Hi Romain :)

Everything is looking good so far with the new cue_cut operator, we release a beta of Airtime 1.9.0 so the real test begins >:)

Martin

2011/6/16 Martin Konečný <martin.konecny@sourcefabric.org>

Ok now I see my problem. I thought cue_out was defined relative to the end of the track, not the beginning. This makes much more sense :)

I've done some small tests and everything is looking good so far! I will report back if I find any other issues.

Martin


 

On Jun 15, 2011 7:11 PM, "Romain Beauxis" <toots@rastageeks.org> wrote:
> Hi and thanks for the feedback!
>
> 2011/6/15 Martin Konečný <martin.konecny@sourcefabric.org>:
>> Hi Romain,
>> Im playing with this now. I have the following script:
>>
>>> set("server.telnet", true)
>>> set("server.telnet.port", 1234)
>>> set("log.file.path", "/home/martin/liquidsoap/ls.log")
>>> set("log.stdout", true)
>>> %include "library/pervasives.liq"
>>> s = request.queue(id="queue", conservative=true)
>>> s = cue_cut(s)
>>> s = mksafe(s)
>>> out(s)
>>
>> and I connect through telnet and issue the following command:
>>>>
>>>> queue.push
>>>> annotate:liq_cue_in="5.0",liq_cue_out="0.0":/home/martin/test.mp3
>>>>
>>>> 0
>>>>
>>>> END
>>>>
>>>> Connection closed by foreign host.
>>
>> This causes the telnet client to be disconnected immediately, and no sound
>> is issued.
>
> So, the problem was that cue_out <= cue_in, which trigered an assert
> false in a situation we did not expect when placing that assertion..
>
> I have removed the assert as I think we should allow cue_out <= cue_in
> even if this is kinda weird to do...
>
> As for the real problem, if you do not want to cue_out, you should not
> pass any cue_out or, if this is not possible, pass a very large
> value..
>
> Romain



--
Martin Konecny
Software Developer, Sourcefabric
martin.konecny@sourcefabric.org

720 Bathurst St. Suite 203
M5S 2R4, Toronto, ON, Canada
+1 (416) 892-8420 (Cell)
Skype: martin.konecny15

http://www.sourcefabric.org
http://www.twitter.com/Sourcefabric