Only clone once, not once per subdir #443
Conversation
|
bors try |
tryBuild failed: |
|
bors try |
|
bors try |
|
What is the state of this? I would like to tackle #408, but having this PR here as a baseline would make things a lot easier. |
|
Just need review iirc |
|
maybe ping them again? After 10 months I find that reasonable |
|
If I know my GitHub notifications these comments will do it :). Also I have merge rights so if you or anyone else wants to review, that would suffice for me to feel comfortable merging. |
|
@fchorney or @mattBrzezinski Could you review this? |
|
Hey sorry, I have changed jobs this summer and am no longer working in Julia. I probably won't really have time to look at stuff like this going forward. |
I'll put it on my TODO list, hopefully get around to it in a few days. Just busy with other things currently :( |
|
The docs failure is already resolved by #453. This needs a rebase now. |
|
bors merge |
|
Build succeeded! The publicly hosted instance of bors-ng is deprecated and will go away soon. If you want to self-host your own instance, instructions are here. If you want to switch to GitHub's built-in merge queue, visit their help page. |
Closes #442
This PR has two commits (+ formatting); the first lets tests pass locally on my machine, by specifying the
init-branchinstead of relying on it being set tomaster. The second does the change to close #442. Besides speeding things up and possibly avoiding weird cloning issues, this also makes the code more "compositional" and simplifies the tests.edit: I've also pushed an additional commit to re-use the same local clone for each PR. It seems wasteful to make so many clones, and in #442 for some reason for me the first clone seems to work and the subsequent ones fail.