On Sat, Feb 10, 2007 at 04:56:09PM +0100, neologix@free.fr wrote: > > Ok. So let's fix a deadline for the freeze then. There was a deadline set for the freeze. The freeze actually took place shortly after that deadline (by two weeks, I think). > And if you want a more specific example: I think Etch is going to ship kernel > 2.6.18, or 2.6.19. OK. Well that is for the kernel team, installer team and release managers to ultimately decide. > The past few kernels have undergone a large amount of new features, and > intrusive patches. They're simply buggy. Last version 2.6.20 as been described > by Linus as a stabilization release: this one is a good one to ship with. By the Umm, so the solution is to release with a kernel that has received little or no testing? Wow. BTW, you do release that there is a tremendous amount of work for the kernel team and installer team whenever the default kernel changes, right? > way, why ain't you supporting Adrian Bunk 2.6.16.y branch, whose goal is exactly > stability and security? > That's great. Of course, there are a number of things to consider. For example, whichever kernel is released will be supported by the kernel team and security team for the life of Etch plus 1 year (or until the release of Lenny). They may have concerns which need to be considered in deciding which kernel will ship. > > Yeah, I was there. I was happy to see that at least one Linux distribution cares > about quality, and not the last bleeding-edge kernel/application. > Except that whole reason that the release continued with delay after delay was the people kept trying to get their pet project done or upgraded. The only two things that absolutely positively needed to be updated were probably the kernel (to 2.6) and X (to 4.3). Of course, since there was no hard target lots of other things "had" to go in. Like repeated updates to gcc and other toolchain utilities. There were lots of others. > > Well, it's your choice. I'm sure you would love the way Ubuntu works then. > In fact I don't. I am just saying that a large groupd of people, lacking a clearly stated obective or hard target will meander about aimlessly. Your proposed "stable kernel + stable toolchain" will not do it. There really needs to be a timeline. Since people will always debate stability and people will always push for inclusion of the next release of foo. Now, had the release managers pushed for freezing and then releasing *regardless* of stability just to meet the deadline, then you might have a point. However, they decided to delay a bit to get the RC bug count down and allow a few new things in, and then the freeze happened. Had the date target not been there in the first place, I am confident that Etch would still not be frozen. > > Really? > Welcome page of dunc tank: > > "The Dunc-Tank is an experiment to see how targetted fund raising can improve > Debian. As our first (and maybe only) project, we're trying to help the release > of etch happen on time." > > "We're trying to help the release of Debian happen on time". Damn. > > This is my last mail. Just do it the way you like. > Well, on kidding. Where on their page does it say that want to sacrifice stability in favor of timliness. It doesn't. Their whole objective was to accelerate the release process so that it would happen "on time" *without* sacrificing the desired stability. Regards, -Roberto -- Roberto C. Sanchez http://people.connexer.com/~roberto http://www.connexer.com
Attachment:
signature.asc
Description: Digital signature