We do actually want CI to pass a branch #1123
Conversation
There was a problem hiding this comment.
looks sensible, but we should document it - https://github.com/matrix-org/sytest/blob/7178c8bd53557c739e99771bf67c9116ec0d731f/docker/README.md lists all the other env vars that the docker image understands, so we should include this one
docker/README.md
Outdated
| homeserver git repository to be mounted at `/src`; otherwise, the image will | ||
| try to fetch that repository from GitHub. |
There was a problem hiding this comment.
Is this correct? I thought that it would just fail if /src didn't exist.
docker/README.md
Outdated
| * `SYTEST_BRANCH`: controls which branch of synapse or dendrite we fetch when `/src` | ||
| is missing. If omitted or if that branch doesn't exist, we'll fetch the appropriate development branch. |
There was a problem hiding this comment.
this isn't quite right. As above: if /src is missing, I don't think it works at all...
There was a problem hiding this comment.
Argh! Thanks. I think I keep confusing /src and /sytest. It's the CI job that has to fetch the source
There was a problem hiding this comment.
No worries! It is pretty confusing.
I was confusing /sytest with /src, and bootstrap.sh with CI config.
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
Co-authored-by: Richard van der Hoff <1389908+richvdh@users.noreply.github.com>
|
I'll run the docker |
PR #1123 broke the `Trying to get same-named sytest branch` stage for Dendrite CI which is still in Buildkite.
this pr not from my fork but from the repo itself