<br><br>On Saturday, June 21, 2014, Thomas Mashos <<a href="mailto:thomas@mashos.com">thomas@mashos.com</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Forgive my ignorance, but then how did this issue make it into an update?<br><br>
</blockquote><div><br></div><div>Not sure I understand your question here. </div><div><br></div><div>While fixing the problem of being unable to play a particular video file (first mentioned here, on this mailing list) I had to make change to the ringbuffer class. </div>
<div><br></div><div>This is a change that could have significant impact on the entire system. While testing I identified an issue where live TV could take more than 10s to start.</div><div>I thought it was related to my recent change. But it turned out it wasn't. </div>
<div>As I was investigating the issue, I found a leak in the backend code. A ringbuffer instance is created, but sometimes it's not deleted. This cause a TFW thread to be created. When you quit the backend at this time you get the error mentioned above. </div>
<div><br></div><div>So I'm hoping the fixes I've pushed fix his problems. But I'm not certain, there may be more<span></span></div><div> </div>