Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
Fix for 794 (chunking input on Windows) #850
@cpitclaudel Thanks. I pushed a commit that shortens the docstrings.
However, following the discussion in the corresponding Emacs bug, I'm not sure whether I still like to default to chunking generally. We probably have to since we can't reasonably choose more specifically, but I do not like to degrade performance for all Windows users that aren't affect. Python, for instance, seems to be fine.
Did you check whether there's a node.js issue? I didn't find any. Should we probably report it to node.js, too?
Yeah, I'm not sure either. I don't really like this kind of bugs. I don't know if there's a way to detect this situation, either. I don't want this small windows tweak to start rippling through the codebase though, so I won't champion per-checker settings :)
I can, however, test out the performance of that thing. Maybe it's not that bad.
I didn't find anything either :/ A report would probably be in order, but I'm not sure if now is the best time. We might want to wait until things are a bit clearer on the Emacs side; then (or if things are not making progress) we can probably ping the node.js people. What do you think? We could also ask them right now.
I think it's a very reasonable timeline. Especially since the bug seems to be fixed on Emacs' side, and Emacs25 will be out soon, so people will have had time to update by then.
I'll try to do that soon. In the meantime I tried out the patch, and it seems to fix the issue!
I will; I'm waiting for more information on Eli's fix to see if this is something worth mentioning to the node.js people, or whether it was all on Emacs' side.
@cpitclaudel Given that the issue appears to have been fixed in Emacs 25 now, we can safely merge this workaround as there's a clear path for obsoleting and removing it again, namely as part of our Emacs 25.1 cleanup milestone.
Based on Eli's commit I'd adapt the condition for enabling the workaround though, in order to skip it if the issue is fixed upstream.
Jan 14, 2016
@cpitclaudel I merged the fix into
May I ask you to test the next MELPA build on Windows in a broken and a fixed Emacs version, and check whether the issue is gone, and whether
I opened #856 for us to track the removal of the workaround once we drop support for Emacs 24, in our Emacs 25 cleanup milestone.
All checks passed :)
My, that was a tricky bug.