Roland Mas wrote:
*Supposing* I were agreeing with you on the existence of a problem, I would probably be of the opinion classified as case 3 above. The reason I could identify for the problem would be that people prefer bitching and complaining about testing being late and stable being old, rather than helping fixing bugs. I agree with the goals. I believe the testing process is appropriate. The problem is not that this process requires software to be tested and declared relatively bug-free before they are admitted into testing. The problem is that the software is not even remotely bug-free.
Apparently the "software" is "bug-free" enough, say, in the case of KDE 3 or Gnome 2.2, for the vast majority of other Linux distributions. I'm not saying there aren't a lot of bugs in those packages, but haven't seen any real show stoppers on SuSE 8.2 or Red Hat 9. The point is that this "sofware" we're talking about is "good enough" for the vast majority of Linux users. (And I'm saying that because in most "What distro do you run" polls, I see the troika of RH, SuSE, and Mandrake pulling the bulk of the votes.)
Is the build system in Debian "better?" I personally *do* think so. I started this thread so that I could understand where the project was headed, not just from a theoretical point of view (which I can get from the docs on debian.org), but from an in-the-trenches point of view. It turns out that it's the same view, but I didn't know that going in. (It was only after pestering the Red Hat list that I found out that their "consumer" versions are going to be more like rolling betas going forward, so I just wanted to poke this list and see what happened.)
The problem with the build process -- to me -- at this point -- is that by being "correct" from a developer's point of view, you've made the system only acceptable for "developers." Now, I'm as "developer" as the next guy. I program for a living. It just happens to be in Visual Studio 6 and .NET, for the most part. But I do LAMP development as well, and I program in every system I'm on. (You know the saying, "Writers write?"...) But there comes a point that I don't want to have to "develop" the system upon which I'm trying to develop. I have several LAMP site migrations and updates to do, and I'm sitting here debating the finer points of Debian's process, NOT because I want to start an argument, and NOT because I want to hack on my system to get it to do all the basic things I *need* it to do, but because I want to know if this is going to allow me to eventually settle in and do the things I really want to do without getting in my way, and yet keep up reasonably with the changing featureset of GNU/Linux at large. If that's NOT the case, then I ought to shut up and go run something else.
> In case 1, I suggest you bring up a discussion on debian-project, so > that the goals for the next release and the more general goals of > releases (in terms of what it means to release a new stable > distribution) can be discussed to death, and we eventually come up > with maybe a different set of targets for the release process.I can't find this on the mailing list listing page. I was sure that there was *some* group dedicated to talking about the project as a project, but I'm missing it now. I'd like to join it and get a feel for it.
In the past few days of living in woody, I've already seen that I can indeed run the NVidia driver and get my video card working under XFree86 4.1, which I didn't think was possible. KDE 2.2 looks an awful lot like 3.x with the addition of the "crystal" icon theme. The only thing I lack at this point is the ability to backup my Tungsten, and I just downloaded the 2.4.20 patch to make that work. I haven't actually built the kernel yet, but I think it will be fairly straightforward given the surprisingly good documentation surrounding this distro. What I'm trying to say by this is that, despite how "old" the software is, I'm not finding a lot of things that Debian isn't able to actually *do*. I may not have the gorgeous OpenGL screensavers in xscreensaver-4.09, but on the other hand, "stable" is proving just that. Everything works as expected.
In short, I think this is going to work for *me*, but I'm still trying to get a look into a crystal ball about the future prospects for Debian.
Regards, dk