1 |
> Firstly, this is not a flame. Please give me the benefit of the doubt in |
2 |
> that respect :) |
3 |
|
4 |
Neither is this. |
5 |
|
6 |
> There is no gentoo stable in the same way that Obsd has stable. Obsd stable |
7 |
> can pretty much be guaranteed to work and play happily. Ebuilds are marked |
8 |
> stable or unstable based on whether the _ebuild_ is known to be reliable, |
9 |
> not the package which the ebuild installs. |
10 |
|
11 |
An ebuild is a bash script -- it doesn't take much for an ebuild to be stable. In this case, rob, your reasoning is completely flawed. The set of ebuilds marked stable for a specific architecture contains programmes and libraries and utilities, etc etc that are known to work reasonably well (given the large variation in customisation and optimisation settings). |
12 |
|
13 |
According to policy, an ebuild is marked stable for a platfrom if and only if recent history (~1 month) shows no new bugs open for it, and all previously opened bugs having been resolved. That gives us a good indication (assuming, of course, that users use our bug tracker -- and surely gentoo-stats and gentoo-stable websites will start to grow an increasing role in this respect) that the package works reasonably well for the majority of users. |
14 |
|
15 |
|
16 |
> There is no indication inside of portage as to whether a program is stable |
17 |
> or not, other than extreme cases where ebuilds are masked because the app |
18 |
> is very broken. Its not possible for us to say "this is a stable platform" |
19 |
> for a gentoo "system" can include any number of programs that we may or may |
20 |
> not have written ebuilds for which can affect the system. |
21 |
|
22 |
Rob, you, your mentor and I need to have a chat. If _we_ did not write it, then _we_ should have checked it, and _carefully_. And if it did not pass basic tests of functionality, syntax, etc, then _we_ should not have put the damned thing into portage in the first place. |
23 |
|
24 |
|
25 |
> |
26 |
> Obsd peeps know exactly what apps are installed in their base system, so |
27 |
> they can mark the stuff stable when they're fairly sure the base system |
28 |
> doesn't blow up. |
29 |
|
30 |
Gentoo peeps know exactly what apps are installed in their base system, so we can mark stuff stable when we're fairly sure the base system doesn't blow up. |
31 |
|
32 |
> Also, I think you misunderstand "releases". 1.4 is a release of an install |
33 |
> CD and maybe GRP. Thats it. It makes no difference to the actual system |
34 |
> once you start running emerge sync you'll be back in the same place as |
35 |
> someone who installed with a 1.3 install CD and has been running emerge |
36 |
> sync. |
37 |
|
38 |
|
39 |
You got this one right. |
40 |
|
41 |
|
42 |
-- |
43 |
Seemant Kulleen |
44 |
Developer and Project Co-ordinator, |
45 |
Gentoo Linux http://www.gentoo.org/~seemant |
46 |
|
47 |
Public Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x3458780E |
48 |
Key fingerprint = 23A9 7CB5 9BBB 4F8D 549B 6593 EDA2 65D8 3458 780E |