1 |
On 2/4/19 7:38 PM, Michał Górny wrote: |
2 |
> Reviews can be done by devs or privileges users. Review by dev gives 3 |
3 |
> rep points, and by privileged user gives 1 rep point. Therefore, |
4 |
> a commit is merged if it's either reviewed by dev or 3 privileged users. |
5 |
|
6 |
This could be 'exploited' with a group of friends. By exploited I mean |
7 |
small inner circles forming, where people just approve their friends |
8 |
commits without looking at them. I think after 3 approvals it should be |
9 |
pushed to be on top of some(/all) dev's to-be-merged list. But then |
10 |
again, would this system be any faster than current proxy-maint |
11 |
workflow, where other proxied maintainers already review and approve PRs? |
12 |
|
13 |
|
14 |
What about ebuilds not many people have knowledge of? Say, java stuff? |
15 |
If no user wants to take a look, it will always require a review from a |
16 |
dev, and judging how that goes even with current Github PRs, will it |
17 |
_ever_ get approved here? |
18 |
|
19 |
|
20 |
What would motivate a developer to review these ebuilds, if there's |
21 |
still separate proxy-maint stuff to work on? |
22 |
|
23 |
|
24 |
> |
25 |
> |
26 |
> Users gain reviewing privilege also via reputation points. If a commit |
27 |
> range including user's commit gets merged to master, user gets 1 rep |
28 |
> point (independently of number of commits in the range). When user gets |
29 |
> 5 rep points, he can start reviewing stuff. |
30 |
|
31 |
So this requires people to make commits to this overlay before being |
32 |
able to review there? Some system should exist, where for example your |
33 |
commits to ::gentoo counts toward this. Otherwise this could encourage |
34 |
people to make meaningless commits just to satisfy the counter. |
35 |
|
36 |
|
37 |
> Your updated thoughts? |
38 |
> |
39 |
_If_ in this approach a dev is still needed for merging stuff in the |
40 |
end, couldn't this somehow be applied to how the current proxy-maint |
41 |
system works? |
42 |
|
43 |
|
44 |
Is there a chance these ebuilds could end up in ::gentoo? _Could_ there |
45 |
be some voting system for that (although I believe that kills the |
46 |
incentive of ever adding this overlay as a user)? |
47 |
|
48 |
|
49 |
Still gotta say, this should prove to be a nice learning ground, and |
50 |
maybe a good place to host popular not-in-tree packages. Or with |
51 |
semi-dead projects where users are still trying to contribute (like MATE). |
52 |
|
53 |
Although I believe everyone's motivation is to get them in ::gentoo in |
54 |
the end. |
55 |
|
56 |
|
57 |
And as asked by someone in the previous thread, what's the plan of |
58 |
cleaning the overlay every now and then? (How?) |
59 |
|
60 |
Who takes care of broken packages? |