W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2014

Re: Priority straw man

From: Tatsuhiro Tsujikawa <tatsuhiro.t@gmail.com>
Date: Wed, 29 Jan 2014 22:51:05 +0900
Message-ID: <CAPyZ6=JbBRSeUs=gi6gYhy+aQrX0QfOKCK68+Z=BvSDFBQ+bAw@mail.gmail.com>
To: Martin Thomson <martin.thomson@gmail.com>
Cc: HTTP Working Group <ietf-http-wg@w3.org>
http://martinthomson.github.io/drafts/priority.html#rfc.section.6.3

says:

"""
The PRIORITY frame is associated with an existing stream. If a PRIORITY
frame is received with a stream identifier of 0x0, the recipient MUST
respond with a connection error (Section
5.4.1<http://martinthomson.github.io/drafts/priority.html#ConnectionErrorHandler>)
of type PROTOCOL_ERROR<http://martinthomson.github.io/drafts/priority.html#PROTOCOL_ERROR>
.
"""

This means that if the endpoint just want to change the weight of the
priority group, it has to
issue PRIORITY frame wit the active stream ID and fill the current correct
dependency in the payload.
The stream ID and stream dependency are completely redundant in this case.

So I propose that PRIORITY can be sent to stream ID = 0. In this case,
stream dependency
field must be 0 and the receiver of this frame changes the weight of the
specified priority group and does not change the relative ordering inside
it.
Using stream ID = 0 makes sense in a mean that priority groups belong to
the connection
rather than the stream.

The valid use case is that if browser assigns priority group for each tab,
and user change the tab, then browser send PRIORITY frames with stream ID =
0 to just change the weight of their priority groups.

Sorry if this has been discussed already.

Best regards,
Tatsuhiro Tsujikawa
Received on Wednesday, 29 January 2014 13:51:57 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:14:23 UTC