[Bug 758179] GstRTSPStream : Create pipeline based on enabled transport type

GStreamer (GNOME Bugzilla) bugzilla at gnome.org
Mon Nov 23 16:22:51 PST 2015


https://bugzilla.gnome.org/show_bug.cgi?id=758179

--- Comment #12 from Srimanta Panda (trollkarlen) <panda_srimanta at yahoo.co.in> ---
(In reply to Sebastian Dröge (slomo) from comment #11)
> Review of attachment 315837 [details] [review]:
> 
> Olivier, looks good to you?
> 
> Srimanta, is there a unit test already that tests all 3 variants? UDP-only,
> TCP-only, UDP & TCP.

Already test are present in the rtspserver test for these cases.

> ::: gst/rtsp-server/rtsp-stream.c
> @@ +2218,3 @@
> +
> +    if (is_udp && is_tcp) {
> +      g_object_set (priv->appsink[i], "async", FALSE, "sync", FALSE, NULL);
> 
> Why and why only on the appsink and not both? As we have queues after the
> tee now on both branches, this should not be required anymore AFAIU and
> might even cause problems.

Only put the async and sync property on appsink, because by during the creation
of the udpsink it is already set to FALSE for async and sync property. 

> @@ +2274,3 @@
> +       * sink used for RTP data, not the RTCP data. */
> +      if (i == 1)
> +        g_object_set (priv->appsink[i], "async", FALSE, "sync", FALSE,
> NULL);
> 
> The comment is wrong I think? This i==1 is the RTCP sink, i==0 is the RTP
> sink. For RTCP it makes sense to disable both (we should do that for RTCP!
> But also for the udpsink)

fixed this comment and uploaded a new patch.

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


More information about the gstreamer-bugs mailing list