Re: Some observations regardig the progress towards Debian 3.1
Must say I agree on alot here, that something has to change with the state
of unstable/testing?
Unstable is to be unstable: why have experimental?
All packages make their way into unstable, after maintainer have done some
work on it.
Some point I wanna make:
-Unstable should be frozen on a sertian time, let`s say 3/4 times a year-
then this makes it`s way into testing.
-Tesing will then have a set of packages that are buildt against each other
as have been mentioned here.
-Debian`s main "problem" is that packages like gnome/kde and xfree are to
old- This does not have anything to say in stable branch since one don`t use
x on server`s (well you`ll probably find some Redhat servers running that)
Focus more on this than having 15000 packages. Winn the enduser! you have
the server marked with the apps that are included.
-Seems that Debian uses to much time to "fiddle" with unstable/testing=
problem
-So I say Unstable should not be considered as a system other than for
upload of packages that have been somewhat tested ( a alpha system) and most
important ONLY after unstable have been frozen should a new package be
uploaded, if not very important security/bug fixes.
-Now new packages enter unstable when the "maintainer feals like it".
-Testing get`s frozen after unstable have been "fixed" and it seems like it
all works.
-Testing can then be worked on for some time and considred as a beta system.
Conclusion:
When testing reaches a certion point eg. less than 500 bugs one can consider
to release this "freeze" it for eg another 3 monts and this be the next
stable.
_________________________________________________________________
Hotmail snakker ditt språk!
http://www.hotmail.msn.com/cgi-bin/sbox?rru=dasp/lang.asp - Få Hotmail på
norsk i dag
Reply to: