| 1 |
On Fri, 6 Dec 2019 04:33:36 -0500 |
| 2 |
Tim Harder <radhermit@g.o> wrote: |
| 3 |
|
| 4 |
> On 2019-12-06 Fri 04:03, Alexis Ballier wrote: |
| 5 |
> > it's not just like repoman and cvs since repoman commit did push ;) |
| 6 |
> > it will never be perfect but i really like repoman commit to refuse |
| 7 |
> > to even commit if there's something obviously wrong |
| 8 |
> |
| 9 |
> I'm more of the opinion (and am working towards that practicality in |
| 10 |
> terms of runtime speed) that a subset of QA checks should be run as a |
| 11 |
> git hook which would cause push failures on certain classes of bad |
| 12 |
> commits. |
| 13 |
|
| 14 |
|
| 15 |
There should be both. Amending the 23rd commit message because one |
| 16 |
mistyped a semicolon is kind of a PITA. |
| 17 |
|
| 18 |
> > as you write below, it's just a matter of checking exit status and |
| 19 |
> > using git, which can be done by scripting, but the script is |
| 20 |
> > standard (*) and mandated to be part of the workflow |
| 21 |
> |
| 22 |
> > it also allows to check or templatize commit messages to follow |
| 23 |
> > policy |
| 24 |
> |
| 25 |
> Technically pkgcheck supports more git-related checks than repoman |
| 26 |
> last I checked, i.e. result keywords including BadCommitSummary, |
| 27 |
> DirectStableKeywords, DroppedUnstableKeywords, DroppedStableKeywords, |
| 28 |
> DirectNoMaintainer, and MissingSignOff; with possible future additions |
| 29 |
> such as warning when modifying deps in an ebuild without revbumping. |
| 30 |
> |
| 31 |
> Futhermore, one can scan against all commits in parallel via `pkgcheck |
| 32 |
> scan --commits` which will enable potential commit results that are |
| 33 |
> otherwise skipped. |
| 34 |
|
| 35 |
All this seems post-commit, not pre-commit. |
| 36 |
|
| 37 |
> Anyway, my main point is that if someone really wants commit |
| 38 |
> functionality it's semi-trivial to script something similar to what |
| 39 |
> repoman does (assuming exit status/api support exists) and if it's |
| 40 |
> decent enough quality (including tests) I'd probably consider adding |
| 41 |
> it to the pkgcheck repo. |
| 42 |
|
| 43 |
It doesn't necessarily have to live in the pkgcheck repo, but it should |
| 44 |
definitely not be "meh, script it yourself, it's trivial" since that |
| 45 |
will probably lead to several scripts with varying degrees of quality |
| 46 |
and brokeness. |
Archives