Hey Jason and the gang; Well if people are leaving because of "stuff" being broke, without putting forth an effort to help fix it: Then I predict said people will be leaving every other linux distro also. i.e. ALL the linux distros go through pains incorporating newer gcc's IMnsHO. <- Al least that is my experience with about 60 distros over the past 10-12 years or so. (YMMV)
Very best regards; Bob Finch P.S...AND archlinux has done a really good job about handling such instability issues so far. > I completely agree with jason on this and that was a fantastic reply > that actually helped explain some stuff for me that I didn't know as > well. Maybe a description like this would be helpful if made available > to the normal non mailing list user.As for users leaving due to such > issues is a completely false statement. If users have further questions > on policys and implementations most join onto the irc chat or hit the > forum for clarity before deciding to leave the distro. > > Maveric-i686- > > Jason Chu wrote: > >>On Wed, Jul 06, 2005 at 11:53:52PM -0700, sn0n wrote: >> >> >>>even moving the kernel from testing, dont stop someone from upgrading >>> their gcc, then getting stuck on nvidia (or any other modules >>>really?), since the kernel is still compiled with 3.x... maybe gcc >>> belongs in unstable... since everyone under the sun is just told to >>> uncomment testing anyways, since everyone wants the 'bleeding edge'.. >>> >>> >> >>I think that's silly. We have current, extra, and unstable for the >> 'bleeding edge'. Testing is for... wait for it... testing! You can't >> complain that something is broken in a repo that essentially means, >> "we're not quite sure if this is broken". >> >>I would also rather not make testing into a regular repo. That's what >> current, extra, and unstable are for. If we made testing just another >> repo, we'd need really-really-testing for the stuff we're actually >> truly testing. >> >>We can't really put gcc4 into unstable because unstable isn't meant to >> be a partial repo, it's meant to be a full one. Anything in unstable >> shouldn't be in any other repo and definitely shouldn't share the same >> name with anything in any of the other big three repos. The only way >> gcc could go in there is if it's called gcc4 and can be installed >> alongside gcc3. >> >> >> >>>just have testing self contained, and compile the kernel with the >>> current testing gcc.. its not that hard.. after all.. Arch uses PHP5 >>> by default.. look how well it works.. ;-) >>> >>> >> >>Yes, testing should be self contained. I don't know if you've been >> around long enough to remember, but we used to have an NPTL repo >> because all the NPTL stuff took too long in testing and it was no >> longer self contained. Because of the amount of work to set up a new >> repo, we try not to do it. From what I understand we're waiting on the >> gcc4.0.1 release to see if it's fit for our use. >> >>"its [sic] not that hard" -- Have you actually tried it? Sometimes >> things are a lot more difficult than you think... Ever thought that >> maybe someone else already thought about this and then *didn't* do it >> for some specific reason? >> >> >> >>>if you 'move the kernel' will you also be setting up another fork, one >>> for 2.4 and one for 2.6, and then a gcc 3.x version and a gcc 4.x >>> version? wow.. 4 kernels now, where all you have to do is start moving >>> over to gcc more.. it is TESTING afterall.. right? >>> >>>or maybe the structure of Current , Testing , Unstable.. needs to be >>> reviesed.. >>> >>> >> >>Just a note. I'm not trying to be critical of you as a person. But >> when you say something like "we really should change this..." no one's >> going to listen to you. Personally, I find it annoying when people say >> we should change something for the "better" but then never actually say >> what we should do. >> >> >> >>>i'm sure users are leaving daily cuz this mess, but dont know where to >>> go or where to look, or have any idea of whats wrong.. >>> >>> >> >>People are leaving because the stuff we say might be broken is broken? >> That seems a little pretentious, doesn't it? They demand perfection >> even where we admit we're not perfect. I don't know how much I like >> users like that anyway... >> >>Jason >> >> >> >>------------------------------------------------------------------------ >> >>_______________________________________________ >>arch mailing list >>[email protected] >>http://www.archlinux.org/mailman/listinfo/arch >> >> > > > _______________________________________________ > arch mailing list > [email protected] > http://www.archlinux.org/mailman/listinfo/arch _______________________________________________ arch mailing list [email protected] http://www.archlinux.org/mailman/listinfo/arch
