| 1 |
Am Sonntag 03 Juni 2012, 10:18:24 schrieb Robin H. Johnson: |
| 2 |
> I propose: |
| 3 |
> - merges are explicitly allowed, even non-fast-forwards |
| 4 |
> - all commits MUST be signed |
| 5 |
> - if you include a commit from a user: |
| 6 |
> author := non-@gentoo |
| 7 |
> committer := @gentoo |
| 8 |
> signer := $committer |
| 9 |
> |
| 10 |
|
| 11 |
Sounds reasonable given the current state of git. Let's just be clear about |
| 12 |
the following consequence (I hope I understand this correctly): |
| 13 |
|
| 14 |
* User makes signed improvements in gentoo-x86 clone |
| 15 |
* Developer pulls from user and >merges< |
| 16 |
* Developer's history contains commits by user, which cannot be pushed to |
| 17 |
gentoo-x86 |
| 18 |
|
| 19 |
Which means in the end "all merges are explicitly allowed, as long as they |
| 20 |
only contain developer commits; commits pulled from users must be rebased". |
| 21 |
|
| 22 |
This is something that (IMHO) we could certainly live with; the only thing I |
| 23 |
am worried about is, how do we automatize it so a developer who is not end-of- |
| 24 |
the-line git guru and ends up with some user-signed commits in his history can |
| 25 |
clean up his local tree again. |
| 26 |
|
| 27 |
|
| 28 |
-- |
| 29 |
|
| 30 |
Andreas K. Huettel |
| 31 |
Gentoo Linux developer |
| 32 |
dilfridge@g.o |
| 33 |
http://www.akhuettel.de/ |
Archives