Fabian Meumertzheim

Results 1684 comments of Fabian Meumertzheim

We currently only support the coverage modes `set` and `atomic`, with the former reporting `0/1` only and the latter reporting actual counts. You can force `atomic` mode by running tests...

@linzhp Could you measure the slowdown of `set` vs. `count` on the Uber repo? If it isn't noticeable, we could even think of making `count` the default rather than introducing...

@bedbad I'm convinced that's the right path forward (see my failed attempt at https://github.com/bazelbuild/rules_foreign_cc/pull/819), but as you remark, could be quite a bit of work. It would be great to...

I agree that the runfiles libraries are currently severely underdocumented. I have derived my knowledge mostly from the actual implementations in mainline Bazel (e.g., https://github.com/bazelbuild/bazel/blob/master/tools/java/runfiles/Runfiles.java and https://github.com/bazelbuild/bazel/blob/master/tools/cpp/runfiles/runfiles_src.h). Which of these...

Just as a side remark: This behavior of the native rules, which is inconsistent with the recommended behavior for Starlark rules, will be fixed with https://github.com/bazelbuild/bazel/pull/15052. But this PR is...

Submitted #3172 to add deprecation warnings. #3173 is a CI test run to check that removing these actions is actually safe.

The example you linked looks interesting, but I don't think I fully grasp what it does yet. What is the current behavior you would like to change and what do...

Thanks for working on this, let me know if I can help. Just one comment for now: It would be great if we could make the runfiles library available at...

Thanks for digging into this! I should be able to take over and debug the Windows issues in the coming weeks.

@sluongng I'm trying some things over at https://github.com/bazelbuild/rules_go/pull/3253. Looks like I got something wrong in my initial implementation of "runfiles in rundirs" lookup, sorry for that.