1 |
Ühel kenal päeval, N, 10.03.2022 kell 18:18, kirjutas Joshua Kinard: |
2 |
> I stick to the officially-published method of checking and committing |
3 |
> changes: |
4 |
> https://devmanual.gentoo.org/ebuild-maintenance/git/index.html |
5 |
> |
6 |
> The two tools highlighted there for the bulk of the work is repoman |
7 |
> and pkgdev. repoman is cited twelve times, pkgdev is cited six |
8 |
> times. pkgcheck is mentioned once. pkgcommit has no mentions. |
9 |
> |
10 |
> From that, one should not be faulted for assuming that repoman is the |
11 |
> more important tool, if not preferred tool, with pkgdev coming in |
12 |
> second place. |
13 |
> pkgcheck comes across as entirely optional and even seems equivalent |
14 |
> to 'repoman full', and how would one know that pkgcommit even exists? |
15 |
|
16 |
I believe the very purpose of this thread is to have a consensus/pre- |
17 |
announcement before actually editing the official documentation as part |
18 |
of the process of deprecating repoman. |
19 |
|
20 |
> That all said, most of my checks are done pre-commit, not pre-push. |
21 |
> It's annoying trying to unwind a commit with a mistake in git, so I |
22 |
> aim to not have any in the commit itself. By the time I get to the |
23 |
> push stage, everything MUST be perfect before I hit 'git push'. |
24 |
> Cause once something is public, there's no going back. In a handful |
25 |
> of cases, I've gone as far as nuking my local copy of the tree, re- |
26 |
> cloning it, and re-doing everything again just to fix a mistake that |
27 |
> I should have caught pre-commit. |
28 |
|
29 |
git commit --amend, interactive rebase, etc (all before push). |
30 |
I suggest to learn them and embrace them. Re-cloning is not fun, just |
31 |
for something that a quick interactive rebase or even a git reset -- |
32 |
hard destructive operation instead of re-clone would solve. |
33 |
|
34 |
Also the benefit of using pkgcheck is to actually be able to make the |
35 |
same checks that CI would do before you push, so you can amend your |
36 |
commits to fix issues before they hit the server and CI and break the |
37 |
tree. pkgcheck is so fast that it can do full tree checks in a |
38 |
reasonable time (repoman would take days on a radiator mips when you go |
39 |
outside single package), and I believe has features to have it check |
40 |
all your commits that haven't been pushed yet at once, checking only |
41 |
what it can to not be too slow to not use (so you don't need to run the |
42 |
check with each commit but for all of them once you commit - and if |
43 |
issues, again, git interactive rebase). |
44 |
|
45 |
repoman does not catch many many QA issues that pkgcheck (and therefore |
46 |
CI) does. |
47 |
|
48 |
> How about making an edit to repoman to have it throw a deprecation |
49 |
> warning for itself and to install whatever we all agree should |
50 |
> replace it? |
51 |
|
52 |
I would think that's a natural next step after this thread, and why |
53 |
this thread exists. |
54 |
|
55 |
> > |
56 |
|
57 |
Mart |