system, installing non-packaged software is one of the biggest ways to hurt that maintainability. Files created by a "make install" usually don't have any way to cleanly remove or upgrade them; you can upgrade by installing a new version over the old, but if the old version included any files that were removed in the new version, you still have that cruft sitting around. You get the idea.

I like the fact that Debian has lots of infrastructure. I like to know that when I install a package, it will cleanly integrate with other related packages, and when I remove it, it will cleanly go away. I like the fact that when I'm looking for a package that performs a certain function, I can often guess its name, thanks to fairly consistent naming patterns, and that when I'm looking for a file, I can usually guess where it's located due to a consistent and sensible filesystem hierarchy.

I hang out in #debian on IRC, and I read some of the mailing lists, and I see a lot more discussion on practical matters than on philosophy, and philosophical rants are pretty rare. The system works quite well for those who use it; your comment about "fixing their distribution" just doesn't apply. Remember that Debian is run democratically: if you don't like the way something's being done, you can always register as a Debian Developer and vote to do things your way. If you don't want to do that, or if you get outvoted by people who like things the way they are, you can use another distribution and nobody will hold it against you.

Re:Why can't they

[Welsh Dwarf]

You have just pointed out the #1 reason why sysadmins who compile from source on production servers need a beating with a clue stick. I'm not going to get all superior, because I know that at one time, I did the same thing. The point is, to put something new on a production machine (like samba with acl support for debian) you: -Compile it on a development box with prefix=/tmp/what_you_want -Make a package of it -test the package on a second test box to make sure it works -install the package on your server This provides several advantages besides the one you stated: Firstly, you never have to have dev tools on your production server (and a lot of rootkits depend on these being present). Secondly, you are sure that when you deploy, it's quick and painless, and you won't brake your server with a botched compile. Thirdly, you can deploy then on multiple servers quickly and efficiently.

Re:Why can't they

[FuegoFuerte]

Standard location? In my mind, the "standard" location for a file is where Debian puts it, and I get confused when it's located somewhere else in another distribution. :-)

I tend to think of the "standard" location as being where I can find a file on better than 75% of the *nix systems I've used (Several linux distros, BSDs, HP-UX, Irix, Solaris, etc). But whatever makes a person happy I suppose. For a linux distro, I'd say the standard location should be where the LSB says it is, which from what I've seen Slackware tends to follow a lot closer than Debian.

I get the sense that you're used to installing things via "configure; make; make install". It's good to have a simple method like that available, but when I talk about maintainability of a system, installing non-packaged software is one of the biggest ways to hurt that maintainability. Files created by a "make install" usually don't have any way to cleanly remove or upgrade them; you can upgrade by installing a new version over the old, but if the old version included any files that were removed in the new version, you still have that cruft sitting around. You get the idea.

Actually, I typically install things via "installpkg ---.tgz" and upgrading things via "upgradepkg ---.tgz". When I install from source, I use "configure; make; checkinstall 'make install'". This makes a package out of it that I can easily install, remove, upgrade, whatever I want.

As far as philosophy goes, well, I think you're well on your way to proving my point with your small dissertation on democracy and voting.

To be honest (and here I'll get philosophical), I think the democracy thing may be part of what's (IMO) wrong with Debian... the masses are often wrong, and most of the people in the masses will never known an entire distribution as intimately as, say, Pat knows Slackware. Therefore things take longer to get fixed, the whole process gets bogged down, and the distribution suffers because of it. I have known people using Debian Unstable who have had things like SSH stop working properly upon 'apt-get dist-upgrade' or whatever the command is. Meanwhile, I am almost always running slackware-current (aka the unstable, in-development branch of slack), and have never once had these problems in the several years I've been using slack.

Re:Why can't they

[spotteddog]

Don't the Debian developers actually work in IT for a living?

Many of us do. Many do not. Many are students. Debian is very diverse.

Some of us who "actually work in IT" view our Debian work as a way to fix what we view as broken in the "mainstream (MS dominated)" IT world. One of those broken things is the lack of accountability, stability, and reliability in all facets of "mainstream/modern" mass produced IT systems.

I compromise my "idealism" with respect to computing systems at the job I get paid for (because I'm not the "big boss"). I am not willing to do so in my volunteer work.

The Debian project will release "when it is ready." It is ready when the software, and all the other bits meet the creiteria set forth in the Debian Social Contract (and the release manager give it their blessing).

NOTE: The opinions expressed are my one and do not necessarily represent those Debian project. I am a Debian Developer.

Re:Why can't they

[dondelelcaro]

For a linux distro, I'd say the standard location should be where the LSB says it is, which from what I've seen Slackware tends to follow a lot closer than Debian.

By default, Debian Policy stays in sync with the LSB. However, there are a few places where the LSB is suboptimal, so Debian documents the differences in policy, and does the right thing. Is there a particular set of files that you're talking about that you feel is in the wrong place?

Actually, I typically install things via "installpkg ---.tgz" and upgrading things via "upgradepkg ---.tgz". When I install from source, I use "configure; make; checkinstall 'make install'". This makes a package out of it that I can easily install, remove, upgrade, whatever I want.

And here, I do the following to upgrade or install a package:
dpkg -i foo.deb;
Or:
aptitude update && aptitude upgrade;
Finally, if I need to build and something from source, it's as simple as:

apt-get source foo;
apt-get install build-essential fakeroot;
apt-get builddep foo;
cd foo-*;
# Change how ./configure or make or R CMD is called
$EDITOR debian/rules;
# Build the .deb
fakeroot debian/rules binary;
# install the .deb
dpkg -i ../foo_*.deb

There's nothing magical there at all. The rules file calls make in the build target, and everything else happens automatically. [Now, if you don't know how to modify a make file, perhaps you shouldn't be building stuff from source?]

Debian standards

[Anonymous Coward]

All I have to say is, good for them for sticking to their standards.

Re:Debian is fading into irrelevence?

[Rura Penthe]

Who the hell uses the 2.6 kernel in a production setting? I know I don't. In fact, I vastly prefer to admin servers on debian because finding updated packages is typically EASIER than on Redhat without a support contract. Some server admins prefer the slower moving target of debian releases and the ease of backports.org for packages they NEED upgraded.

Re:Debian is fading into irrelevence?

[dilvie]

Sarge does not need to be "released" to be relevant. The software still runs on the debian OS. You can still upgrade your kernel. IMO, the best thing about debian is that you can start with a very minimal linux install and add the packages you need as you need them a lot easier than is possible with many other distributions.

up to date?

[linuxbert]

i use and like debian. but i want new packages to be released, and for it to generally appear to be supported.

if you dont want non-free stuff, fine, release sarge, its almost ready (and long delayed) and make removal of non free packages a goal of the next release.

Re:Debian is fading into irrelevence?

[ObviousGuy]

Replying to my own post.

I look at the landscape and wonder when a distribution that takes a pragmatic look at Linux stability like Debian does without the associated religious zealotry will come into being.

It is an amazing thing to see the focus on stability and completeness in Debian. I deride the distro in jest, but it is a sign of good thinking that they don't declare a stable release every other weekend like some other distros.

However, the zeal to make Debian a "Free" distro is hampering it, causing the maintainers to shun obviously useful and necessary utilities because of the flavor of its license.

I wonder how long it will take for Debian to fork between the GNU-like religious faction and the pragmatists.

Re:Debian is fading into irrelevence?

[AstrumPreliator]

I think there's a point you missed though. Even though Debian is well supported and is infact a good linux distro, it's still aged. There is a definite trend towards open source operating systems right now and I think Debian is going to miss out on this to sa large extent. People (and by people I don't mean individual computer enthusiasts) are going for the more main stream and updated distros such as Fedora. Once you install a distro on a box you tend not to just replace it with a different distro on a whim. I know people who use Red Hat 7.3 for servers still and as long as the hard drive is alive and assuming they don't kill the OS, it will be running for a long time to come.

In that respect I do think distros such as Debian and Gentoo will fade away to a large extent. They will always be around, but not widely used.

Of course thats just my opinion.

Rock solid stability

[rufey]

Although I am disappointed that Sarge will take a bit longer, for my two servers at home, Woody has been rock solid. Sure I grabbed, compiled, configured, and am running 2.6.4 on one of them, but its still Woody under the hood. If I need something newer than what Woody has, I grab the source and build it myself (OSS is great because of this). I like machines that just run and don't need to be upgraded every other month. The bleeding edge releases of other vendors simply doesn't offer anything I need for a web, mySql, Sendmail/Postfix , and DNS server.

We have machines at work that are currently running Redhat 7.2. A couple are RedHat 8, 9, and RHEL 2.1. Why are they not all running the latest and greatest RedHat? Because we either can't afford the downtime (not to mention configuration) to upgrade every time that RedHat comes out with its next release, or the bleeding edge releases break things. Unless a newer release provides some feature/function that we need in production and we can't get any other way, we don't upgrade each time a release comes out. We've even downgraded a couple of machines from RHEL 3.0 to 2.1 because getting some Oracle software installed was near imposible (even with Oracle consultants on-site!)

I'd much have a rock solid server that performs its job all the time than have a bleeding edge server that requires 2 or 3 upgrades a year just to stay bleeding edge.

Re:Rock solid stability

[matlhDam]

The good people at Fedora Legacy are providing updated RPMs for RH 7.2, 7.3, and 8.0, and will be providing support for 9.0 when it's EOLed in a few days. They've been a bit sluggish getting update RPMs out recently due to download server problems, but the support's still there.

Re:Rock solid stability

[rufey]

While I agree that having older operating systems can pose a security risk, with proper administration, you can reduce that greatly.

Our older RedHat systems don't run all the default services. In fact, most of them can only be gotten to (from the outside) on port 80 and or 443, which runs the latest and greatest Apache/Java/Tomcat. Sure, inside we also have ssh access, but with a firewall and intrusion detection between it and the Internet, its a good bet that port 80/443 is all thats open. We have most of nfs, ftp, telnet, smtp, snmp, pop, imap, finger, echo, nis, and any other service turned off. If they are not running, they are not vulnerable.

Running something like Apache 2.0.48 on RH 7.2 isn't necessarily going to be any less secure than running Apache 2.0.48 on RHEL 3.0. Any successful compromise of Apache is likely to be due to Apache, not the underlying OS. Fixing Apache, not upgrading the OS, would be the solution here, assuming you build Apache yourself from source, which I tend to do.

Re:"Grand Resolution" ?

[The Musician]

The slashdot summary that says "Grand Resolution" in wrong. The proper expansion of GR is "General Resolution".

See http://www.debian.org/devel/constitution, in particular sections #4 and A.

Great. Just great.

[JessLeah]

As more and more pieces of hardware will be protected by the ever-intensifying "intellectual property" laws, Debian will get more and more worthless. It's quite simple: In the distant past, manufacturers made specs available to those who asked. Then they stopped doing it, but you could reverse-engineer them. Then a few of them succumbed to commercial pressure (and the desire to look like "nice guys" to geeks who might influence corporate purchasing decisions) and released binary-only, proprietary drivers for the most popular Linux distro(s) (read: Red Hat and, if you're lucky, Mandrake and/or SuSE). Now most companies don't even bother doing that, and there is a growing trend towards the use of wrappers and such around Windows (!!!) DLL-based drivers. Linux's future is one of proprietary drivers and payware wrappers around proprietary Windows drivers.

And the Debian people are rejecting this sort of thing because of their morals. That's really great. It's also, unfortunately, a wonderful way to ensure that Debian only has primitive, reverse-engineered, DMCA-illegal, flaky support for newer hardware.

Let's see. nVidia and ATI both have proprietary binary-only drivers for Linux (which of course ONLY work on Linux/x86, not Linux/PPC or Linux/ARM or Linux/SPARC or whateverthehell), right? DriverLoader is required to use a bunch of WiFi chipsets under Linux, using Windows .DLLs. Mplayer (that favourite of rebellious geeks) uses Windows .DLLs. Am I forgetting any similar projects? And the kernel is full of various drivers (think sound drivers) which ask for proprietary pieces of firmware, right? I suppose the Debian folks are going to rip out support for all of these devices?

I LIKE the Debian project's inherent sense of morality. I DON'T like their ridiculous lack of pragmatism. This sort of antic is only going to drive off more moderate users towards the likes of Fedora (bloatbloatbloat), Lindows^WLinspire (Windows wannabe, bloat), and ... well, and Windows itself. Way to go, guys.

Re:Great. Just great.

[Phleg]

This is what contrib and non-free are for. Debian's new social contract simply says that it will not depend upon non-free goods of any sort--not that it won't be provided.

Re:Great. Just great.

[Wyzard]

Most likely, the non-free stuff will not be completely removed, but rather, moved to the non-free section of the distribution. (Strictly speaking, non-free isn't part of the distribution, so things moved there have been removed from Debian, but the packages are available from the same servers, and interoperate with the free stuff.)

The NVidia drivers, for example, work just fine in Debian. Not only is the nvidia-kernel-source package available via apt-get, but it works with Debian's kernel-package build system to produce .deb packages of the built modules. The fact that the drivers are non-free don't affect me in the least; I use them the same way I use any other third-party module package.

I'm not particularly bothered by this change. I'm slightly bothered by the delay of the Sarge release, but since I run Sid on my desktop, and my servers do fine with Woody plus an occasional backport, it's not that big a deal. And it does pretty much answer the question of whether GNOME 2.6 will make it into Sarge. :-)

Re:Debian is fading into irrelevence?

[nuclear305]

Oh yea?

Sometimes I wonder if all the distro zealots have stopped to realize this one simple fact:

Every distro is using OSS. Yes, it may be tweaked and patched here and there..but beneath all the branding and logos...it's STILL the same software.

[Of course, I'll throw in my 2 cents as well. I've used both Debian and Gentoo. Honestly, Gentoo is my path to take...I now run it on 3 servers [1 is production] and a firewall and once they're set up..they just run. Not to mention extremely easy to update. ]

Rather than whine, help

[Morganth]

I can see why some people think Debian is fading into irrelevance. Even running sid feels like you're "behind the curve" (in terms of what my Gentoo friends are emerging), and sid is already well ahead of sarge and perhaps years ahead of woody.

Nonetheless, instead of complaining about it, why not help a hand. One first project ot look is Debian on the Desktop.

Maybe from there desktop users can pull enough weight to a) get the latest desktop packages into sid (or at least, at worst, experimental); b) utilize existing apt-get source framework to allow for rapid from-source installs of bleeding edge apps, to reduce packaging time; c) further tweak app to prevent the already-rare occurences of dependecy hell (or, more appropriately, must-remove-to-upgrade hell).

But please, don't do what I do. Don't whine. Just try and help. I think Debian needs a community of [young] desktop users to sort of provide a voice alongside the old-timers who care more about stable servers than Gnome 2.6 or whatnot.

It's a shame

[stootles]

After reading this we are now discussing what distro to look at? I don't feel like fighting with management about another distro - Debian was so good that the battle with them was worth it, we just found out we have Suse licenses, so that could be the go - plus we will actually have a big red support button - running 20 servers and I have never had a support button before, that'll be different.

Seriously Debian is great, but, this is a ?harsh? reminder that Debian is not developed for users - never has been - it is developed for the developers making Debian.

However, I see a possible bonus here for those commercial distro's using Debian as they will be able to insert the non-free stuff into their own distro's. From what I saw it seems alot of people would start with something like progeny, but end up migrating to Debian proper - maybe this will give those companies a fighting chance to keep their linux users.

Cheers,

Stewart

Interesting

[aws4y]

I think that the Debian people are right on this one , however the FSF foundation is partly to blame because of the invariant sections in the FDL . (why glibc wont have documentation)

I think the solution, since non-free is being kept, should be to include the non-free repositories in the default "sources.list" file and allow tasksel to use non-free packages for documentation under a "Non-Free documentation" header, no non-free stuff should needed for the bootstrap installation(although binary kernel module won't be available by default). Thats the best comprimise, IMHO.

Could we stop the Microsoft, Debian, Gentoo and Fedora, and *BSD astroturfing please?

A good push-back

[Animats]

This is a good thing. Debian is pushing back against increasingly proprietary hardware.

Now we need a logo for open-source hardware, so people know what to buy. Preferably one designed by a competent icon designer, like Susan Kare.

Re:A good push-back

[TWX]

"Now we need a logo for open-source hardware, so people know what to buy. Preferably one designed by a competent icon designer, like Susan Kare."

No, we need the DMCA repealed, so developers within the USA won't be afraid to reverse-engineer some company's hardware. We need companies to release their own source, or to at least provide a high quality binary driver if the development is slow enough (XFree86 as an example) so that people can use Linux on their new cheap machines.

For the longest time there was no decent support for the sound chipset on my motherboard. Fortunately the OEM didn't skimp on the PCI slots, so adding a Sound Blaster with an EMU10K wasn't a big deal, but it would be nice for this stuff to just work. I had an i815 based machine at work that has problems with sound, video, and ethernet. The video stuff came in first, but I had added cards to handle sound and ethernet, and even if they have a solution at this point it's irrevelant to me since I came up with a stable workaround. Most users aren't going to want to do that though, they just want the damn thing to work.

I use Debian on my computers. I like it. It's easy to maintain, stable on the servers, and fairly easy to keep current enough for my tastes with Sid. I install it and I don't think about it anymore, excepting security updates. I have a computer that's been up for the better part of a year (non-public:) and doesn't give me any fits. I could probably automate the apt-get update && apt-get upgrade procedure and still not worry.

Debian is for Slackware admins that got lazy.

Re:This will help?

[Anonymous Coward]

With this focus, Debian will never become a more widely-used distro and remain the distro of the uber-geek, but perhaps that's the point after all.

Er, no on both counts. Debian's goal is not to be widely used. And neither is it Debian's goal to be the distro of the uber-geek.

Debian's goal is to assemble the best operating system it can that you can freely redistribute to your friends and family. Debian's goal is to assemble the best operating system it can that you can sell to others for profit. Debian's goal is to assemble the best operating system that it can whereby you can modify it to fit your needs. Debian's goal is to respect you, the customer.

How many other operating systems can say that?

Perspective

[The Musician]

A few disclaimers, from someone who reads debian-* lists regularly, but isn't part of the project...

(1) Much of what is proposed is about moving pieces of the OS from the "main" archive to the "non-free" archive; "main" is what you get on the Debian CDs, "non-free" is available via ftp. So it is probably less convenient to obtain, but not totally expunged from debian.

Of course, components that affect your initial installation are more sensitive to the method of distribution, but other projects are welcome to build mixed installer tools that combine the default debian installer with the non-free firmware.

(2) This was only announced about 24 hours ago. Things are still in a state of flux, so don't take the "all this is happening and sarge is now year(s) away" too literally.

(3) Don't read into the summary that this solely a personal decision by Anthony Towns, or that he is necessarily in favor of the proposed changes.

Re:Debian is fading into irrelevence?

[Grail]

You mistake the Debian maintainers' pragmatic licencing approach for religious zealotry.

They are approaching the Debian GNU/Linux as a Free Software project, not a feature rich distribution project. Once you yourself can understand what the philosophy of the Debian project, you might understand that they are being incredibly pragmatic.

Regardless of how long Copyright is extended for (eg: Disney's current goal of forever - 1 day), no matter how tight the DMCA becomes, you will always be allowed to run the complete Debian GNU/Linux operating system.

Licencing and legal restrictions on your hardware may prevent you running Debian on your specific hardware (thanks to "Trustworthy Computing" taking over from "binary only"), but there will be no licence or legal restrictions to your using Debian on any hardware that it does work on.

You have to be a special type of person to be a Debian developer - these are people who want to dedicate their time to having an operating system they can safely give to their friends and family without risking a gaol term. People who aren't Debian developers (or fanatical users) are the ones who'd hand over their soul for the next cool gimmick ("yes, I'll accept the condition of only running the software you let me, if you'll let me pay $200 for Halo 4! That game's so cool I don't need freedom!").

Re:Debian is fading into irrelevence?

[ian mills]

Why do people continually compare Debian Stable to bleeding edge Linux distros? If you want bleeding edge use Debian Unstable, that's what it's there for. In anycase 2.4 is not "years behind." There is nothing stopping you from using bleeding edge software and proprietary software with Debian, I do it all the time, they just seperate out the nonfree stuff for the people who care about that. It really is a "feature" not a "bug."

Businesses use Redhat because they offer commerical support, something I don't believe Debian offers, as Debian is not a commercial enterprise.

This is actually an advantage.

[rice_burners_suck]

I think my first reaction at reading this announcement was one of disappointment. This seems a lot like a step backwards, especially when many important components are affected. However, after a bit of thought on this subject, I have come to the conclusion that this is a good idea anyway.

First, if I am interpreting this correctly, this entire issue revolves around Debian remaining 100% free (under a certain definition of "free"), and not requiring the use of any non-free component. This is in stark contrast to, say, the NetBSD project, which is a bit more lax on which "free"/"open" licenses qualify for inclusion in their software. Their idea is that they do not have infinite time to reproduce every single component under the BSD license, so inclusion of other software benefits the community. This would seem to place Debian at a disadvantage.

But upon further reflection, I reminded myself that free software is all about the freedom to choose. In other words, I can choose to use Debian, or not, and further, if I choose to use Debian, nobody said that I can't install components from other distros, specific developers whose software was not included, or even earlier versions of Debian. Therefore, this becomes a great advantage to the community: A 100% "free" distribution, into which you can add whatever components, free or not, that you wish.

Just great? Could be awesome.

[green_crocadilian]

I certainly see your point that for some popular devices (e.g. nVidia cards), the proprietary drivers are much better than the open ones. I also agree that going the moral road will turn away some users.

However, Debian is imho big enough to pull this sort of thing. If some micro-league, half-assed distro went this route, it would die in obscurity, but a major product like Debian will survive. Furthermore, by rejecting proprietary firmware and documentation, Debian is raising awareness of some important issues (like Fedora not including mp3 support raises awareness of patent encumbered technologies).

Say you get a shiny new pci card with a little tux on the box, and a proprietary driver on the CD. Cool, huh? No. Not cool. The driver will work with your Linux system provided:
-you use kernel 2.4 or maybe 2.6
-you compiled said kernel with gcc 3.2 or 3.3
-you use glibc 2.somethingorother
-your /etc, /dev, and /proc are set up just right.

Years pass. Linux gains 20% desktop market share. Duke Nukem Forever is released for Mac and Lintel. You fish out an old computer from your closet; you want to install a Linux (kernel 3.0; compiled with gcc 3.5; with glibc 2.somethingelse; and a GNU/Darwin directory layout) to turn into a streaming virtual reality server for your apartment. Guess what's the probability of your closed-source driver still working?

Open source drivers might be a hassle to use in the short term, but C source is still the most portable way to distribute software.

Re:Debian has shot itself in the foot

[The Musician]

It is true that the "apt" part of debian -- a modular packaging system with dependencies that are carefully expressed and automatically installed -- is a beautiful thing, and that other distros have similar tools.

However, even just looking at the practical (rather that philosophical) side, apt is not the best part of Debian.

The best part of Debian is a set of packaging conventions (the Debian Policy Manual), and a set of tools and a QA system to support that system, with the result that the software you use from Debian is consistently well-integrated, even though it is crafted by a distributed group of volunteers.

The social norms and continual build-up and exposition of best practices, expressed in part by the Debian Policy Manual, is really the best practical characteristic of Debian.

The freedom thing (and corresponding attention to software licensing) is nice, too.

You know what this means.

[Visceral Monkey]

It's obviously the result of a grand conspiracy involving the masons, jews and gentoo users. Those dirty, dirty Gentoo users..

Re:Debian has shot itself in the foot

[Phleg]

My company could never have put up with such a slow and unreliable release schedule.

I'd say the release schedule is quite reliable. "Never" is about as reliable as you can get. Joking aside, I don't see what the big problem is. Does your company actually *need* bleeding edge features provided by many packages? The truth is usually no, and that the unmatched stability and reliability of having older packages with fewer features is a better investment. If the answer is yes, it's very rarely for anything but a few packages, which can be upgraded easily through apt pinning.

The only updates that are absolutely critical are security patches. And thankfully, unlike some Operating Systems and distributions, Debian only provides the security fix for its stable branch, and doesn't require you to update the package to a newer version. This means that less bugs have a chance of being introduced in a security patch, which in turn allows companies to install patches with less worrying about whether or not it will break a current installation. It's still possible, but massively less likely.

In fact, when you get down to it, Sarge is pretty much completely usable as it is. The servers I administer which *do* happen to need those newer features are all running Sarge problem-free.

Really, for all the complaining about Debian, almost none of it is founded on anything rational. Think it's outdated? Run sarge or sid--you lose nothing. Think they're being too pedantic about code and documentation released under non-free licenses? Point apt at the contrib and non-free branches. They're even included into the Debian architecture, including bug reports, mailing lists, and apt entries, so you don't have to go out of your way to do anything special. It's literally no more than changing one word in a configuration file to fix both of these "problems".

Re:Debian has shot itself in the foot

[pnot]

From the Debian website:

The Debian Project is an association of individuals who have made common cause to create a free operating system.

It's a question of defining your goals. You're criticising Debian because their project isn't achieving what *you* see as the ideal goal of a Linux distro.

Debian is not accountable to you. Debian is accountable to its developers -- and as the vote shows, they overwhelmingly support freedom over world domination.

As you point out, there are other distributions which settle on a different compromise between freedom and ease of use. You are, of course, welcome to use these. But frankly I think it's a little cheeky to lambast Debian for not conforming to YOUR idea of what THEIR goals should be. Why do you unleash such bitterness against something you profess not to care about? If you're right, Debian will die quietly and it won't make any difference to you.

Debian is quickly becoming the dinosaur of Linux distributions and is pulling an RMS and hurting the cause of Free Software by marginalizing itself with extremism such that no serious users or organisations will want to be associated with it.

Oddly enough, eweek doesn't agree:

According to a Netcraft Ltd. report covering July 2003 to January 2004, Debian was the fastest-growing distribution among Linux Web servers, and Debian trailed only Red Hat Inc.'s Red Hat Linux in the number of Web sites it serves.

But hey, I'm happy with Debian, you're happy with Fedora. No need to make a flamefest of it.

Will RMS finally recommend Debian?

[steveha]

One of the cool things about Debian is they way they relentlessly review things to make sure they are free. The Debian Free Software Guidlines (DFSG) are rigorously applied, and anything that doesn't meet DFSG is not allowed in Debian's "main" area.

(I tell people "Debian is fanatic about this stuff so we don't have to be." If you just use Debian main, you are using nothing but free software. Easy!)

Debian has two areas for software that doesn't meet the DFSG: "contrib" and "non-free". Now that this proposal has passed, not only software but documentation and firmware will be migrated out of main and into contrib or non-free.

The first thing I thought when I read this was: I wonder if Richard Stallman will finally be satisfied?

Last August, RMS was asked in an interview, which distribution of GNU/Linux he would recommend. He said he would recommend GNU/LinEx, because it contains no non-free software. As it turns out, he was mistaken about that; GNU/LinEx still has traces of non-free software in it, just as Debian has. He withdrew the recommendation of GNU/LinEx (without, to my knowledge, offering any recommendation to replace it).

RMS has said that he cannot recommend any distro that offers up free and non-free software from the same servers, or contains references to any servers that offer non-free software. (Keep in mind that his definition of non-free is not identical to the "non-free" of the Debian project.) So Debian, the most free distro I know, is still not recommended by RMS.

You can read a somewhat acrimonious discussion thread about this here if you like:

linux.debian.legal discussion archived by groups.google.com

Note that Debian is so committed to free software that they are booting FSF documentation from main, because of the newest version of the "Free Documentation License" that allows invariant sections. Invariant sections are clearly free according to the FSF, but they are not in compliance with the DFSG, and thus do not go in main anymore. Discussion here:

another linux.debian.legal discussion archived by groups.google.com

I will close with a final quote from RMS, on the possibility that Debian might one day strip out the non-free software to his satisfaction:

The change that I asked Debian developers to make, some years ago, was to separate the two, such that we could refer people to Debian GNU/Linux without in the same act referring them also to the non-free software. This would make it possible for us to refer the public to Debian GNU/Linux. If in the future Debian GNU/Linux does not include the GNU manuals, this reference could not be wholeheartedly positive, but we could still make the reference.

P.S. If you asked me for a recommendation for a truly free distro, I'd suggest Debian main. If you don't put contrib and non-free in your sources.list file, you will never get any contrib or non-free software and yours system will be fully free software. That's good enough for me, even though it's not good enough for RMS.

steveha

Re:Will RMS finally recommend Debian?

[black mariah]

At some point everyone just needs to stop giving a fuck what some bearded hippie thinks (not that being a bearded hippie is a BAD thing) and just get the fuck on with it. Look, I respect RMS and the FSF as much as anyone else that uses Linux but I seriously don't give a flying fuck what RMS does and doesn't like. It makes no difference to me, and it shouldn't make a difference to anyone else. RMS is NOT divine. He's just a guy. His opinion is no more relevant than anyone else's.

And there's more!

[RAMMS EIN]

Adding to your post, I would like to make the case for Debian unstable.

Unstable in no way means it's really unstable. What it means is that while packages have had some basic testing, the distribution as a whole hasn't been religiously tested, and, consequently, isn't years behind the curve as stable is.

Packages in unstable often provide improvements and bugfixes that the versions in stable didn't yet have.* This means that, while stable is guaranteed to be stable, many people will find unstable more usable (especially people using Gaim, as the IM networks change protocols once in a while, breaking older versions).

The message is, if you want guaranteed stability, use Debian stable. If you want to stay current, but still want to have the benefits of Debian (easy software installation, automatic dependency resolution), use Debian unstable. Don't use testing, unless you really intend to test it - it's almost guaranteed to be broken.

* Note that security fixes are backported to stable. This means that you can keep using the version of the package you have always used, and be sure your configuration keeps working, while still getting security updates that are only available through upgrading for other distros.

Re:And there's more!

[Anonymous Coward]

You appear to have confused "testing" and "unstable."

Re:And there's more!

[pohl]

I'm clear on the intended semantics of "stable", "testing", and "unstable", but my experience is exactly the same as the grandparent post. There may be an occasional burp in the updates, but they go away fairly quickly. In contrast it seems like it takes forever for fixes to get promoted to testing. I've been a debian unstable user for years, and feel like it's the sweet spot to be at for the best balance of recent software and ease of maintenance. Face it, the debian development process is so freaking conservative that even their unstable branch is more stable than other distributions.

This is for my workstation, of course...I wouldn't maintain a server that way.

Slow release cycle? It is not that slow

[gnuman99]

For people that don't know, Debian has 3 distributions: unstable (Sid), testing (Sarge) and stable (Woody). This means that if you want most up-to-date software, you run Sid and cope with some possible breakage (I didn't have anything broken badly over the last few years).

If you want stuff up to date, but want to have something that would be considered "stable" by other distros, you run Sarge (or testing).

The Woody distribution is for cases when you want to run a bunch of applications predictably. This means that your production application will run the same on day one as it does on day 100. An update will not break your application. An update will not change the way the application works. That is the point of stable - stable operation for a long period of time.

And yes, you can install 2.4.26 in Woody (from kernel.org). Woody actually has 2.4.x kernels no matter what the trolls are talking about.

Debians own social standards???

[quinkin]

Hrm... Having read the majority of the newsgroup postings on the topic I think I can say with some certainty that the Debianites really need to work on their own social standards.

They are astonishingly rude and confrontational in an entirely unproductive way. Sure it's probably unfair to point the finger at Debian alone (especially on /. - oh the irony) but I can say with some certainty that nothing positive will come from that thread. Conflict resolution amongst egotistical (come on , we can admit it) geeks is damn difficult - especially when programmer opinions take on the form of religious zealotry (free vs. libre).

These are big changes, and many people are expressing that they felt misled with the "editorial changes" description of the vote in question. I am not going to get involved in an internal dispute, except to say that it is in the best interests of the project for the majority not to feel manipulated and/or deceived. Again, I'm not saying they have been, I am saying that is what some are expressing.

Q.

Re:Debians own social standards???

[Karora]

As with any group of over 1000 individuals, your statement is a vast generalisation.

How can you call us "rude and confrontational" when all you are basing that on are some mailing list posts, primarily in a mailing list which is renowned for that behaviour.

Personally, as a Debian Developer, I try to assist people and fix bugs in my packages, as my way of contributing back to a phenomenal set of software. Debian has over 5000 packages in the distribution, and while those are mostly not written by DDs, they are packaged, and made to play together nicely and install, upgrade and uninstall cleanly, and the whole damn thing just works.

Yes, of course Debian Developers are principled people who care passionately about things other than software, and if you stick a thousand of them in a mailing list together there is bound to be fire! And hell, some of us are geeks without social skill. Cry me a river.

Re:Debians own social standards???

[DAldredge]

You imply he is wrong in the first half of your sentence, then in the second have say he was right.

Every thought of running a presidental campaign?

Re:Woody's Age

[Wyzard]

Then again, it begs the question, is Debian really a desktop OS? Debian developers seem to argue that one a lot.

Stable isn't, but with a little care and feeding, Sid is. And the care and feeding is educational -- just doing an upgrade every few days with apt-listchanges installed is a great way to learn by osmosis little tidbits about the system's workings.

Debian's Identity Crisis

[Elf-friend]

It seems to me that Debian can't decide what it is these days. Half the time, the just want to be the most secure and stable Linux distro out there. The other half of the time, all that matters is the licensing aspect.

I really wish they would make up their minds. Are they trying to provide the most stable distro out there, or are they trying to be an unofficial organ of the FSF? Both perhaps? If this last is the case, then they ought to have been more balanced about this decision. Why push back the release cycle by a whole year just to make the GNU zealots happy? Why not wait until the next release for these change and bring Sarge out on time? At least the message there would have been that "we agree with FSF/GNU ideals in principal, but we have other goals which are as important as far as this release goes."

Instead, the message they are sending is that "Debian is for GNU zealots only. We don't give a damn about anyone else. If you have a need for any closed-source program or proprietary hardware, you are evil." I am sick of this attitude, frankly.

Don't get me wrong, I respect what RMS and FSF/GNU have done for the cause of free/OSS software, but I simply can't agree with the notion that closed-source is evil. I prefer Linus' approach which is essentially to say that we think free/OSS is a better idea, but that authors have a right to go closed-source if they want. Personally, I tend to think that the BSD license is often, maybe even generally, superior to the GPL. I use Linux because it ofers more choices than BSD, not because I dislike BSD or its license. I had thought that Debian was distancing itself from GNU, but I guess they've done a 180.

I have used Debian for over three years, because I like the package system. I am not a GNU zealot. Over the last two years, I have become increasingly annoyed with holdups in the release cycle, but promises of a quick Sarge release went a long way to apease me. This is the last straw. There are other distros (Gentoo for one) with as good or better package/ports systems, and that at least pretend to care about real-world users. Goodbye Debian.

P.S. Before anyone flames me, keep in mind that in part I am blowing off steam out of utter frustration. If I spoke overly harshly, I apologize to anyone I offended.

Re:Debian's Identity Crisis

[twilight30]

I hear what you're saying, I've been more than a little annoyed with Debian as well lately. But...

I don't use the Stable release, outside of sysadmins, I don't think I know anyone else who does.

They have an obligation to release free stuff that has only become stricter in the last few years. What they are saying about proprietary hardware is not, 'Fuck you, you are evil,' but rather, 'We're sorry, but we have no control nor proper information over the hardware you're using and can't really help you, because if we did have that information, we would be constrained by the makers of your hardware not to divulge it to you.'

To take a completely trivial example, the computer I'm typing this on has an NVidia video card, and I run Debian testing on it. I knew I'd have a hard time using Debian-only software on it, so off I went to Google. Within an hour I had everything downloaded and installed.

OK, perhaps not as easy as grabbing Fedora. But I know as a result of this what I have, what Debian can provide, and what they can't. In other words, I've learned something, and I didn't have it done for me.

A lot of times people forget or get confused as to what the Debian project is about. But they tell you this upfront: they aim to make a free computer operating system. Not to make the easiest, or most convenient -- though it would be nice -- or to make the most secure or most complete. Simply to make it free, so that when a new user, a business, an organisation or a government picks it up, they won't have to --ever-- worry about licensing costs or any other shit.

Limiting the GNU stuff to exclude FDL documentation is, I agree, a total pain in the ass. Granted, disk space is no problem for users these days. But what would be the alternative be? Gloss over the very real difference in opinion as to the modification of texts as just a side issue? It creates problems because the GNU project says 'You have to accept this part and parcel of the original package even if you don't agree with any of it nor does it have any real use'.

And no, this is not a flame, it's an honest question -- if this is really a problem, why aren't you using a FreeBSD system instead? They have ports. They have lots of free stuff. They have a large userbase, they have lots of online support and are pragmatic. What's the holdup?

For fuck's sake...

[theantix]

That's it... I'm giving up on Debian. I know they mean well but some users just want a stable system that has had application updates since 1994. I agree with the ideology of their actions, I think the unfree documenation should be removed from the project. But that should be a project goal for the next release, because we were nearly ready for one in the coming months.

It's sad, because the idea of a community driven project is noble, and I hate to see it fail. But this is failure -- they have abandoned their release goals and further postponed an already rediculously overdue rlease. They just aren't serious about maintaining a stable release, and thus I'm not going to take them seriously.

Not that they owe me anything -- I appreciate all the hard work that the Debian Developers do. But this is just the last straw...

In other news...

[luna1ix]

Debian has decided to change the codename of the next release to GNU/Longhorn.

too far

[tacocat]

I was reading on this email post and it's really discouraging.

many pieces of hardware will not be supported by the Debian system itself

The single biggest problem that I've seen in getting people to adopt Linux, and Debian in particular, is the function of hardware connected to the computer. If my widget doesn't work with Linux then what's the point of using Linux?

I've grown accustomed to the practice of due diligence on researching any hardware support for a product X before I buy it. But if I have to start doing this, and then perform another search just for Debian, it's making Debian very unattractive.

I am really doubtful that this is a smart move on their part. I am a HUGE fan of Debian and very supportive of their work. But the implications of this are not good from where I sit. Their ideologies are making their product non-useful to the community which they attempt to serve.

They are creating an overly complex architecture at a time when Linux does not have the support necessary from the commercial entities controlling the market (hardware and software). This will tend to isolate Debian from the rest of the Linux community and may give them the label of "Oh... Those guys over there in the orange sheets."

I hope I'm wrong, but I think Debian really screwed this up in a big way. The fact that they have just incurred an entire year of delays to their release cycle at a time when they were months away and years behind the rest really doesn't help them in the least.

I really don't understand their motives with this one.

Re:Debian is wrong about documentation

[boots@work]

Kicking out even FSF's documentation, on the grounds that is doesn't meet Debian's criterion of freeness, was a really stupid thing to do.

OK, with you so far...

Writing software is its own reward for a lot of people. Writing docs is a vital chore which nobody likes and which gets little recognition.

Sorry, but this is completely bogus. We went through this argument a few years ago: "people won't write nice GUIs for fun", "people won't write open-source databases", "people won't release open source kernels for big iron". Now it may well be that there is not enough good open documentation on some particular topics at the moment, but that is no reason to assume that it is not worth trying to get free documentation.

The FSF docs are free enough for all practical purposes.

They are free enough for many purposes, but the licence is buggy and needs to be fixed. Some projects such as Arch have gone back to the GPL for documentation, which is the sensible choice at the moment. Both FSF and Debian need to stop fucking around.

More Free than GNU?

[Nutcase]

I always thought Debian was essentially the linux implementation of the FSF/GNU ideals. Most other distros make compromises for usability, but debian never compromises on freedom. This is just the latest example of that. And more power too them for it.

The amazing thing here is this: In reaffiming their commitment to freedom, they are finding that they have to exclude some GNU documentation because it is considered non-free. In other words, Debian now seems to value free software more than the Free Software Foundation.

Thats disappointing, but at least Debian is sticking to their ideals without compromise. Too bad the FSF can't say the same.

stable=woody, sarge=testing, sid=unstable

[kungfujew]

Just thought I'd clear the confusion here... I saw some of you guys refer to sarge as "unstable" and sid as "testing".....this is not the case. the correct names are: stable=woody- for production servers (current stable release) sarge=testing - this you run on your home machine sid=unstable -you dont want to run this because it breaks almost every day trust me, backing out of a dist upgrade to unstable is a painfull and involved process.

A brief review of history.

[Daniel]

Here's some context for people who don't follow Debian matters habitually.

Debian has a document called the DFSG, or Debian Free Software Guidelines. These guidelines are used to determine whether software included in Debian is free: they require that the software be freely distributable, freely modifiable, etc. Stuff that doesn't meet these guidelines doesn't go on the CD images and is segregated into the "non-free" section of the archive; this policy is enshrined in Debian's Social Contract. More contextual information on the DFSG and its application is available here.

Now, historically, these guidelines have been applied to everything distributed by Debian. For instance, the Doom shareware .WAD went into non-free because its license forbade modification. However, some controversy has arisen in the last few years due to two developments: first, the FSF started using a new license (the "GNU Free Documentation License") for its documentation; more recently, there has been a trend for hardware manufacturers to require drivers to upload binary firmware code upon initialization.

Despite its name, the "GNU Free Documentation License" turned out not to meet the DFSG (you can read some unofficial explanations [URL redacted because I believe the author wishes to keep it private for the time being; I will post it later if he tells me it's ok; I'll badly summarize it by saying that Invariant Sections are the major issue but not the only problem]). Because this license was applied to documentation of large packages, such as libc and Emacs, because it claimed to be "Free", and because it was published by the FSF, some people felt that Debian should find a way to distribute software under this license in "main" even though it was clearly non-free according to the DFSG. The typical argument advanced to support this position was that "documentation is not software, so it doesn't need to meet the DFSG". This argument relied on an ambiguity in the meaning of the word "software": it can mean either "anything that's not hardware", or "sequences of instructions to be executed on the host microprocessor".

The firmware issue is somewhat different; there were some recent arguments on the debian-devel mailing list over whether binary firmware that is uploaded by an otherwise free driver should be moved to non-free. I haven't followed this as closely, and it only came up in the last month or two. (well, it has been discussed in the past, but the first serious discussion I'm aware of is in the last month or two)

The amendment that was recently passed changes the text of the Social Contract to make it clear that everything in the Debian archives (not just executable programs) should meet the DFSG. This was intended to settle the GFDL question once and for all.

The message referenced by this /. article is a post from the Release Manager indicating that he is changing his policy as a result of the GR. Until now, certain things that were unambiguously non-free, but where it was felt that the non-freeness was either not a regression (ie, they were non-free before and we didn't realize it and distributed them anyway), or where it would cause significant disruption to force the non-free item out of main (for instance, binary firmware), were being allowed to remain in Debian main until the release of sarge. Assuming that this message was sent in good faith, Anthony is indicating that he honestly believes that this was not previously a pragmatic exception to the Social Contract, and that no such pragmatic exception is possible now. Thus, he is now holding up the release until all this non-free stuff gets removed from main.

Discussion is ongoing on several Debian lists, and I don't think it's appropriate to make assumptions about the final outcome until things have settled down again.

Daniel

Re:A brief review of history.

[Daniel]

Just a note: a reference on the GFDL that has been publically posted is available here.

Daniel