Right now, the code will create branches locally when doing updates, and then push them as branches to their origin repo.
This is fine for @pacchettibotti because it has write access to all of them, but it's not fine if someone else wants to run the project.
I think the main issue is this line
|
, "git push --set-upstream origin " <> branchName |
..where we set
origin as the upstream. I suspect that's not the only place where this idea is hardcoded though.
So we should extend the code so that:
- we can specify a username in the
config.json or with an envvar
- ..and then when pushing to upstream, we'd check if that username has a fork of the repo, and if not we'd create it and push to it.
Right now, the code will create branches locally when doing updates, and then push them as branches to their origin repo.
This is fine for @pacchettibotti because it has write access to all of them, but it's not fine if someone else wants to run the project.
I think the main issue is this line
pacchettibotti/src/PacchettiBotti/Run.hs
Line 98 in 85234a5
..where we set
originas the upstream. I suspect that's not the only place where this idea is hardcoded though.So we should extend the code so that:
config.jsonor with an envvar