[Live-devel] Seeking

Derk-Jan Hartman hartman at videolan.org
Wed Jun 3 12:58:18 PDT 2009


On 3 jun 2009, at 20:29, Ross Finlayson wrote:

>> This relates to:
>> https://trac.videolan.org/vlc/ticket/2806
>>
>> Is it true that any seek action needs to be preceded with a PAUSE ?
>> Reading the RFC, i kinda deduce the same information as the user is  
>> stating, but to be honest, I've never seen a server that supports  
>> queue'd PLAY requests, and they all seem to not mind the "missing"  
>> PAUSE of VLC. (Although i have to see that few people will ever use  
>> this functionality)
>>
>> What is your interpretation of the standard Ross ?
>
> I don't see anything in the spec that suggests that that a seek  
> (i.e., a "PLAY" with a Range: start time) needs to be preceded by a  
> "PAUSE".  Our RTSP server implementation certainly does not require  
> this, and I don't know of any other implementation that does either.

Well http://tools.ietf.org/html/rfc2326 says this:

"PLAY requests may be pipelined (queued); a server MUST queue PLAY  
requests to be executed in order. That is, a PLAY request arriving  
while a previous PLAY request is still active is delayed until the  
first has been completed."

So if you are playing, then PLAY with starttime set, it needs to  
finish the previous PLAY first.
And PAUSE says:
A PAUSE request discards all queued PLAY requests.
So this made me think that perhaps the user is right (even though no  
one seems to do this).

DJ



More information about the live-devel mailing list