1 |
On Thu, Jun 16, 2022 at 2:06 PM Robin H. Johnson <robbat2@g.o> wrote: |
2 |
> A question I'd like the candidates to answer. |
3 |
> |
4 |
> What do you feel that Council can do to increase the contributions to |
5 |
> Gentoo? |
6 |
|
7 |
Continue encouraging tooling improvements. Our tooling has improved by |
8 |
leaps and bounds in recent years: |
9 |
|
10 |
- app-portage/iwdevtools for all kinds of useful QA checks |
11 |
- app-portage/gentoolkit's merge-driver-ekeyword for automatically |
12 |
handling rebase conflicts in KEYWORDS |
13 |
- app-portage/mgorny-dev-scripts for lots of useful scripts |
14 |
- app-portage/nattka for stable and keyword requests |
15 |
- app-portage/pram for applying GitHub PRs |
16 |
- dev-util/pkgcheck for static analysis checks |
17 |
|
18 |
In the same vein, we need to deprecate inferior tooling (like I |
19 |
proposed and Council approved with app-portage/repoman). Too much time |
20 |
is wasted correcting mistakes that e.g. repoman doesn't catch that |
21 |
pkgcheck does. Too much time is wasted fixing mistakes that would have |
22 |
been caught with pre-push CI (that Gitlab could provide in the |
23 |
future). |
24 |
|
25 |
But we still would stand to benefit from other improvements. I'd like |
26 |
to make a tool that emails maintainers monthly with a summary of |
27 |
things like |
28 |
|
29 |
- list of maintained packages that have newer versions available |
30 |
- list of maintained packages that are ready for stabilization |
31 |
- list of bugs assigned to the maintainer |
32 |
- maybe something about the last time the maintainer touched a |
33 |
particular package |
34 |
|
35 |
I think this would go a long way to resolving some issues I've seen in |
36 |
Gentoo, where developers don't remember that they are the maintainer |
37 |
of a package, or they don't notice that there's a newer version |
38 |
available, or that it's time to stabilize one of their packages, etc. |
39 |
These situations often end up blocking the work of more proactive |
40 |
developers. |
41 |
|
42 |
> - how do you feel development velocity of Gentoo should be measured? |
43 |
> - how do you feel technical debt of Gentoo should be measured? |
44 |
|
45 |
I don't think the answers to either of these questions is particularly |
46 |
instructive. |
47 |
|
48 |
For development velocity, I guess https://www.openhub.net/p/gentoo/ is |
49 |
at least interesting—particularly the number of contributors ("Up + 28 |
50 |
(4%) from previous 12 months"). |
51 |
|
52 |
> - how do we make it easier to get orders of magnitude more contributors? |
53 |
|
54 |
I don't think aiming for order-of-magnitude increases in contributors |
55 |
is a realistic goal. We've been within ±10% of 200 contributors |
56 |
(according to the OpenHub data) for the last 5 years. |
57 |
|
58 |
Gaining contributors is important though, and I think the most |
59 |
significant thing we can do here is to foster a community that people |
60 |
enjoy being a part of. The last few years has seen an exodus, |
61 |
voluntary and otherwise, or some rather antisocial characters. I think |
62 |
Gentoo is significantly more fun as a result. |
63 |
|
64 |
> - how do we make it easier to do QA on significantly more contributions? |
65 |
|
66 |
Gitlab. See below. |
67 |
|
68 |
> - what blockers do you perceive in the contribution processes, and how |
69 |
> do you think they should be tackled? |
70 |
|
71 |
Getting Gitlab going on Gentoo's infrastructure is in progress, and I |
72 |
have high hopes for it to provide significant benefits—including |
73 |
reducing the barrier to entry for contributors; at the same time I'm |
74 |
concerned about our infrastructure team's capacity to support it. |
75 |
|
76 |
Generally speaking, my most significant concerns about Gentoo are |
77 |
related to the infrastructure team's capacity to handle issues that |
78 |
only it can handle (or has permissions to handle). There are many |
79 |
ideas for improvement and/or requests that have languished for years. |
80 |
A few that come to mind from the top of my head: |
81 |
|
82 |
- Forums migration (was brought up as a potential blocker to doing |
83 |
something about the Off The Wall forum during Council discussions in |
84 |
December 2020) |
85 |
- Distfiles hosting (bug #176186, opened April 2007) |
86 |
- Allow retirement team to process retirements without infra |
87 |
involvement (bug #311511) |
88 |
- Various packages.gentoo.org bugs |
89 |
- Setting up the new $25k servers Council approved |
90 |
- Integration with RelEng processes is a pain |
91 |
- Gitlab + CI |
92 |
- Something needed for the new glsamaker? |
93 |
- Probably more I've forgotten |
94 |
|
95 |
It's just generally difficult to get infra to do almost anything in a |
96 |
timely fashion. I don't know what Council can do to resolve this |
97 |
situation other than to highlight it as a problem. |