Slashdot Mirror
Debian Core Consortium Releases First Code
daria42 writes "It looks like the Debian Common Core Alliance announced a while ago is going to make good on its promises: the project has released its first code this week. The release consists of a base installation of Debian 3.1 with the Linux Standard Base and security updates attached. But the project also looks like it has attracted some criticism from within the Debian developer community - with a spoof Web site having already been set up to poke fun at the Alliance."
Screw the real site, the spoof is what's important: http://www.dccalliance.biz.nyud.net:8090/
...already. Looks like the spoof web site is down.
Bradley Holt
must of taken ages to think that one up
The story includes a link to the spoof website but not to the actual one. Great reporting.
The address is http://www.dccalliance.org/ btw.
There should be no problem with this as long as they're following the proper licensing for all the code they distribute.
It won't matter anyway to the Debian groups.
Everyone wants a Tux in their life.
Just what we need: some more kids (or grown-ups acting like kids) fighting among themselves. This is all we need to project that trustworthy Linux and open-source image.
Now UserLinux has died a painful death of disinterest, Bruce Perens has a new toy which he can use to avoid becoming next years recipient of the "Eric S Raymond Memorial Prize for Open Source Irrelevance."
Altogether now:
Will the real Bruce Perens please shut up,
Please shut up
Please shut up
Bah! I tried ".nyud.net:8090" for the first time and thought I'd be cool for mentioning it here first!
Anyway, it was my first real test of it. Pretty cool. Even if it didn't work very well.
let's face it if more Linux Distros worked the same way and had the same layout, plus if all lib,Sources were the same that would help out a lot.
CH
there is a greasemonkey script which shows Slashdot index page with links as having coral links, mirrirdot links and google cache links next to every link!
The real benefit of the alliance that I see is that .deb packages should be compatible across multiple distros. Unfortunately, Ubuntu is not part of the alliance, and there are a lot of 3rd-party Ubuntu .debs out there.
What is the actual difference between "Debian Common Core" and "Debian" (since Dv3.1)? Is DCC just an organization that certifies that (its own) Debian-based distros are actually both Debian-based, and comply with "Linux Standards Base" specs? Does Debian v3.1 itself not pass that test?
--
make install -not war
Somewhat on topic is the issue of fragmenting. For a while, if an application or OS didn't do something you like, the common response was:
- Dont like it? Fork it! - Dont like it? Roll your own!
Problem is that it leads to a lot of confusion and fragmentation within the community that confuses the hell out of outsiders.
I think consolidation is a good thing and folks should work together more often rather then just splintering a code base.
(Note, fragmentation CAN be a good thing in the cases like Security Knoppix or RTLinux)
http://www.dccalliance.org.nyud.net:8090/
Seems to work when you get the correct URL. Fancy.
"Hello world, we released an open source operating system so that all may benefit from our efforts and... Oh noes! People are modifying it to suit their needs! Evil! Strike them down!"
(slightly edited)
o uille.htm
http://www.splorp.com/critique/
Spirals all come from Ubu Roi's dread Gidouille :
http://expositions.bnf.fr/utopie/pistes/grand/gid
aurelien
This seems very reasonable to me. There's something I'm missing -- Why the resistance and the spoof site?
Conflict often brings about the biggest changes, and conflict between OS developers is nothing new.
Take OpenBSD. Had it not been for Theo quarreling with the NetBSD elite, then we would not have the ultrasecure system that we have today.
And of course there's the revolutionary DragonflyBSD. If Matt had not been ostracized by the FreeBSD team, then we wouldn't have what will most likely become the premiere workstation BSD in the near future.
Then there's the whole CTSS/ITS/Multics debacle of yesteryear.
While not an operating system in itself, the whole XFree86/Xorg licensing incident has proved to be one of the greatest influences on UNIX GUI development in the past 20 years.
I believe that conflict is essential for open source projects. For if it were not for conflict, we would not have such great products as OpenBSD, DragonflyBSD, and Xorg. I, for one, support this sort of conflict. It often leads to increased productivity in the long run.
Cyric Zndovzny at your service.
I was expecting a "spoof site poking fun" to be, you know, funny.
Everybody loves Eric Raymond:
http://geekz.co.uk/lovesraymond
then they ridicule you,
then they fight you,
then you win.
-- Mahatma Gandhi
"Computers are useless. They can only give you answers."
-- Pablo Picasso
Sorry to be harsh, but when I started using Debian 3 years back, I wasn't treated well as a 'n00b' even though I had 2 yrs prior Slackware experience, and just felt like the entire project was too splintered. I mean, running on multiple archs is cool and all, but if it pulls down the medium range then what's been gained? The plus of this approach is it was ripe for someone to come along, take what's good (APT-GET!) and create something specialized, which is now Ubuntu Linux. Building on the Debian base was just their beginnning, but it was an ace move.
bad_outlook
--
Is this vague enough for you?
What else do you need to know?
Best Slashdot Co
Frankly, I think it's excellent for such spoof sites to be created. And hopefully they do incite much hatred, angst and conflict. I want the DCC people to have the urge to, as certain Microsoft executives might say, "fucking destroy" Debian. I want them to have that urge so badly that they put out a far superior product.
Cyric Zndovzny at your service.
Always poking fun at the Alliance. Why is it that I always find myself drinking in an alliance friendly bar on Unification Day?
You like your new Mac more than you like me, don't you, Dave? Dave? I asked...She said Yes.
It appears that all the major Debian derived distros have joined this group (Mepis, Xandros, Linspire, Knoppix, Linspire, etc.) but not Ubuntu. Why not?
But you wasted it, your moment has come and gone now to exist only as a series of 'what ifs' for the rest of your life.
/. post than the fact that they messed it up?
Its sad, really...
WTF!! Isn't it more sad that anyone would actually worry about messing up a
the spoof site at http://www.dccalliance.biz.nyud.net:8090/faq.html is pretty slow too, here is my analysis (and a copy of their 'faq':
So somebody is upset about basing the name of a separate organisation on 'Debian' and abbreviate that to a 'D'. Well wanker, I tell you something: you cannot trademark a single letter, or we'd only have about 36 possible companies.
(this seems a rip-off from the 'real' DCC faq entry. see above, no trademarks on single letters.
Aha, a somewhat real-ish bone to pick. Except that creating a patched kernel is not such a big deal. You can find several in testing, does that mean that testing has been forked with every new kernel release? As long as the new kernel is interoperable with the one it replaces you can hardly call that forking.
is DCC necesary?
Debian has grown into a big organisation, and thus also has it's share of people with 'uncommon personalities'. It is all a volunteer effort (and thus?) some people in debian react a little allergic to commerce baseed on Debian, even though the licence allows it. Commercial Debian-based distro's have a vested interest in Debian, so they seek some influence. It can be vey hard to have to argue with every maintainer whose package they have altered to get him to accept the changes(There are 1000's of developers and and at least ten times more packages in Debian). Even with proper conflict resolution it quickly becomes a nightmare, so a lot of distro makers don't feed their changes upstream/to Debian at all.
That is a problem and something that a separate repository can solve. Yes that is in effect a fork, in the same sense that Ubuntu or Knoppix is a fork (not for the silly reason above). If the Debian derived distromakers have their own repository where they can work together changing Debian to their common goals without getting bogged down in Debian rules/games, then that is just great, IMHO.
It is great for the Debian-derived-distro-makers(DDDM?), as it allows them to cooperate and improve Debian while they are at it. It is great because it avoids conflict and bottlenecks. Commercials distro's (can) have a different interest than induvidual Debain developers. With this construction no single Debain developer can obstruct a DDDM. It is great because It will concentrate all enhancements made by DDDM's into one place, so the Debian developers don't need to track all different DDDMs for changes to their packages. And most of all, it will concentrate efforts into coding and cooperating, and that is good for all.
This space is intentionally staring blankly at you
Debian/kFreeBSD has its web site's pages copyrighted by SPI, web pages which mention that "Debian" is a registered trademark without mentioning the status of "FreeBSD".
But the people I blame are the directors of the FreeBSD Foundation which now owns the FreeBSD trademark at least as far as it applies to "CD ROMs featuring an archive of computer programs which may be accessed for use archived on a CDROM." (And it appears the FreeBSD Foundation is working to expand the applicability of the FreeBSD trademark.) But there is already a Debian/kFreeBSD iso.
Considering that a simple cease and desist was sufficient to force CentOS to scrub references on its web site to the phrase "Red Hat" and other such trademarks (other than apparently a link to someone else's article), I am baffled what either Debian/kFreeBSD or the FreeBSD Foundation is waiting for.
The package repository is indeed important. I think Debian's strict packaging guidelines and quality control have made first-class repository possible.
For several months I have tried using RHEL4 system and it has been quite frustrating. APT is there and RPM is actually quite good low-level package manager. However there are no software packages! It seems like nobody wants to build packages for RHEL. Several important things are missing: Totem video player, Evince document viewer, Gtkmm devel libs, Epiphany, AucTeX and Inkscape to name a few. As soon as I have a chance I will convert the box to Debian or BSD.
I don't know how futile resistance will be to a guy who put Linspire *twice* in a list of what would've been 5 distros... He must really love it, and that's scary. That's the kind of guy who makes a simple story into a story of mythical proportions.
Yes they try to avoid the word fork, probably for the connotations it has ('forks ar *bad*'). But there is nothing wrong with a fork an-sich. The size of the Debian project is what made it great(lots of packages, lots of testing, lots of development), but the size of Debian also makes it hard for commercial enities to cooperate with them directly.
Usually forks are considered 'bad' because of the duplication of effort. This fork is good because it prevents the structure of the Debian organisation to slow the efforts of the DDDMs.
This space is intentionally staring blankly at you
Since Debian has been my Linux desktop choice for over 3 years, I don't really feel comfortable using a candy-cane Linux distro (Linspire, Mandrake, Suse, Mandrake, Lindows, Linspire, etc.).
I higlighted and middle pasted that list, not checking it for grammatical errors. I'm sorry if I hurt your feelings. I'll pay more attention next time. Ubuntu ubove ull.
Everything else is the fork including what is called "Debian" now.
I mean, running on multiple archs is cool and all, but if it pulls down the medium range then what's been gained?
Exactly. Throwing out features to appeal to the lowest common denominator in the name of portability is great for the those fringe architectures that hardly anyone uses, but lowers the standard for the vast majority of users (i386 etc).
What's wrong with going on a feature by feature basis?
eg
"
Feature X: Supported platforms:
i386, PPC, AMD64
Unsupported platforms:
the rest
"
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
Ubuntu has shown us the way. Time-based releases every six months, supported for 18 months. Then, every fourth release or so (about every 2 years), "bless" that release as enterprise-ready by giving it 5 years of suppport instead of 18 months.
Genius. Avoids the problem of split distributions like Fedora/RHEL, where only one is really suitable for enterprise use and only one is really suitable for enthusiast or home use - and they're not the same. That model works great when you're trying to sell a refined RHEL or NLD to corporations, but with a community distro like Debian, there's no need for such a split. Don't change the distro, keep pumping out time-based releases and change the support angle - want to stay more bleeding edge, keep upgrading every 6 months; want to run a production server, install a 5-year-supported release and relax.
Time-based releases have proven themselves again and again, while "when it's ready" releases have failed again and again. Time-based releases are a perfect fit for Debian, as a (perhaps overly) democratic community project. If wrangling a release out of wild developers is too hard for human beings, make the CALENDAR the boss. It works.
Wake up, Debian! I know you hate Ubuntu because you guys hate... well, everything. (Really, never saw such an ornery group of guys before - talk about needed to get laid.) But they've designed an excellent release process, perfectly designed for the needs of both enterprise and hobbyist users without 'ghetto-izing' any group. Why not swallow your pride and adopt it?
Disagree, don't think Ubuntu has the answer? Well, then come up with something else. The Debian release process is BROKEN. It doesn't work. People who are skeptical of the DCCA are skeptical because it relies upon your release process. Pick something, hopefully something that has worked for others, and let's get moving already, you god-damned ornery slackers.
As a server administrator, I love Debian stable. The fact that new releases (and thus new features) are years apart is a great advantage: everything that is working stays working. The Debian project works very hard to keep it that way; they backport security fixes so that you don't have to upgrade to a later version of a package, and when the new release comes, they make every imaginable effort that your upgrade will be seamless.
It's good that one distro caters to the needs of people who believe in "if it ain't broken, don't fix it". This is the real beauty of having so many distros: a few of them to fill every niche, so that all niches are filled and there are some alternatives within each niche.
There are advantages to schedule-based releases, and you do a good job at pointing them out. However, it's a fundamental truth in software development that you can't decide in advance when something is going to be ready. The policy for Debian stable is to make every possible effort to ensure that everything is working before releasing it. This is the only way to ensure a smooth-working release, and, to me, outweighs the inconvenience of not knowing when the release will come.
The beauty of Debian is that the project is so immense that it manages to cater both to extremely conservative types (me when administrating servers) and to bleeding edge types (me when playing with my PC at home), and do a great job at both. If you want assurance that everything works more than anything else, use stable. If you want the latest and greatest, use unstable. If you want a smoothly polished desktop distro, use Ubuntu (you could use other desktop distros, but they usually can't match the convenience of Debian's wealth of packages). If you want an embedded or otherwise really small system, Debian is probably too heavy; but note that busybox provides a dpkg command...
Please correct me if I got my facts wrong.