Boris Zbarsky [:bzbarsky] Reporter
	Description • 7 years ago

http://searchfox.org/mozilla-central/search?q=&path=testing%2Fweb-platform%2Fmeta%2Fwebcryptoapi shows a ton of failures in various web platform tests for webcrypto.

This is popping up in marketing materials from Apple and whatnot; see things like https://twitter.com/webkit/status/888466862002106368

Some of these are fairly simple (e.g. incorrect type of exception thrown), but some are us actually not following the specs correctly (e.g. the very first one I tried at http://w3c-test.org/WebCryptoAPI/generateKey/test_failures_ECDH.https.html shows us not failing operations that should fail).

	Tim Taubert [:ttaubert] (inactive)
	Comment 1 • 7 years ago

I would love to fix those but I definitely won't find the time to do so in the next months :( They should be relatively easy fixes, but then OTOH getting WebCrypto work prioritized is hard. I tried.

	Daniel Veditz [:dveditz]
	Comment 2 • 7 years ago

JC: this is hurting us currently. I've tagged this as "backlog" for now based on Tim's comment, but you ought to determine where it's scheduled in fact.

	Tim Taubert [:ttaubert] (inactive)
	Comment 3 • 7 years ago

This isn't just about error codes, these would be somewhat easy fixes, we're missing features too. We deemed WebCrypto not important enough to spend time on for years, why does one WebKit blog post change that suddenly?

	J.C. Jones [:jcj] (he/him)
	Comment 4 • 7 years ago

I've added this to the SecEng roadmap and my team's backlog, Dan.