1 |
On Wed, 2009-05-13 at 20:47 -0500, Jeremy Olexa wrote: |
2 |
> Mart Raudsepp wrote: |
3 |
> > Hello, |
4 |
|
5 |
Hey, |
6 |
|
7 |
> > I have had this project in my mind for a while, so it's about time to |
8 |
> > get it out there, as to see if feedback finds it a good one - and if |
9 |
> > that is so, if there are people who want to make it happen. |
10 |
> <snip> |
11 |
> |
12 |
> Hmm, I wonder what the point is when there is 400 maintainer-needed bugs |
13 |
> open.. |
14 |
|
15 |
The point is making popular packages available in the _portage tree_. |
16 |
When widely used popular packages end up in maintainer-needed, they tend |
17 |
to be relatively quickly picked up by other teams or developers. |
18 |
maintainer-wanted hopes for the same, that they will tend to be picked |
19 |
up by others. Most of the 400 packages affected by maintainer-needed |
20 |
bugs are probably as such because close to no-one cares about them, and |
21 |
caring by the project proposed here doesn't get us any closer to world |
22 |
domination(tm) - we'd just have more packages of quality, except no-one |
23 |
uses those packages. We already have a set of people, of which you |
24 |
Jeremy seem to be participating in, that takes care of maintainer-needed |
25 |
bugs that have user patches available, hence the packages people |
26 |
actually care about are eventually taken care of as well as I |
27 |
understand. Maybe that project should formalize themselves as well. Or |
28 |
we can add that set of tasks to this one. |
29 |
|
30 |
That said, my initial idea months ago was related to the |
31 |
maintainer-needed alias, taking care of the packages of greater interest |
32 |
marked maintainer-needed and introducing new ones that are encouraged to |
33 |
be taken over. But by now I don't think mixing those together is a good |
34 |
idea. The community maintained packages idea from another e-mail was |
35 |
quite good to not mix with maintainer-wanted either (the point about |
36 |
making sure new package requests and bugs against in-tree |
37 |
maintainer-wanted packages don't get mixed), except I think that naming |
38 |
doesn't so strongly express that other dedicated maintainers are |
39 |
desired. |
40 |
|
41 |
|
42 |
> I think a maintainer-wanted team won't accomplish much because it just |
43 |
> uses more developer time from a pool of "not enough time" that exists |
44 |
> already. |
45 |
|
46 |
Volunteer manpower doesn't work quite like that. |
47 |
Volunteers have as much time as they make for this as a hobby of |
48 |
interest. Developer A is interested in keeping old crufty stuff dropped |
49 |
to maintainer-needed in quality condition as they like that sort of |
50 |
thing, while developer B doesn't and he likes the satisfaction of making |
51 |
much desired new packages available to the wider user base instead. If |
52 |
you don't have a project encouraging that, developer B can end up just |
53 |
not taking more time for Gentoo and does more of other stuff instead, |
54 |
lets say gardening or watching random TV. |
55 |
|
56 |
Because we don't provide monetary motivation to take care of exotic and |
57 |
outdated packages to get that out of the way shouldn't block people |
58 |
motivated in providing popular packages that would be used by a larger |
59 |
set of the user base and contribute to the popularity of Gentoo. |
60 |
|
61 |
> If people are a) too lazy to contribute to sunrise, b) don't |
62 |
> know about sunrise, or c) don't know enough about ebuilds to contribute |
63 |
> to sunrise, then we need to fix[1] that. |
64 |
|
65 |
Sure, the sunrise project can do all of that. That doesn't make the |
66 |
packages available in the official tree. The maintainer-wanted project |
67 |
however can tap into the work done by sunrise when a popular package is |
68 |
to be added to the tree that already exists in Sunrise. If certain |
69 |
interested in the package users are contributing to Sunrise for that |
70 |
package, they could hopefully be turned into proxy maintainers in the |
71 |
official tree version and perhaps even eventually become developers as |
72 |
well when they have interest in a larger set of packages. If they have |
73 |
been maintaining a lot of different package in Sunrise before that, they |
74 |
seem to be a good candidate in joining the maintainer-wanted team too |
75 |
then, as they seem to be motivated by the kind of work they do, as same |
76 |
work was done in an overlay by him/her before. |
77 |
Close collaboration with Sunrise is good, that is. So the end outcome |
78 |
would be that the packages that are used by many people are available in |
79 |
the official tree. |
80 |
|
81 |
> I don't see any reason to create a team that duplicates the sunrise |
82 |
> work. Keep in mind, I am against pretty much any overlay, I think work |
83 |
> should be kept in the main tree. But, for ebuild maintenance with |
84 |
> limited developer time, sunrise just makes sense(tm). |
85 |
|
86 |
Developer time is not so strictly limited. The motivation to spend more |
87 |
time on Gentoo instead of other daily non-work not-so productive |
88 |
activities might be limited. Yes, we also have the small set of |
89 |
developers that do a very large chunk of the work - they are limited in |
90 |
time indeed because they already are so motivated to use so much time |
91 |
for Gentoo. |
92 |
I am a strong believer in the correlation of motivation and time spent |
93 |
on volunteer hobby work as you can see. |
94 |
|
95 |
> Some other possible directions include: |
96 |
> 1) maintainer-needed team - Where a group maintains the set of 761 |
97 |
> m-needed packages. |
98 |
|
99 |
Sounds reasonable for achieving different goals. The popular packages of |
100 |
these 761 could be taken over by the maintainer-wanted team as well when |
101 |
that team is interested, while the maintainer-needed team would make |
102 |
sure those remaining 700 packages don't block stabilization of newer |
103 |
system packages, etc |
104 |
|
105 |
> 2) proxy maint project[2] - Where a group helps users commit to the main |
106 |
> tree, very similar to the sunrise project. Very similar to this proposal |
107 |
> but better conserves our developer time. |
108 |
|
109 |
Maybe adding this blurb to the proposal text: |
110 |
When the bug request for the package has contributors for the package, |
111 |
it is encouraged to draft them as a proxy maintainer(s), with the |
112 |
maintainer-wanted team being the committers and ensuring quality. |
113 |
|
114 |
> -Jeremy |
115 |
> |
116 |
> [1]: http://dev.gentoo.org/~darkside/sunrise_proposal.txt |
117 |
> http://dev.gentoo.org/~darkside/sunrise_status.txt |
118 |
> [2]: http://dev.gentoo.org/~antarus/projects/proxy-maint/ |
119 |
|
120 |
-- |
121 |
Mart Raudsepp |
122 |
Gentoo Developer |
123 |
Mail: leio@g.o |
124 |
Weblog: http://planet.gentoo.org/developers/leio |