[Bug 785730] New elements: libSRT src/sink

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=785730

--- Comment #19 from Justin J. Kim <justin.joy.9to5 at gmail.com> ---
(In reply to George Kiagiadakis from comment #16)
> Wouldn't it be easier if the server/client selection was done via the URI
> instead of creating new elements?

I thought socket type per element would be easier than URI parameters.
>From the beginning, I considered a feature to switching socket type by setting
property (or parsing URL) in an element, but I felt I needed to use more
if-else to set socket options.

How about introducing a kind of wrapper bin to deploy a proper socket type
element by parsing URI parameters?

> it seems there is also the simultaneous open mode,
> which is not covered by these elements.

Yes, I wasn't sure that mode and suffered from lack of testing for server and
client connections.

(In reply to Brendan Lockhart from comment #17)
> Would it be possible to set the rank of srtserversrc to secondary?

I think it could be. This idea could be related to the above (introducing a
wrapper bin). My simple idea is lowering the ranks of source elements to
secondary or marginal, then the wrapper bin has primary rank. What do you guys
think about this idea?

> Also, from my testing it appears that srtserversink will accept and handle
> multiple client connections. Is this intended behaviour? I don't believe the
> stransmit app works in this way, but supporting multiple clients would be a
> fantastic feature :)

Yes, it's an intended implementation because I thought we can get a live stream
via srt so any client can be connected, even though stransmit doesn't
provide/support multiple clients.

(In reply to luckychou from comment #18)
> Btw,it seems that SRT did not provide compilation of Android platform
> instructions, no way to quickly test demo on Android.

Are there any tutorial or instruction of using 'libsrt' on Android? I can't
find that kind of information while quick searching. It would be a starting
point anyway.

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