[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Re: bugzilla 2.18-6, sarge candidate



Hi,

Users ask for many things that it's not feasible for us to give them. The fact is that the diff between 2.16.7-5 and 2.18-6 is 197,000 lines long; while I accept that many of these changes are improvements, there is also the possibility of regressions, and bugzilla 2.18 did not get uploaded to unstable until April 18 -- i.e., 17 days after our "last call" email went out on d-d-a, and providing only 16 days of testing of this upstream branch by users of unstable prior to the freeze (and with four uploads during that period, no less).

I understand why you are arguing for its inclusion, and I realize (from comments on IRC) that you've put a lot of effort into trying to get 2.18 ready for sarge, but I'm afraid I just don't see that it's ready in time, sorry.

I was told that Debian, unlike commercial vendors, release only when it is ready and do not "rush for releasing on time". I was told that Debian package versions are old but they are stable. If I need newer version with more features I need to get something "testing" and "unstable". Otherwise I need to use a 3+ year old release.

One year from now, I cannot imagine how Debian can promise me my package is stable. Upstream security fixes most probably fall in the 197,000 lines of changed/moved code. Upgrade scripts from upstream may not work for jumping over 3 versions, and I need to depend on one coded specifically for Debian. "How am I going to explain these to my boss," not shouted by me but the mass of Debian users (because I am running the new packaged version even before it is in experimental).

I totally agree that RMs have your procedures, and I believe that I cannot change your decision anyway. But I really have no choice except to stay with Debian, no matter how flawed I felt. I like Ubuntu for desktops, and thanks vorlon. I hope Debian can keep its promise to provide stable but good enough software, like how Ubuntu achieves its own set of objectives. I don't think a finite set of written rules is enough. The decision made by smart people is always smarter. Thanks for reading.

--
Regards,
Alan



Reply to: