Shuttleworth Calls For Coordinated Release Cycles 238
voodoosws points out on Mark Shuttleworth's blog Shuttleworth's call for synchronized publication of Linux distributions, excerpting: "There's one thing that could convince me to change the date of the next Ubuntu LTS: the opportunity to collaborate with the other, large distributions on a coordinated major / minor release cycle. If two out of three of Red Hat (RHEL), Novell (SLES) and Debian are willing to agree in advance on a date to the nearest month, and thereby on a combination of kernel, compiler toolchain, GNOME/KDE, X and OpenOffice versions, and agree to a six-month and 2-3 year long term cycle, then I would happily realign Ubuntu's short and long-term cycles around that. I think the benefits of this sort of alignment to users, upstreams and the distributions themselves would be enormous. I'll write more about this idea in due course, for now let's just call it my dream of true free software syncronicity."
Anhy reasons not to? (Score:2, Interesting)
Effect on testing? (Score:2, Interesting)
Matching toolchains, binary compatible apps (Score:3, Interesting)
Re:err Gentoo? (Score:4, Interesting)
Users often want to have access to the latest software. Many distributions provide updated packages over time, so that when a new official release comes out, many users already have a good portion of the changes. Gentoo takes this to the extreme, having eliminated the concept of release entirely (except for the installation system). So how does a synchronized release schedule help anyone when users will be upgrading various packages when they are updated?
I just don't see the value in synchronizing releases.
Re:Translation? (Score:5, Interesting)
The other half of the idea is that it provides a more uniform platform for software/hardware support -- the idea being that Adobe or NVidia could make a release that works with all the March 2019 distros, and support people could offer support for the March 2019 distros. It is, after all, one thing to be familiar with the package management of 5 different distros, and another thing entirely to be familiar with the specific point release quirks of the software packages on 5 different distros.
Of course reality would be nowhere near as elegant as this theory.
Re:Effect on testing? (Score:4, Interesting)
Anyway, the idea is to have MORE testers, since they would all be at one big bazaar instead of each at their own little one.
More patches would work across distros, and more time could be spent on other things.
The has its problems though. The agreement would force innovation to make people go to one distro over another, but that would only last a cycle or so as the other distros implement the idea (Since they all share in the end).
Maybe if Ubuntu decided to go desktop only, And Suse and Red Hat agreed to split up the server side, maybe, but that is unlikely. Debian will not not do it, they are the stable granddaddy, which can be a good thing.
I like having news on new distros throughout the year also, so it keeps people interested.
It's an interesting concept that might free up some coding time by solving a lot of problems faster and with less overlap, but I don't think it is likely to happen.
One of the strengths of Linux is its diversity (plus we geeks all like our pet projects). IT may help newcomers though, who are scared enough by the number of Vista versions, and are terrified of picking a distro, not to mention learning a new OS.
One explanation (Score:3, Interesting)
I'm surprised that nobody yet has mentioned the most obvious reason why Shuttleworth would want this.
Canonical shipped shipped their latest Ubuntu release, 8.04 ("Hardy Heron"), last month. Canonical wanted Hardy to include Firefox 3 instead of the getting-a-bit-long-in-the-tooth Firefox 2. However, since Mozilla's release target for FF3 was a month after the release target for Hardy, Ubuntu had three choices, none of them ideal: put off the move to FF3 for the next release (six months down the road), ship Hardy with a beta version of FF3, or delay shipping Hardy until FF3 had shipped. They ended up shipping with Firefox 3 Beta 5.
This matters because Hardy is a so-called "Long Term Support [ubuntu.com]" (LTS) release, which Ubuntu commits to supporting for eighteen months instead of the standard six (for customers who don't want to be updating their OS every six months). So now Canonical is on the hook to support a beta release of Firefox until long after FF3 final is released and the betas are forgotten.
I would presume that Shuttleworth wants coordination in distro release cycles because it would give him more leverage with third parties in situations like this. If Canonical says to Mozilla "hurry up and finish FF3 so we can make our release date", that's easy to ignore because it's just one distro. If EVERY distro is saying that to Mozilla, they'd be more inclined to listen -- and probably more likely to orient their own release cycles to fit with those of the distros.
Mod Parent Down (Score:2, Interesting)
1. It's a Trojan Horse to legitimize Shuttleworth's business prospects. Simultaneously he'll out-shout Debian and give him a platform with which to more easily compete with Novell. Given Novell's history of mismanagement, I'd say they are helping him already.
2. There will be _lots_ of ubuntu users ready to mod me down for comment #1. Let's entertain the notion that I am completely misguided and I fall in line with the Ubuntu groupthink. What exactly would be accomplished? Nothing. Ubuntu already forks from Debian early on and as far as I can tell, never the twain shall meet.
3. Debian still releases when ready. Ubuntu releases on a calendar, warts and all. Imagine if the Debian community fell under Shuttleworth's and changes to a firm "ready or not" release schedule. It would slowly turn the Debian project into Shuttleworth's fiefdom rotting it from the inside.
Re:Diversity is Stronger (Score:3, Interesting)
What I said was the frequency of opportunity to switch is important. Right now, if your distro doesn't do what you need, but another one issues a release that does do it, you can switch. Any one person might switch that way only once in 10 years. But overall, thousands of people probably switch that way every time a new distro is released with significant differences in version or inclusions from the last significant release of any distro.
Many more people probably do so at the even finer granularity of unofficial updates to existing installs, but that's not the same as getting an official release, with all its support, and even with support contracts (like with the distros Shuttleworth is talking about). Which is why we're talking about the actual release schedules, because unofficial updates are so frequent and self-organizing that they don't need to be synchronized. It's the value of the full, supported (and tested) releases that Shuttleworth and I are talking about.
Re:[CITATION NEEDED] (Score:2, Interesting)
If Suse or RH want to hold back on a new release or RH wants to jump up and be the beta tester for a new version then more power to them. I think that while Shuttleworth may be well intentioned, it would ultimately be a disservice to the linux community for multiple reasons. Security and reliability being the number one reason.
I really think Shuttleworth knows this and is making this statement more to get advertising for his distro than for anything else.
RH and Novell have large customers and are well acquainted with their needs. If in the process they find that certain packages cause problems, they may postpone a release. RH and Novell accepting a release cycle plan based on the communities "We just want to know" or "I have a dream", just isn't going to be a good business decision at this time. And I don't think it makes sense for the Linux Community as a whole either.
I actually stumbled across this last night on firehose or something. No comments had been posted yet and I thought it was a joke along the lines of something in the Onion.
Re:One explanation (Score:4, Interesting)
Re:Diversity is Stronger (Score:2, Interesting)
The Bazaar will survive. There'll still be Puppy, PCLinuxOS, Mandriva, MEPIS. Gentoo, Slack/Slax, CentOS, Arch, Vector. NetBSD, FreeBSD, OpenBSD, DesktopBSD, PC-BSD. OpenSolaris. gOS, Xandros, Freespire.
The Bazaar and the Cathedral is not a dichotomy. It's symbiotic. They not only can coexist, they must.
The Bazaar certainly will survive this because of the people behind it. If Shuttleworth's clan strays to far towards the Cathedral, the Bazaar's backers will throw their weight elsewhere.
But if Canonical never does anything to provide services to the Cathedral, only the Bazaar - no matter how vocal, always the minority - will ever benefit.
The people who only have time for the Cathedral's simplicity benefit from collaboration, and the people with time to sample everything in the Bazaar benefit from competition. It's long past time for the Bazaar and the Cathedral to work together and share those benefits with each other.