[MPlayer-users] Seekable http streams are not reporting stream length. Can they?

Ulion ulion2002 at gmail.com
Tue Aug 5 01:52:18 CEST 2008


Content-Length/end_pos could be an int64, so you'd better take good care of it.

2008/8/3, The Wanderer <inverseparadox at comcast.net>:
> Brian L. wrote:
>
>> I am using mplayer to play audio files over http. I was immediately
>> impressed that seeking in the media file was supported out of the box
>> using http ranges.
>>
>> Unfortunately, the http stream implementation doesn't report the
>> stream length back to mplayer, even when the stream length is known
>> based on a Content-Length header, so streams appear with
>> unknown/zero/bad lengths, like this:
>>
>> A:   6.1 (06.1) of -0.3 (unknown)  0.0% 45%
>>
>> The same file played from the local filesystem shows up like this:
>>
>> A:   2.0 (01.9) of 237.0 (03:57.0)  0.1%
>>
>> It seems like adding something to stream/http.c like this:
>>
>> char *tmp = http_get_field(http_hdr, "Content-Length");
>> if (tmp)
>> {
>>     stream->end_pos = atoi(tmp);
>> }
>>
>> in nop_streaming_start in the case 200: // OK case would cause the
>> information to be reported back and (presumably) show up in the
>> feedback on mplayer's stdout.
>>
>> Is there a reason not to do this? Can this feature be added?
>
> Write up a patch (following DOCS/tech/patches.txt) and submit it to
> mplayer-dev-eng, and see what they say. At a glance, I don't think
> there's any reason this would be rejected.
>
> --
>        The Wanderer
>
> Warning: Simply because I argue an issue does not mean I agree with any
> side of it.
>
> Secrecy is the beginning of tyranny.
> _______________________________________________
> MPlayer-users mailing list
> MPlayer-users at mplayerhq.hu
> https://lists.mplayerhq.hu/mailman/listinfo/mplayer-users
>


-- 
Ulion



More information about the MPlayer-users mailing list