| 1 |
On Mon, 21 Oct 2019 17:58:51 -0700 |
| 2 |
Matt Turner <mattst88@g.o> wrote: |
| 3 |
|
| 4 |
> I'm not sure what this is in reference to so it seems to be a |
| 5 |
> non-sequitur, but I like the policy of at least waiting a day for |
| 6 |
> review of non-critical fixes. Phrased another way, let people in every |
| 7 |
> timezone have a chance. |
| 8 |
|
| 9 |
Its not aimed so much at the person writing this proposal, but more so, |
| 10 |
something that has happened before and was annoying and I had to have |
| 11 |
annoying conversations about why this was not-ok, where the agent in |
| 12 |
question didn't attempt to reach out to any team member of a very |
| 13 |
critical package before "just doing it". |
| 14 |
|
| 15 |
The nature of the change *was* simple enough that had we seen it, we'd |
| 16 |
have ACKed it easily. |
| 17 |
|
| 18 |
But the need to get lucky and spot the commit in the history and review |
| 19 |
it after-the-fact hoping the agent didn't break anything is where this |
| 20 |
is "not-ok". |
| 21 |
|
| 22 |
Yes, I hate to have to re-iterate in policy behaviour that I consider |
| 23 |
to be a sensible reasonable default... but it turns out, not everyone |
| 24 |
has that same sensibility. |
| 25 |
|
| 26 |
I'm fine with a policy that allows for non-maintainer contributions, |
| 27 |
just the stipulations of "try contact the maintainer, wait a reasonable |
| 28 |
amount of time based on the overall combination of that packages |
| 29 |
importance and the effective availability of people who maintain it to |
| 30 |
even respond to a ping". |
| 31 |
|
| 32 |
Because the fact is, non-maintainers have substantially less |
| 33 |
understanding of the total net of complexities, both in portage and |
| 34 |
outside of portage (by way of project specific tracking projects), and |
| 35 |
they're less equipped to make a judgement call as to wether or not a |
| 36 |
change that looks trivial, is trivial in the grand scheme of things. |
| 37 |
|
| 38 |
( For instance, tweaking the package-keywords entries on bulk keyword |
| 39 |
request I filed riles me up, because I need a lot of tooling to |
| 40 |
maintain that stuff, and currently, other people tweaking stuff that is |
| 41 |
outside their responsibilities messes with my ability to do this |
| 42 |
effectively. And this isn't even visible 'in portage', its just more of |
| 43 |
the same pattern of 'touching stuff that's not your responsibility to |
| 44 |
touch without contacting the people whos responsibility it is' ) |
Archives