Submitted by “We are pleased to announce the immediate availability of CentOS-6.0 for i386 and x86_64 Architectures. CentOS-6.0 is based on the upstream release EL 6.0 and includes packages from all variants. All upstream repositories have been combined into one, to make it easier for end users to work with. There are some important changes to this release compared with the previous versions of CentOS and we highly recommend reading this announcement along with the Release Notes.”
8 months delay definitely gives a new meaning to “immediate availability”. Sheesh, “Duke Nukem Forever” shipped before CentOS 6 (sorry, just couldn’t resist the temptation to say it).
Congratz to the release team. You finally got it out the door. I hope they will consider fixing their community interaction problems. Someone on the SL site mentioned a new project to create a 100% compatible RHEL clone in an open community model. I hope CentOS will contribute to it.
That would be nice. SL could be a nice basis. Delete the SL specific patches, and there’s the code base.
I considered switching to SL, but it diverges just enough to make me uncomfortable when switching back to RHEL. With CentOS, it has bug for bug compatibility.
SL is not and is not aiming at being compatible with RHEL. They don’t care about ABI compatibility if what I have heard is right.
But yea, CentOS have double failed in 1 year on community relation.
well, the delay shows that community driven enterprise versions should be something so stay away from. And yes we have “the other” RHEL clone as well. But then again, how large is that community?
Thanks but no thanks. RHEL or SLES (and my vote goes for the latter — better support; better tooling, less anoying bugs)
Well the problem is that it isn’t community driven. Its like 3 guys doing it whenever they feel like it. And they will tell you if you don’t like it, tough crap. Fedora is community driven and they manage to put out a distro every 6 months. True they get a lot of support from Red Hat, but still. CentOS is used widely by ISP’s while SL is used around the world. There are resources to make the whole process much faster and smoother if enlisted. The new project (wish I could remember the name of it) aims for 100% binary compatibility but with complete transparency in the process. Transparency is 90% of the complaints against the CentOS project.
Personally a lot of this would just go away if Red Hat started a license program for non-profits/academia and open source projects. Maybe its not worth their effort, but a lot of effort is spent creating clones so that must mean there is a need. Maybe something done like the Fedora project. A community based version of their enterprise OS could be a good thing for them.
I believe it is called Ascendos.
http://www.ascendos.org/
In my opinion, this is not likely to happen because RH needs the community clones to exist, but it does not need them to be especially good.
They need to exist because without them (and with the official distro being entirely behind the paywall) not enough people would keep up the skills and “mind share” needed to keep the RHEL ecosystem healthy.
They don’t need to be that great because it’s a good sales tactic – “Call us when you’re serious enough to pay for the real thing, with a real roadmap”
The real question, for me anyway, is did the current situation turn off enough people (ie, turn them on to Ubuntu or whatever) that the ecosystem was damaged? In other words, from a selling-RHEL perspective, is the good-but-not-that-good balance right? I don’t know. I think it could be better. In the short term, the switchers are probably not the people they care about (the paying types) but in the long term, mind share might have gone down.
What I’d like to see is RH just making patches without any support free. But apparently their market analysis tells them no–otherwise they’d be doing it.
“Real roadmap”
RHEL6 itself was 2 years late
What kind of salesman lets mere facts get in the way?
Quite. There’re two insightful threads on @centos-devel from February and March. If you have the patience to go through all messages, you’ll see things like
On documenting undocumented build-time dependencies (“we don’t use any super magic juice anywhere”):
And the main problem with CentOS as I see it:
By 2011-03-10 CentOS 6 was supposed to be 30 packages away from ready. It took them another four full months to get it out the door. The devteam seems to have some very serious communication&community issues to handle.
Let’s be honest here. Most serious, as in real world day to day business relying on it, users of an Enterprise OS aren’t going to really care about jumping on a new major version the moment it’s released or even a month or three late for that matter. No sane admin is frothing at the mouth to toss an OS they haven’t personally tested with their setup upon release when the existing OS is working well and still supported.
What hurt CentOS wasn’t so much the slowness of release, but the relative lack of communication about tangible progress/holdups and community rumblings around the reasons for the slow release.
Edited 2011-07-12 08:46 UTC
While waiting for the CentOS releas I setup SF Linux (another Red Hat clone) on a test machine. It turned out to be terribly buggy. If that reflected the state of Red Hat at that time or if it was something the SF Linux people had managed to add I don’t know, but it certainly wasn’t production ready.
In the mean time my systems have rlun CentOS 5.x without any problems. If the CentOS guys now have something that actually works, I will congratulate them and say that it was time well spent. If not I will happily run 5.x a bit longer as long as I get bug fixes. To me upgrading isn’t a that big deal. The software I need run just fine on 5.x.
However, I think they should have given us more information on the state of things. I realize that it takes time to provide such information, and that time may have to be taken from the time available to get the release ready, but if you don’t provide it everybody is going to ask how far you are from a finished distro, and that will consume time too.