[webkit-reviews] review requested: [Bug 115354] [gstreamer] Disable HTTP request "Accept-Encoding:" header field on gstreamer source element to avoid receiving the wrong size when retrieving data : [Attachment 209514] Patch

[bugzilla-daemon at webkit.org]

Andre Moreira Magalhaes <andrunko at gmail.com> has asked	for review:
Bug 115354: [gstreamer] Disable HTTP request "Accept-Encoding:" header field on
gstreamer source element to avoid receiving the wrong size when retrieving data
https://bugs.webkit.org/show_bug.cgi?id=115354

Attachment 209514: Patch
https://bugs.webkit.org/attachment.cgi?id=209514&action=review

------- Additional Comments from Andre Moreira Magalhaes <andrunko at gmail.com>
(In reply to comment #83)
> (In reply to comment #82)
> > So, the patch failed on mac-ews even with the cgi range support, so
expanding on the options we have:
> > 
> > 1 - Push this fix but add an exception for mac-ews on
platform/mac/TestExpectations
> > 2 - Push this fix but without sending "Accept-Encoding: identity" and
update the test accordingly. This would probably make the test pass on mac also
but still could bring "issues" on the gst source impl when the server replies
with compressed data when no Accept-Encoding header is received as described in
the rfc2616.
> > 3 - Push this fix but move test to some specific port (gtk?)
> > 4 - Ignore this fix
> > 5 - Ignore this fix and revert patch already applied
> 
> Ok, I checked here and with the patches for bug #116533 and bug #116534
already applied this patch should not be required anymore, so I guess the best
to do here is option 5. Either that or we pick option 2 and do not send any
Accept-Encoding header for media streams.

Ok, new patch implementing option 2 (not sending any accept-encoding header)
just in case. This should pass on mac also, lets see. Not sure we want to
include the test though, maybe this should be considered implementation
specific.

Let me know what you think.