On Tue, 30 Oct 2012 17:20:33 +0200, Simon Pieters <[email protected]> wrote:
On Fri, 26 Oct 2012 17:23:53 +0300, Anne van Kesteren <[email protected]>
wrote:
Currently encoding the query component of a URL using the document's
encoding affects all URLs with a "relative scheme" (http/ws/file/...).
Should we restrict this to http/https/file so new schemes such as
ws/wss and others will not be affected by this weird legacy quirk?
So in Opera this quirk does not apply to ws: or wss:. We have a test
case for this. I believe the spec required this at the time we
implemented it. Firefox passes the test as well, but Chrome fails it
(Chrome has the quirk). I don't know what IE does. I tentatively suggest
we go with Opera/Firefox here and limit this quirk so it does not apply
to ws: or wss:.
The above applies to what gets sent over the wire when using the
WebSocket(...) constructor. For <a href>, the results are different:
http://simon.html5.org/test/url/url-encoding.html
I don't have an opinion at this point about what to do here.
Is
that something implementors would consider following?
The parsing section of http://url.spec.whatwg.org/ has this as an open
issue for now.
cheers
--
Simon Pieters
Opera Software
