Adam Di Carlo wrote: > >IMO, the purpose of freezing is to stabilize. I don't care much if Debian is > >stable before the freeze, just so long as it freezes. Yes, a short freeze is > >good, but better a long freeze than an interminable unfrozen unstable. > > I don't make the decision. > > I personally feel that a *feature-complete* boot-floppies is essential > *prior* to freezing.
Okay, but my concern is that based upon your last posting, there may be other essential packages which are undergoing severe breakage during the interval while boot-floppies works towards this goal. Somebody suggested a "light freeze" before, which would entail freezing all essential packages until all could be brought to a sufficient state of completion and stability, and then freezing the rest of the distribution. > >There should be a *firm* date for freezing. > > There is -- Jan 15th. I understand that this is the presently announced date, but if it gets pushed back again because of other broken essential packages, I know that I and many other people are going to be very unhappy about it. And the more fundamental point is that Jan 15th 2000 is way, way too long since the last release -- I recognize that there is nothing to be done about the past, but we need to do much better in the future. FWIW, I have imposed a partial freeze upon my own packages, and for instance have no intention of updating oo2c to version 1.5.0 unless I am certain it will not break anything else -- and presently it will break voberon. If that means potato releases with a slightly obsolete oo2c 1.4.7, so be it, and this should not concern anyone -- unless it takes another year and a half before the next stable release. In that case, "slightly obsolete" becomes an extreme understatement.
