Only run Travis for the master branch & run sanity checks with DMD too#441
Only run Travis for the master branch & run sanity checks with DMD too#441PetarKirov merged 2 commits intodlang-tour:masterfrom
Conversation
|
👍 |
.travis.yml
Outdated
| # Compiler to static binary with gdc | ||
| - if [[ "${DC}" == "dmd" ]]; then dub test --compiler=${DC}; fi | ||
| - if [[ "${DC}" == "dmd" ]]; then dub --compiler=${DC} -- --sanitycheck; fi | ||
| # Compiler to static binary with ldc |
There was a problem hiding this comment.
Small typo - I think it should be "Compile to ...".
|
This leads to about 30 seconds increase in the run time of the DMD build, but as it's still faster than LDC, I think we don't need to care ;-) |
|
Apart from small nit, LGTM. Feel free to merge when ready. |
|
I tried the new squash feature of GitHub and looks like it works good - it added to the commit message a link to this PR. |
Thanks & good catch!
If we like it, we should be able to make it the default (by disallowing normal merge commits). This way I can at least not accidentally press on "merge" on my phonem :D Btw a minor thing that i noticed - it copies all commit messages in the description (like the usual
I think it does this by default for merge commits too: |
Yeah, I remember that one. I guess that guy could have squashed his commits into an "appeasing the git Nazis" message :D
The good thing is that it opens a textbox where you can manually edit the commit message. I just decided to leave it as it is for this PR.
I already new that it does that for merge messages, I was just worried that we might loose the PR context if squashing was used. C.f. Vladimir's post on the forum.
I don't think this would be a good idea, because it would make larger PRs harder to review, which usually are not if the they're built from a number of small commits. |
A bit more of Travis magic applied:
masterbranch (currently it runs twice with all PRs made with branches ondlang-tour, e.g. hard to avoid when coming from the web UI)