Matthew R. Becker comments

Results 684 comments of


                                            Matthew R. Becker

Shorten config & job names

Sure I get that. We could do better on extra underscores for example. But having the variable names in the config file path is very useful especially for the build-locally...

Support for Travis as a windows provider

PRs welcome!

python 10 weirdness in ci-setup

code versions: ``` + conda-build 3.21.8 py38h578d9bd_0 conda-forge/linux-64 573 KB + conda-forge-pinning 2022.02.19.23.26.53 hd8ed1ab_0 conda-forge/noarch 19 KB + conda-smithy 3.17.1 pyhd8ed1ab_0 conda-forge/noarch 79 KB ```

docker_image: cos7-cuda:10.2 chosen over cos7-x86_64 for os_version: { linux_64: cos7 }

cc @isuruf I think we might need to specify the sort order explicitly somehow?

docker_image: cos7-cuda:10.2 chosen over cos7-x86_64 for os_version: { linux_64: cos7 }

I don't follow what you are saying. Can explain a bit more? In particular, this comment from @mbargull > Just omitting the sort would not work either since conda-build's output...

docker_image: cos7-cuda:10.2 chosen over cos7-x86_64 for os_version: { linux_64: cos7 }

Ah ok. So what I said above then.

Linter will now hint if the license_family fields in the about section doesn't exist

This PR recommends, not requires, `license_family`. As such I think it is fine. LGTM to merge!

Linter will now hint if the license_family fields in the about section doesn't exist

I gave my LGTM above and it still stands! Merge merge merge! Sorry I did not leave an official approval.

Linter will now hint if the license_family fields in the about section doesn't exist

Would you like me to press merge? I’m happy if you do.

Linter will now hint if the license_family fields in the about section doesn't exist

I agree with @scopatz here. We can merge this without issue and deprecate it later as we gain our footing and decide on things.