This didn't end up being the cause of bug 921180, but I think we want to land this anyway.. right jgriffin?

Comment on attachment 815417 [details] [diff] [review] Patch 1.0 - use b2g-bin if it exists Review of attachment 815417 [details] [diff] [review]: ----------------------------------------------------------------- ::: scripts/b2g_desktop_unittest.py @@ +140,5 @@ > + # if the b2g-bin binary exists, use it > + if not self.binary_path.endswith('-bin'): > + self.binary_path = '%s-bin' % self.binary_path > + if not os.path.isfile(self.binary_path): > + self.binary_path = self.binary_path[:-len('-bin')] Wouldn't this be a little cleaner like: bin_path = '%s-bin' % self.binary_path if (os.path.isfile(bin_path): self.binary_path = bin_path

(In reply to Jonathan Griffin (:jgriffin) from comment #1) > Wouldn't this be a little cleaner like: > > bin_path = '%s-bin' % self.binary_path > if (os.path.isfile(bin_path): > self.binary_path = bin_path Sure, if we can count on mozharness always returning the non-bin version. I guess that is a fair assumption to make though.

You're probably right; we shouldn't assume what mozharness will pass to us, since that could change in the future.

Because of the patch in bug 921180, I think this is no longer necessary. No point in having the check in two separate places, and if it's something we always want to do, then I think it belongs in the harness anyway.