utgg wrote:
> Actually, the real reason the streams were often erroring before they
> could start was due to a bug in HLS.pm. I don't know if you are
> monitoring any of this Triode, but you may want to include this in your
> next update. Near the end of HLS.pm:
> >
Code:
--------------------
> > if (!${*$self}{'_pl_noupdate'} && time() >
${*$self}{'_pl_fetched'} + ${*$self}{'_pl_lastint'}) {
--------------------
> >
> wants to be:
> >
Code:
--------------------
> > if (length ${*$self}{'_pl_url'} && !${*$self}{'_pl_noupdate'}
&& time() > ${*$self}{'_pl_fetched'} + ${*$self}{'_pl_lastint'}) {
--------------------
> >
>
> It still doesn't seem to do any harm in reducing FETCH_CHUNKS to 1 as
> I suggested originally. It gives a much faster start-up to the streams
> - the start-up time could be a bit variable with 5 chunks. This is
> because several async http chunk fetches are launched at once and it
> is random which one arrives first - if the first chunk arrives last it
> will take longer to start than if the first chunk arrives first. And
> having several chunks being fetched simultaneously will slow the first
> one down anyway.
Thanks for this.
It works perfectly with Squeezeplug on a Raspberry Pi B. The previous
change to reduce the number of chunks to be fetched had no effect for
me, but I have kept this set to 1.
------------------------------------------------------------------------
AngloCuencano's Profile: http://forums.slimdevices.com/member.php?userid=64040
View this thread: http://forums.slimdevices.com/showthread.php?t=53229
_______________________________________________
plugins mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/plugins
