1 |
Brad Laue wrote: |
2 |
> Debian may be stable, but Woody's sheer age really does indicate that |
3 |
> something is wrong with their development model. I don't imagine |
4 |
> suggesting a faster paced release model to their core developers would |
5 |
> be met with much openness, nor would it be put to a vote. FreeBSD and |
6 |
> RedHat, to name two, have live package build systems (ports/rawhide |
7 |
> respectively) and cut a release from these every four months on the dot |
8 |
> with impeccable QA. I don't think I could suggest such a thing to Debian |
9 |
> developers without being laughed out of the discussion. |
10 |
> |
11 |
|
12 |
My impression is that the paralysis we've all seen in the Debian |
13 |
community is due to two factors: 1) ridiculously inflexible rules |
14 |
regarding free vs. non-free software and 2) a |
15 |
we're-the-best-and-don't-need-to-change attitude. I've seen long and |
16 |
apparently serious discussions on the Debian lists over whether gcc |
17 |
should be removed from the main distro because the man pages aren't |
18 |
released under a "free" license. I've seen hundreds (who knows, |
19 |
thousands?) of posts describing problems with apt-get and deselect |
20 |
dismissed cavilierly. In my opinion THAT is what's killing Debian - the |
21 |
inability to distinguish worthwhile questions from intellectual |
22 |
masturbation and the unwillingness to accept criticism. I don't see |
23 |
either of these elements present in Gentoo now, and I don't see zhen's |
24 |
proposals changing that. |
25 |
|
26 |
As for the long time between Debian releases, it seems to me that trying |
27 |
to support 10,000 packages on 11 architectures with "rock-solid |
28 |
stability" is a big part of that. Again, I don't see anything in zhen's |
29 |
proposal on that order of magnitude. Certainly, asking one attendee of |
30 |
a meeting to dump a log to the website isn't going to slow development. |
31 |
|
32 |
Heck, the glacial pace of Debian releases wouldn't even be a problem if |
33 |
their upgrade system was robust. "Install the old version, then run |
34 |
apt-get upgrade repeatedly until the conflicts go away" isn't |
35 |
acceptable. Again, Gentoo does not suffer from this problem - it's |
36 |
essentially a releaseless distro. |
37 |
|
38 |
|
39 |
|
40 |
-- |
41 |
Stephen C. Martin |
42 |
|
43 |
PGP/GnuPG key 1024D/8C4FCA5D |
44 |
|
45 |
|
46 |
|
47 |
|
48 |
|
49 |
|
50 |
-- |
51 |
gentoo-dev@g.o mailing list |