> Interesting. What happens happens if the WebKit patch then fails to land > in WebKit, perhaps because some bot fails the test, a conflict, or anything > else? Is resolving that a manual affair? >
If patch lands in WPT and not in WebKit due to conflicts, that is fine. One will need to resolve the conflict manually as done for other conflicts. > > When imports are automated and much more frequent, I take it the upside is > that you never need to reapply still-being-exported patches like we do in > Blink. But what about the above case, when the change hasn't been applied > in WebKit yet? Seems like the importer instead needs to *revert* the > in-flight change? > I guess the question is if the change is made in WPT, not yet in WebKit and we are reimporting in the middle. In that case, we will probably have the new test failing, timing out or crashing. This will get fixed once the patch lands in WebKit (after conflict resolution).
_______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org https://lists.webkit.org/mailman/listinfo/webkit-dev