Bugzilla – Bug 10085
Track advance sometimes fails for SB1s
Last modified: 2009-07-31 10:32:16 UTC
A race condition connected with the timer for Web::HTTP:tryStreamingLater() can, in the callback for an old streaming connection, discard the timer for a new connection. Thus no streaming occurs if it had not been possible to select for read on the input stream, which could be the case when transcoding and on Windows where the pipeline_reader socket is set-up asynchronously. This should only happen: on Windows, using Transcoding and (I think) with SB1s.
Change 23983.
Since installing build 24011 I've had 2 SB1s and 1 SB2 on Random mix over night and they're all still player after 8 hours! Craig
I had to stop 2 of the players but 1 SB1 ran all weekend without a hiccup so I think we can call this fixed. Thanks Craig
Just to spoil things, I've just noticed that I'm not getting gapless playback on the SB1s but the same tracks play seamlessly on the Transporter. Craig
You only get (pseudo-)gapless playback on SB1s for MP3s and when not synced.
Well these are flacs and the player's not synced. Maybe I've just never noticed before. It's a consequence of playing my entire Pink Floyd collection all weekend to give it a good test :-) Craig
Actually, SC7.2 used to do pseudo-gapless for SB1s for wav but the algorithm was not safe and could screw up horribly if the following wav did not have the same properties (endianness/sample-rate/...) as the preceeding one.
Ok that explains it. Is it worth me raising a bug/enhancement request or is it not possible with 7.3's streaming? Craig
You could file an enhancement request but it would be pretty difficult to implement properly. I have just looked at the code and there would need to be quite a bit of refactoring to get it right. I'm not sure that we would do that for SB1s any more. Sorry.
This bug has been fixed in the 7.3.0 release version of SqueezeCenter! Please download the new version from http://www.slimdevices.com/su_downloads.html if you haven't already. If you are still experiencing this problem, feel free to reopen the bug with your new comments and we'll have another look.
Reduce number of active targets for SC