Editorial Issue: RST_STREAM, GOAWAY and FINAL flag

[jasnell]

The current draft (-02) says, "A RST_STREAM is the last frame that an endpoint can send on a stream." ... what is not explicitly stated anywhere in the draft is that the FINAL flag for the RST_STREAM frame MUST be set. Likewise with GOAWAY

Also, what happens if an endpoint sends a RST_STREAM with stream identifier #0 ?

[jpinner]

Seems to be a side-effect of defining the FINAL flag as part of the common frame header. It is also misleading as the final flag places the stream into a half-closed state (Sec 3.4.3) but the RST_STREAM frame places the stream into the closed state (Sec 3.4.4).

Also IMHO, RST_STREAM w/ stream id 0 should be a session error.

[jasnell]

Perhaps in the definitions of the RST_STREAM and GOAWAY frames we can say that the FINAL flag has no defined meaning and MUST be ignored.

[jasnell]

And, fwiw, I agree with session error for RST_STREAM w/ stream id 0.

[martinthomson]

Let's run this one by the list.

On 25 April 2013 10:55, Jeff Pinner notifications@github.com wrote:

Seems to be a side-effect of defining the FINAL flag as part of the common
frame header. It is also misleading as the final flag places the stream
into a half-closed state (Sec 3.4.3) but the RST_STREAM frame places the
stream into the closed state (Sec 3.4.4).

Also IMHO, RST_STREAM w/ stream id 0 should be a session error.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/77#issuecomment-17029027
.

[jasnell]

Addressed by recent commits. #1527cea3b6