On Mon, Feb 26, 2007 at 10:45:24AM +0100, Bernhard R. Link wrote: > * Pierre Habouzit <[EMAIL PROTECTED]> [070226 03:03]: > > So now let's do a simple calculation. 100 bugs, 20 minutes, that's > > 2000 minutes, over 6 weeks, that's 333 minutes a week, meaning at least > > 6 hours a half of work. Just to keep up with bugs. Of completely tedious > > work. > > > > Add to that: working on the backlog, working on the bugs that in fact > > need 1hour of work, packaging new upstreams, doing some maintenance on > > the repository and so on, and KABOOOM, either you have a time machine, > > or there is not enough time. > > > > so well, hmm let me think again ... YES THIS IS A DAMN PERFECT > > ARGUMENT. > > Sorry to say it that way. But after I read this, not letting such > packages in a stable release before they get enough manpower to be > handled, sound a better idea then before.
Stripping KDE, php, xorg, gnome, iceweasel, the libc out of stable would indeed make releases a lot less painful. Any other brilliant idea around ? Also consider that in the huge mass of bugs, for any of the mentioned packages, I'd say 25 to 50% of the bugs are just either invalid or long gone. Those teams do an amazing job dealing with the most urgent bugs, _AND_ keeping up with upstream. Because such packages aren't supported very long, because upstreams have the same man power issues, and they only support _some_ releases, if we do not keep up, then it means putting an unbearable pressure over the already quite loaded security team. Could we go back to sanity and reality, please ? -- ·O· Pierre Habouzit ··O [EMAIL PROTECTED] OOO http://www.madism.org
pgpJfr4WAWtSO.pgp
Description: PGP signature
