1 |
On 06/01/06, Brian Harring <ferringb@g.o> wrote: |
2 |
|
3 |
> Probably better to iron out what y'all actually need and what the dev |
4 |
> community is willing to put up with. |
5 |
> |
6 |
> Eg, would do some research into it, read the archives from last few |
7 |
> wars over it, in general find (and address) the issues that lead to |
8 |
> glep19 going still born. |
9 |
|
10 |
The problems being: |
11 |
|
12 |
1) Manpower. There are already 10,000 open bugs in bugzilla (and |
13 |
growing) without adding more. |
14 |
2) Lack of interest. Most developers aren't interested in supporting |
15 |
"old" packages. |
16 |
3) The enterprise. Both of the above problems would be fixed if |
17 |
enterprises were contributing developers and/or money. However, they |
18 |
aren't, so why is that? The truth is most enterprises want to go to a |
19 |
big company to buy their software. They want one homogeneous binary |
20 |
system, not a flexible way of building packages from source, and they |
21 |
want someone else to do it and be responsible for it. |
22 |
|
23 |
Look at the arch/~arch split - as a guideline ebuilds are supposed to |
24 |
move from ~arch -> arch after some reasonable period of time with no |
25 |
bugs reported (eg. 30 days). Yet as the friendly script shows us, |
26 |
there are many ebuilds that that stuck in ~arch for a long long time. |
27 |
Why? The main reason is developer interest - a lot of people only run |
28 |
~arch, so the motivation is reduced. It's difficult for users to help |
29 |
- in the end it's easier to mark stuff ~arch in |
30 |
/etc/portage/package.keywords than to file a bug requesting that some |
31 |
developer test and change the ebuild. Proper testing of a tree |
32 |
requires developers to run both arch and ~arch. The stable proposals |
33 |
would add yet another tree to test. |
34 |
|
35 |
The only way I can see to solve these problems is more automation. |
36 |
Link bugs directly to individual versions (or sets) of ebuilds. Have a |
37 |
defined process for marking stuff stable - either x days with no bugs, |
38 |
and/or through sampling of users installed packages to see what is |
39 |
actually installed out there. Bug voting would fix the disconnect |
40 |
between users and devs (sometimes people are interested in working on |
41 |
a random problem to help gentoo - at the moment it's difficult to see |
42 |
which bugs are really the important ones to users). Decentralise |
43 |
development - allow users to share their own patches in a searchable |
44 |
system, rather than force everything through bugzilla; add support to |
45 |
portage to utilise other peoples trees - the current system of devs |
46 |
publishing overlays on their web pages/svn and users having to |
47 |
manually wget or svn up is too difficult for users and removes ebuilds |
48 |
from the tree - so they are less visible and get less testing. For QA |
49 |
gentoo really needs a compile farm with automated compile, install and |
50 |
test (from those ebuilds that support it). Make the system smarter, |
51 |
instead of throwing more people at the problem. |
52 |
|
53 |
-- |
54 |
gentoo-dev@g.o mailing list |