[Bug 770935] RTP/RTSP/TCP transport RTSP TIMEOUT should start at once when tcpi_ca_state is TCP_CA_Loss

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=770935

--- Comment #3 from Dag Gullberg <opaque.se at gmail.com> ---
Yes, you are right. 

The intent is simply to leave things as they are, but allow the RTSP to start
at the earliest possible moment.

But true, if we set it to GST_RTSP_EINTR, then the BACKLOG will be exercised
and we do not want that, because if the result of the send to the socket was
OK, then the data is either in the window or in the socket buffer, and should
be treated as *handled*. 

It worked out for my disconnect test scenarios for the obvious reason.

I will contemplate these comments and get back with an updated patch.

cheers.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.