[vlc-devel] input_EscapeDiscontinuity() and client VLC instances

[j.zorko at att.net]

Hello, all ...

The macroblocking artifacts when a server VLC seeks on the source file while streaming to a 
client VLC are due to the client VLC not calling input_EscapeDiscontinuity() to prepare its' 
decoders for the stream gap.  The server calls it, but it does not propogate to the client.  So, i'm 
wondering what the best way to do this is:

1. Make use of the TS discontintuity indicator to trigger the client to call 
input_EscapeDiscontinuity()
2. Make another communications channel between server and client (a socket)
3. Overload some other never-used TS header information to trigger the client to make the 
function call.

We're working with the 0.6.0 codebase, and while we plan on eventually moving to 0.6.3, it's not 
really feasable right now, so i'd like to incorporate this in the context of 0.6.0 code.  Option 1 
seems like the best way, though I don't know how much 0.6.0 differs from 0.6.3 with respect to 
TS discontinuity handling.  Option 2 isn't so hot -- it doesn't scale, doesn't handle IP multicast,  
and it introduces dependencies between server and client i.e. a set-top box wouldn't know how 
to talk on this special channel.  Option 3 means doping the TS stream with our own data, which 
just sounds bad for the same reasons option 2 does.

So, anyone have any ideas, pertaining to 0.6.0 code if at all possible, on how to express 
discontinuity in the TS packets, so that the client VLC can see this discontinuity and call 
input_EscapeDiscontinuity()?  I'm hunting, but if someone already knows, cool ...

Regards,

John

--
Falling You - exploring the beauty 
of voice and sound
http://www.mp3.com/fallingyou

-- 
This is the vlc-devel mailing-list, see http://www.videolan.org/vlc/
To unsubscribe, please read http://developers.videolan.org/lists.html
If you are in trouble, please contact <postmaster at videolan.org>