I encountered, but have not completely isolated, a problem which seems to
indicate that isExtensible is not sufficient for purpose of detecting
whether a hidden record can be added to a key of a WeakMap. Making the
following change solved the problem I encountered:

https://github.com/drses/weak-map/commit/7a89a00bc11b50626db0bdc824c32cd83e5d79e0

However, I have not managed to create a test that reproduces the problem on
my own machine, with the same versions of Node.js and npm as are seen on
TravisCI where I encountered the problem. The problem may arise only if the
weak map shim is executed multiple times, which can occur and is
furthermore expected with Node.js and npm when there are version mismatches
between dependencies.

On an unrelated note, WeakMap in Subversion is now coupled with SES’s
problem/fix system, which are changes I can’t bring to bear with a stand
alone WeakMap shim. The only viable solution may be to manually port all
subsequent patches. Alternately, it might be necessary for WeakMap to
entrain all of SES and just pick out the WeakMap part at run-time, which
limits its viability. However, I’m keenly aware these are issues that are
probably out of scope for the Caja project.

Kris Kowal

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"Google Caja Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to