PDA

View Full Version : prioritizing softsqueeze on Fedora



Ben Klaas
2005-01-22, 13:29
hi all-

I've just been playing around with running softsqueeze on my FC3
machine. It plays fine, but whenever I do anything mildly processor
intensive like loading images into a web browser, the music skips.

I had previously had this trouble with xmms on this machine and found I
could get around it by writing a hack to launch xmms from a perl script
that automatically "niced" the process to give it kernel priority. Has
anyone experimented with nicing softsqueeze/slimserver? Should I be
renicing softsqueeze or slimserver (they are running on the same
machine)? I've experimented with both and neither seemed to work (I
could be doing something wrong). Also, is there another way to
workaround this issue?

cheers,
#!/ben

Robin Bowes
2005-01-22, 13:59
Ben Klaas wrote:
> hi all-
>
> I've just been playing around with running softsqueeze on my FC3
> machine. It plays fine, but whenever I do anything mildly processor
> intensive like loading images into a web browser, the music skips.
>
> I had previously had this trouble with xmms on this machine and found I
> could get around it by writing a hack to launch xmms from a perl script
> that automatically "niced" the process to give it kernel priority. Has
> anyone experimented with nicing softsqueeze/slimserver? Should I be
> renicing softsqueeze or slimserver (they are running on the same
> machine)? I've experimented with both and neither seemed to work (I
> could be doing something wrong). Also, is there another way to
> workaround this issue?

Ben,

I don't think it will make any difference - the code streaming music is
in the same thread as music scanning / web server, etc. so giving it
higher priority won't help.

If you're running slimserver from the /etc/init.d/slimserver script then
you could add a different nice value immediately after the "daemon"
line, e.g.:

daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS

R.
--
http://robinbowes.com

Chris O'Shannassy
2005-01-22, 15:02
Robin Bowes wrote:

> Ben Klaas wrote:
>
>> hi all-
>>
>> I've just been playing around with running softsqueeze on my FC3
>> machine. It plays fine, but whenever I do anything mildly processor
>> intensive like loading images into a web browser, the music skips.
>>
>> I had previously had this trouble with xmms on this machine and found
>> I could get around it by writing a hack to launch xmms from a perl
>> script that automatically "niced" the process to give it kernel
>> priority. Has anyone experimented with nicing softsqueeze/slimserver?
>> Should I be renicing softsqueeze or slimserver (they are running on
>> the same machine)? I've experimented with both and neither seemed to
>> work (I could be doing something wrong). Also, is there another way
>> to workaround this issue?
>
>
> Ben,
>
> I don't think it will make any difference - the code streaming music
> is in the same thread as music scanning / web server, etc. so giving
> it higher priority won't help.
>
> If you're running slimserver from the /etc/init.d/slimserver script
> then you could add a different nice value immediately after the
> "daemon" line, e.g.:
>
> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>
> R.

Hi;

I used to have the same problem, on my setup there's an
"/etc/conf.d/slimserver " file where some slimserver daemon stuff gets
configured, look for the "SLIMSERVER_OPTS" line and add the priority...
(mine cut&pasted below) It should definately help, (I don't have any
issues at all anymore) though I guess it depends how fast your machine is.

SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
--logfile=$SLIMSERVER_LOG --priority=-5

(Negative values mean higher priority too btw, making it positive
will make things worse) The only downside is that it'll do the library
rescans at a higher priority as well, which you don't want. Maybe
that's something the software guys could look into?

Chris

Ben Klaas
2005-01-22, 16:50
Chris, Robin,

thanks much...spot on, as usual from this list.

I'll have to figure out a workaround for the scanning taking high
priority. Isn't it possible to configure slimserver to only rescan in
the middle of the night? That'd do it as far as I'm concerned...

thanks again for the rapid response, team.
#!/ben

Chris O'Shannassy wrote:
> Robin Bowes wrote:
>
>> Ben Klaas wrote:
>>
>>> hi all-
>>>
>>> I've just been playing around with running softsqueeze on my FC3
>>> machine. It plays fine, but whenever I do anything mildly processor
>>> intensive like loading images into a web browser, the music skips.
>>>
>>> I had previously had this trouble with xmms on this machine and found
>>> I could get around it by writing a hack to launch xmms from a perl
>>> script that automatically "niced" the process to give it kernel
>>> priority. Has anyone experimented with nicing softsqueeze/slimserver?
>>> Should I be renicing softsqueeze or slimserver (they are running on
>>> the same machine)? I've experimented with both and neither seemed to
>>> work (I could be doing something wrong). Also, is there another way
>>> to workaround this issue?
>>
>>
>>
>> Ben,
>>
>> I don't think it will make any difference - the code streaming music
>> is in the same thread as music scanning / web server, etc. so giving
>> it higher priority won't help.
>>
>> If you're running slimserver from the /etc/init.d/slimserver script
>> then you could add a different nice value immediately after the
>> "daemon" line, e.g.:
>>
>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>
>> R.
>
>
> Hi;
>
> I used to have the same problem, on my setup there's an
> "/etc/conf.d/slimserver " file where some slimserver daemon stuff gets
> configured, look for the "SLIMSERVER_OPTS" line and add the priority...
> (mine cut&pasted below) It should definately help, (I don't have any
> issues at all anymore) though I guess it depends how fast your machine is.
>
> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
> --logfile=$SLIMSERVER_LOG --priority=-5
>
> (Negative values mean higher priority too btw, making it positive
> will make things worse) The only downside is that it'll do the library
> rescans at a higher priority as well, which you don't want. Maybe
> that's something the software guys could look into?
>
> Chris
>

Robin Bowes
2005-01-23, 04:41
Ben Klaas wrote:
> Chris, Robin,
>
> thanks much...spot on, as usual from this list.
>
> I'll have to figure out a workaround for the scanning taking high
> priority. Isn't it possible to configure slimserver to only rescan in
> the middle of the night? That'd do it as far as I'm concerned...
>
> thanks again for the rapid response, team.
> #!/ben

Ben,

This is actually one of my hobby horses - I recognised some time ago
that it would require an overhaul of the slimserver architecture to
solve this problem. Sure, the DB backend will improve things slightly,
but to really crack this issue will require breaking each task out into
a separate process - either through use of threads, or separate programs.

I've heard some rumours that the slimdevices team are looking at this,
but not heard anything official yet.

R.
--
http://robinbowes.com

dean
2005-01-23, 15:57
The groundwork for this has been completed by Dan Sully. SlimServer
6.0 can run as two processes, one that doesn't do any scanning and one
that only does scanning. This hasn't been rolled into a standard
configuration, but it will likely be before we release 6.0.


On Jan 23, 2005, at 3:41 AM, Robin Bowes wrote:

> Ben Klaas wrote:
>> Chris, Robin,
>> thanks much...spot on, as usual from this list.
>> I'll have to figure out a workaround for the scanning taking high
>> priority. Isn't it possible to configure slimserver to only rescan in
>> the middle of the night? That'd do it as far as I'm concerned...
>> thanks again for the rapid response, team.
>> #!/ben
>
> Ben,
>
> This is actually one of my hobby horses - I recognised some time ago
> that it would require an overhaul of the slimserver architecture to
> solve this problem. Sure, the DB backend will improve things slightly,
> but to really crack this issue will require breaking each task out
> into a separate process - either through use of threads, or separate
> programs.
>
> I've heard some rumours that the slimdevices team are looking at this,
> but not heard anything official yet.
>
> R.
> --
> http://robinbowes.com
>
>

Marc Sherman
2005-01-24, 06:13
dean blackketter wrote:
> The groundwork for this has been completed by Dan Sully. SlimServer 6.0
> can run as two processes, one that doesn't do any scanning and one that
> only does scanning. This hasn't been rolled into a standard
> configuration, but it will likely be before we release 6.0.

How about web serving? Has that been pulled out into a seperate process
as well? Really, I think the line should be drawn at "one process that
only does _streaming_, and one or more processes that do everything
else". That way, the streaming can be run with a negative nice level,
and everything else can take a back seat.

Regardless, that's great news to hear that scanning's been pulled out
into a separate process.

- Marc

Ben Klaas
2005-01-24, 13:01
rats, I spoke too soon.

I reniced slimserver to -11 and, just for good measure, did the same for
the java softsqueeze process. I checked top to make sure both processes
were running at the new priority. The music cuts out whenever any other
kernel intensive process happens. This, to me, doesn't make sense with
the renicing of the processes, but my ears are not deceiving me.

I'm about to scrap the idea and go back to streaming through xmms, which
is fine. I just thought softsqueeze seemed like a cool way to do it,
plus it would force me to learn navigating the squeezebox remote better.
Alas, my anti-java bias grows...

#!/ben

Ben Klaas wrote:
> Chris, Robin,
>
> thanks much...spot on, as usual from this list.
>
> I'll have to figure out a workaround for the scanning taking high
> priority. Isn't it possible to configure slimserver to only rescan in
> the middle of the night? That'd do it as far as I'm concerned...
>
> thanks again for the rapid response, team.
> #!/ben
>
> Chris O'Shannassy wrote:
>
>> Robin Bowes wrote:
>>
>>> Ben Klaas wrote:
>>>
>>>> hi all-
>>>>
>>>> I've just been playing around with running softsqueeze on my FC3
>>>> machine. It plays fine, but whenever I do anything mildly processor
>>>> intensive like loading images into a web browser, the music skips.
>>>>
>>>> I had previously had this trouble with xmms on this machine and
>>>> found I could get around it by writing a hack to launch xmms from a
>>>> perl script that automatically "niced" the process to give it kernel
>>>> priority. Has anyone experimented with nicing
>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>> slimserver (they are running on the same machine)? I've experimented
>>>> with both and neither seemed to work (I could be doing something
>>>> wrong). Also, is there another way to workaround this issue?
>>>
>>>
>>>
>>>
>>> Ben,
>>>
>>> I don't think it will make any difference - the code streaming music
>>> is in the same thread as music scanning / web server, etc. so giving
>>> it higher priority won't help.
>>>
>>> If you're running slimserver from the /etc/init.d/slimserver script
>>> then you could add a different nice value immediately after the
>>> "daemon" line, e.g.:
>>>
>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>>
>>> R.
>>
>>
>>
>> Hi;
>>
>> I used to have the same problem, on my setup there's an
>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff gets
>> configured, look for the "SLIMSERVER_OPTS" line and add the
>> priority... (mine cut&pasted below) It should definately help, (I
>> don't have any issues at all anymore) though I guess it depends how
>> fast your machine is.
>>
>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>> --logfile=$SLIMSERVER_LOG --priority=-5
>>
>> (Negative values mean higher priority too btw, making it positive
>> will make things worse) The only downside is that it'll do the
>> library rescans at a higher priority as well, which you don't want.
>> Maybe that's something the software guys could look into?
>>
>> Chris
>>

rtitmuss
2005-01-24, 13:16
Hi Ben,

Sorry I am joining this discussion late, we got a new puppy in the
house last weekend and spare time has dropped to zero at the moment!!

I think that your problems are more likely to be with Softsqueeze, not
the slimserver. Could you let me know what version of Java and
Softsqueeze you are using. You may find some improvements upgrading to
Softsqueeze 1.16. Also in the Softsqueeze preferences what audio mixer
is selected, and what other options do you have? And final what spec is
your PC?

Regards,
Richard



On 24 Jan 2005, at 20:01, Ben Klaas wrote:

> rats, I spoke too soon.
>
> I reniced slimserver to -11 and, just for good measure, did the same
> for the java softsqueeze process. I checked top to make sure both
> processes were running at the new priority. The music cuts out
> whenever any other kernel intensive process happens. This, to me,
> doesn't make sense with the renicing of the processes, but my ears are
> not deceiving me.
>
> I'm about to scrap the idea and go back to streaming through xmms,
> which is fine. I just thought softsqueeze seemed like a cool way to do
> it, plus it would force me to learn navigating the squeezebox remote
> better. Alas, my anti-java bias grows...
>
> #!/ben
>
> Ben Klaas wrote:
>> Chris, Robin,
>> thanks much...spot on, as usual from this list.
>> I'll have to figure out a workaround for the scanning taking high
>> priority. Isn't it possible to configure slimserver to only rescan in
>> the middle of the night? That'd do it as far as I'm concerned...
>> thanks again for the rapid response, team.
>> #!/ben
>> Chris O'Shannassy wrote:
>>> Robin Bowes wrote:
>>>
>>>> Ben Klaas wrote:
>>>>
>>>>> hi all-
>>>>>
>>>>> I've just been playing around with running softsqueeze on my FC3
>>>>> machine. It plays fine, but whenever I do anything mildly
>>>>> processor intensive like loading images into a web browser, the
>>>>> music skips.
>>>>>
>>>>> I had previously had this trouble with xmms on this machine and
>>>>> found I could get around it by writing a hack to launch xmms from
>>>>> a perl script that automatically "niced" the process to give it
>>>>> kernel priority. Has anyone experimented with nicing
>>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>>> slimserver (they are running on the same machine)? I've
>>>>> experimented with both and neither seemed to work (I could be
>>>>> doing something wrong). Also, is there another way to workaround
>>>>> this issue?
>>>>
>>>>
>>>>
>>>>
>>>> Ben,
>>>>
>>>> I don't think it will make any difference - the code streaming
>>>> music is in the same thread as music scanning / web server, etc. so
>>>> giving it higher priority won't help.
>>>>
>>>> If you're running slimserver from the /etc/init.d/slimserver script
>>>> then you could add a different nice value immediately after the
>>>> "daemon" line, e.g.:
>>>>
>>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>>>
>>>> R.
>>>
>>>
>>>
>>> Hi;
>>>
>>> I used to have the same problem, on my setup there's an
>>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff
>>> gets configured, look for the "SLIMSERVER_OPTS" line and add the
>>> priority... (mine cut&pasted below) It should definately help, (I
>>> don't have any issues at all anymore) though I guess it depends how
>>> fast your machine is.
>>>
>>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>>> --logfile=$SLIMSERVER_LOG --priority=-5
>>>
>>> (Negative values mean higher priority too btw, making it
>>> positive will make things worse) The only downside is that it'll do
>>> the library rescans at a higher priority as well, which you don't
>>> want. Maybe that's something the software guys could look into?
>>>
>>> Chris
>>>

Ben Klaas
2005-01-24, 13:35
hi Richard-

thanks for the input. here's my info--

Softsqueeze: 1.15
Java: 1.5.0_01

Audio mixer: AudioPCI
MP3 Decoder: Java MP3 Plugin

PC: somewhat ancient Gateway PIII running at about 800mHz with 512MB
RAM. Running linux kernel 2.6.9-1.681_FC3.

aside from going to softsqueeze 1.16 (my apologies, didn't see there was
a new version...should have tried that first), anything else blaringly
obvious in my setup? Should I be using javasound for the audio mixer?

thanks again
#!/ben

Richard Titmuss wrote:
> Hi Ben,
>
> Sorry I am joining this discussion late, we got a new puppy in the house
> last weekend and spare time has dropped to zero at the moment!!
>
> I think that your problems are more likely to be with Softsqueeze, not
> the slimserver. Could you let me know what version of Java and
> Softsqueeze you are using. You may find some improvements upgrading to
> Softsqueeze 1.16. Also in the Softsqueeze preferences what audio mixer
> is selected, and what other options do you have? And final what spec is
> your PC?
>
> Regards,
> Richard
>
>
>
> On 24 Jan 2005, at 20:01, Ben Klaas wrote:
>
>> rats, I spoke too soon.
>>
>> I reniced slimserver to -11 and, just for good measure, did the same
>> for the java softsqueeze process. I checked top to make sure both
>> processes were running at the new priority. The music cuts out
>> whenever any other kernel intensive process happens. This, to me,
>> doesn't make sense with the renicing of the processes, but my ears are
>> not deceiving me.
>>
>> I'm about to scrap the idea and go back to streaming through xmms,
>> which is fine. I just thought softsqueeze seemed like a cool way to do
>> it, plus it would force me to learn navigating the squeezebox remote
>> better. Alas, my anti-java bias grows...
>>
>> #!/ben
>>
>> Ben Klaas wrote:
>>
>>> Chris, Robin,
>>> thanks much...spot on, as usual from this list.
>>> I'll have to figure out a workaround for the scanning taking high
>>> priority. Isn't it possible to configure slimserver to only rescan in
>>> the middle of the night? That'd do it as far as I'm concerned...
>>> thanks again for the rapid response, team.
>>> #!/ben
>>> Chris O'Shannassy wrote:
>>>
>>>> Robin Bowes wrote:
>>>>
>>>>> Ben Klaas wrote:
>>>>>
>>>>>> hi all-
>>>>>>
>>>>>> I've just been playing around with running softsqueeze on my FC3
>>>>>> machine. It plays fine, but whenever I do anything mildly
>>>>>> processor intensive like loading images into a web browser, the
>>>>>> music skips.
>>>>>>
>>>>>> I had previously had this trouble with xmms on this machine and
>>>>>> found I could get around it by writing a hack to launch xmms from
>>>>>> a perl script that automatically "niced" the process to give it
>>>>>> kernel priority. Has anyone experimented with nicing
>>>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>>>> slimserver (they are running on the same machine)? I've
>>>>>> experimented with both and neither seemed to work (I could be
>>>>>> doing something wrong). Also, is there another way to workaround
>>>>>> this issue?
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Ben,
>>>>>
>>>>> I don't think it will make any difference - the code streaming
>>>>> music is in the same thread as music scanning / web server, etc. so
>>>>> giving it higher priority won't help.
>>>>>
>>>>> If you're running slimserver from the /etc/init.d/slimserver script
>>>>> then you could add a different nice value immediately after the
>>>>> "daemon" line, e.g.:
>>>>>
>>>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>>>>
>>>>> R.
>>>>
>>>>
>>>>
>>>>
>>>> Hi;
>>>>
>>>> I used to have the same problem, on my setup there's an
>>>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff
>>>> gets configured, look for the "SLIMSERVER_OPTS" line and add the
>>>> priority... (mine cut&pasted below) It should definately help, (I
>>>> don't have any issues at all anymore) though I guess it depends how
>>>> fast your machine is.
>>>>
>>>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>>>> --logfile=$SLIMSERVER_LOG --priority=-5
>>>>
>>>> (Negative values mean higher priority too btw, making it
>>>> positive will make things worse) The only downside is that it'll do
>>>> the library rescans at a higher priority as well, which you don't
>>>> want. Maybe that's something the software guys could look into?
>>>>
>>>> Chris
>>>>

Daniel Cohen
2005-01-24, 13:39
I've just learned that the BBC is redesigning its Player interface,
to be launched in the next couple of days (The Independent, Media
section, Monday 24th January).

Will AlienBBC cope with this, or will it need a reworking?
--
Daniel Cohen

Ben Klaas
2005-01-24, 13:48
extra info:

downloaded softsqueeze 1.16
tried running through both java sound engine and audiopci mixer

no change in behavior...

Ben Klaas wrote:
> hi Richard-
>
> thanks for the input. here's my info--
>
> Softsqueeze: 1.15
> Java: 1.5.0_01
>
> Audio mixer: AudioPCI
> MP3 Decoder: Java MP3 Plugin
>
> PC: somewhat ancient Gateway PIII running at about 800mHz with 512MB
> RAM. Running linux kernel 2.6.9-1.681_FC3.
>
> aside from going to softsqueeze 1.16 (my apologies, didn't see there was
> a new version...should have tried that first), anything else blaringly
> obvious in my setup? Should I be using javasound for the audio mixer?
>
> thanks again
> #!/ben
>
> Richard Titmuss wrote:
>
>> Hi Ben,
>>
>> Sorry I am joining this discussion late, we got a new puppy in the
>> house last weekend and spare time has dropped to zero at the moment!!
>>
>> I think that your problems are more likely to be with Softsqueeze, not
>> the slimserver. Could you let me know what version of Java and
>> Softsqueeze you are using. You may find some improvements upgrading to
>> Softsqueeze 1.16. Also in the Softsqueeze preferences what audio mixer
>> is selected, and what other options do you have? And final what spec
>> is your PC?
>>
>> Regards,
>> Richard
>>
>>
>>
>> On 24 Jan 2005, at 20:01, Ben Klaas wrote:
>>
>>> rats, I spoke too soon.
>>>
>>> I reniced slimserver to -11 and, just for good measure, did the same
>>> for the java softsqueeze process. I checked top to make sure both
>>> processes were running at the new priority. The music cuts out
>>> whenever any other kernel intensive process happens. This, to me,
>>> doesn't make sense with the renicing of the processes, but my ears
>>> are not deceiving me.
>>>
>>> I'm about to scrap the idea and go back to streaming through xmms,
>>> which is fine. I just thought softsqueeze seemed like a cool way to
>>> do it, plus it would force me to learn navigating the squeezebox
>>> remote better. Alas, my anti-java bias grows...
>>>
>>> #!/ben
>>>
>>> Ben Klaas wrote:
>>>
>>>> Chris, Robin,
>>>> thanks much...spot on, as usual from this list.
>>>> I'll have to figure out a workaround for the scanning taking high
>>>> priority. Isn't it possible to configure slimserver to only rescan
>>>> in the middle of the night? That'd do it as far as I'm concerned...
>>>> thanks again for the rapid response, team.
>>>> #!/ben
>>>> Chris O'Shannassy wrote:
>>>>
>>>>> Robin Bowes wrote:
>>>>>
>>>>>> Ben Klaas wrote:
>>>>>>
>>>>>>> hi all-
>>>>>>>
>>>>>>> I've just been playing around with running softsqueeze on my FC3
>>>>>>> machine. It plays fine, but whenever I do anything mildly
>>>>>>> processor intensive like loading images into a web browser, the
>>>>>>> music skips.
>>>>>>>
>>>>>>> I had previously had this trouble with xmms on this machine and
>>>>>>> found I could get around it by writing a hack to launch xmms from
>>>>>>> a perl script that automatically "niced" the process to give it
>>>>>>> kernel priority. Has anyone experimented with nicing
>>>>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>>>>> slimserver (they are running on the same machine)? I've
>>>>>>> experimented with both and neither seemed to work (I could be
>>>>>>> doing something wrong). Also, is there another way to workaround
>>>>>>> this issue?
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Ben,
>>>>>>
>>>>>> I don't think it will make any difference - the code streaming
>>>>>> music is in the same thread as music scanning / web server, etc.
>>>>>> so giving it higher priority won't help.
>>>>>>
>>>>>> If you're running slimserver from the /etc/init.d/slimserver
>>>>>> script then you could add a different nice value immediately after
>>>>>> the "daemon" line, e.g.:
>>>>>>
>>>>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>>>>>
>>>>>> R.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> Hi;
>>>>>
>>>>> I used to have the same problem, on my setup there's an
>>>>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff
>>>>> gets configured, look for the "SLIMSERVER_OPTS" line and add the
>>>>> priority... (mine cut&pasted below) It should definately help, (I
>>>>> don't have any issues at all anymore) though I guess it depends how
>>>>> fast your machine is.
>>>>>
>>>>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>>>>> --logfile=$SLIMSERVER_LOG --priority=-5
>>>>>
>>>>> (Negative values mean higher priority too btw, making it
>>>>> positive will make things worse) The only downside is that it'll
>>>>> do the library rescans at a higher priority as well, which you
>>>>> don't want. Maybe that's something the software guys could look into?
>>>>>
>>>>> Chris
>>>>>

rtitmuss
2005-01-24, 14:09
OK, that all looks fine. The audiopci mixer should be better than the
java sound engine. Could you enable the 'Java Sound (Verbose)' debug in
Softsqueeze and play a few tracks. When you look back in the
Softsqueeze console it will show the buffer how full the java sound
audio buffer is before a write, you need to look for 'available=xxx' on
the end of the line. Playing back now on my mac this settles at 24%
available, e.g.

15010 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
15326 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
written=63992,open:true
15375 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
15686 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
written=63992,open:true
15731 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%

Does yours stop at a constant number, or go up and down? It would be
useful to see this log during a skin.

Also looking back at your original post I notice that you said that the
music skips when loading images into a web browser. Is this images from
any site, or from the slimserver?

Regards,
Richard




On 24 Jan 2005, at 20:48, Ben Klaas wrote:

> extra info:
>
> downloaded softsqueeze 1.16
> tried running through both java sound engine and audiopci mixer
>
> no change in behavior...
>
> Ben Klaas wrote:
>> hi Richard-
>> thanks for the input. here's my info--
>> Softsqueeze: 1.15
>> Java: 1.5.0_01
>> Audio mixer: AudioPCI
>> MP3 Decoder: Java MP3 Plugin
>> PC: somewhat ancient Gateway PIII running at about 800mHz with 512MB
>> RAM. Running linux kernel 2.6.9-1.681_FC3.
>> aside from going to softsqueeze 1.16 (my apologies, didn't see there
>> was a new version...should have tried that first), anything else
>> blaringly obvious in my setup? Should I be using javasound for the
>> audio mixer?
>> thanks again
>> #!/ben
>> Richard Titmuss wrote:
>>> Hi Ben,
>>>
>>> Sorry I am joining this discussion late, we got a new puppy in the
>>> house last weekend and spare time has dropped to zero at the
>>> moment!!
>>>
>>> I think that your problems are more likely to be with Softsqueeze,
>>> not the slimserver. Could you let me know what version of Java and
>>> Softsqueeze you are using. You may find some improvements upgrading
>>> to Softsqueeze 1.16. Also in the Softsqueeze preferences what audio
>>> mixer is selected, and what other options do you have? And final
>>> what spec is your PC?
>>>
>>> Regards,
>>> Richard
>>>
>>>
>>>
>>> On 24 Jan 2005, at 20:01, Ben Klaas wrote:
>>>
>>>> rats, I spoke too soon.
>>>>
>>>> I reniced slimserver to -11 and, just for good measure, did the
>>>> same for the java softsqueeze process. I checked top to make sure
>>>> both processes were running at the new priority. The music cuts out
>>>> whenever any other kernel intensive process happens. This, to me,
>>>> doesn't make sense with the renicing of the processes, but my ears
>>>> are not deceiving me.
>>>>
>>>> I'm about to scrap the idea and go back to streaming through xmms,
>>>> which is fine. I just thought softsqueeze seemed like a cool way to
>>>> do it, plus it would force me to learn navigating the squeezebox
>>>> remote better. Alas, my anti-java bias grows...
>>>>
>>>> #!/ben
>>>>
>>>> Ben Klaas wrote:
>>>>
>>>>> Chris, Robin,
>>>>> thanks much...spot on, as usual from this list.
>>>>> I'll have to figure out a workaround for the scanning taking high
>>>>> priority. Isn't it possible to configure slimserver to only rescan
>>>>> in the middle of the night? That'd do it as far as I'm
>>>>> concerned...
>>>>> thanks again for the rapid response, team.
>>>>> #!/ben
>>>>> Chris O'Shannassy wrote:
>>>>>
>>>>>> Robin Bowes wrote:
>>>>>>
>>>>>>> Ben Klaas wrote:
>>>>>>>
>>>>>>>> hi all-
>>>>>>>>
>>>>>>>> I've just been playing around with running softsqueeze on my
>>>>>>>> FC3 machine. It plays fine, but whenever I do anything mildly
>>>>>>>> processor intensive like loading images into a web browser, the
>>>>>>>> music skips.
>>>>>>>>
>>>>>>>> I had previously had this trouble with xmms on this machine and
>>>>>>>> found I could get around it by writing a hack to launch xmms
>>>>>>>> from a perl script that automatically "niced" the process to
>>>>>>>> give it kernel priority. Has anyone experimented with nicing
>>>>>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>>>>>> slimserver (they are running on the same machine)? I've
>>>>>>>> experimented with both and neither seemed to work (I could be
>>>>>>>> doing something wrong). Also, is there another way to
>>>>>>>> workaround this issue?
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Ben,
>>>>>>>
>>>>>>> I don't think it will make any difference - the code streaming
>>>>>>> music is in the same thread as music scanning / web server, etc.
>>>>>>> so giving it higher priority won't help.
>>>>>>>
>>>>>>> If you're running slimserver from the /etc/init.d/slimserver
>>>>>>> script then you could add a different nice value immediately
>>>>>>> after the "daemon" line, e.g.:
>>>>>>>
>>>>>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN
>>>>>>> $SLIMSERVER_ARGS
>>>>>>>
>>>>>>> R.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> Hi;
>>>>>>
>>>>>> I used to have the same problem, on my setup there's an
>>>>>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff
>>>>>> gets configured, look for the "SLIMSERVER_OPTS" line and add the
>>>>>> priority... (mine cut&pasted below) It should definately help,
>>>>>> (I don't have any issues at all anymore) though I guess it
>>>>>> depends how fast your machine is.
>>>>>>
>>>>>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>>>>>> --logfile=$SLIMSERVER_LOG --priority=-5
>>>>>>
>>>>>> (Negative values mean higher priority too btw, making it
>>>>>> positive will make things worse) The only downside is that it'll
>>>>>> do the library rescans at a higher priority as well, which you
>>>>>> don't want. Maybe that's something the software guys could look
>>>>>> into?
>>>>>>
>>>>>> Chris
>>>>>>

Ben Klaas
2005-01-24, 14:21
it seems to be quite stable at the same percentage on my system.

160631 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
160897 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
written=63992,open:true
160920 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
161273 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
written=63992,open:true

as far as the "loading images from a browser" comment, I just gave that
as an example of the type of activity that would cause softsqueeze to
skip...it will skip when anything remotely processor-intensive takes
place...reloading nytimes.com has been what I've been using to test this
issue, but I can reproduce the softsqueeze skips by just running a perl
script that does some number crunching.

Richard Titmuss wrote:
> OK, that all looks fine. The audiopci mixer should be better than the
> java sound engine. Could you enable the 'Java Sound (Verbose)' debug in
> Softsqueeze and play a few tracks. When you look back in the Softsqueeze
> console it will show the buffer how full the java sound audio buffer is
> before a write, you need to look for 'available=xxx' on the end of the
> line. Playing back now on my mac this settles at 24% available, e.g.
>
> 15010 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
> 15326 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
> written=63992,open:true
> 15375 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
> 15686 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
> written=63992,open:true
> 15731 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>
> Does yours stop at a constant number, or go up and down? It would be
> useful to see this log during a skin.
>
> Also looking back at your original post I notice that you said that the
> music skips when loading images into a web browser. Is this images from
> any site, or from the slimserver?
>
> Regards,
> Richard
>
>
>
>
> On 24 Jan 2005, at 20:48, Ben Klaas wrote:
>
>> extra info:
>>
>> downloaded softsqueeze 1.16
>> tried running through both java sound engine and audiopci mixer
>>
>> no change in behavior...
>>
>> Ben Klaas wrote:
>>
>>> hi Richard-
>>> thanks for the input. here's my info--
>>> Softsqueeze: 1.15
>>> Java: 1.5.0_01
>>> Audio mixer: AudioPCI
>>> MP3 Decoder: Java MP3 Plugin
>>> PC: somewhat ancient Gateway PIII running at about 800mHz with 512MB
>>> RAM. Running linux kernel 2.6.9-1.681_FC3.
>>> aside from going to softsqueeze 1.16 (my apologies, didn't see there
>>> was a new version...should have tried that first), anything else
>>> blaringly obvious in my setup? Should I be using javasound for the
>>> audio mixer?
>>> thanks again
>>> #!/ben
>>> Richard Titmuss wrote:
>>>
>>>> Hi Ben,
>>>>
>>>> Sorry I am joining this discussion late, we got a new puppy in the
>>>> house last weekend and spare time has dropped to zero at the moment!!
>>>>
>>>> I think that your problems are more likely to be with Softsqueeze,
>>>> not the slimserver. Could you let me know what version of Java and
>>>> Softsqueeze you are using. You may find some improvements upgrading
>>>> to Softsqueeze 1.16. Also in the Softsqueeze preferences what audio
>>>> mixer is selected, and what other options do you have? And final
>>>> what spec is your PC?
>>>>
>>>> Regards,
>>>> Richard
>>>>
>>>>
>>>>
>>>> On 24 Jan 2005, at 20:01, Ben Klaas wrote:
>>>>
>>>>> rats, I spoke too soon.
>>>>>
>>>>> I reniced slimserver to -11 and, just for good measure, did the
>>>>> same for the java softsqueeze process. I checked top to make sure
>>>>> both processes were running at the new priority. The music cuts out
>>>>> whenever any other kernel intensive process happens. This, to me,
>>>>> doesn't make sense with the renicing of the processes, but my ears
>>>>> are not deceiving me.
>>>>>
>>>>> I'm about to scrap the idea and go back to streaming through xmms,
>>>>> which is fine. I just thought softsqueeze seemed like a cool way to
>>>>> do it, plus it would force me to learn navigating the squeezebox
>>>>> remote better. Alas, my anti-java bias grows...
>>>>>
>>>>> #!/ben
>>>>>
>>>>> Ben Klaas wrote:
>>>>>
>>>>>> Chris, Robin,
>>>>>> thanks much...spot on, as usual from this list.
>>>>>> I'll have to figure out a workaround for the scanning taking high
>>>>>> priority. Isn't it possible to configure slimserver to only rescan
>>>>>> in the middle of the night? That'd do it as far as I'm concerned...
>>>>>> thanks again for the rapid response, team.
>>>>>> #!/ben
>>>>>> Chris O'Shannassy wrote:
>>>>>>
>>>>>>> Robin Bowes wrote:
>>>>>>>
>>>>>>>> Ben Klaas wrote:
>>>>>>>>
>>>>>>>>> hi all-
>>>>>>>>>
>>>>>>>>> I've just been playing around with running softsqueeze on my
>>>>>>>>> FC3 machine. It plays fine, but whenever I do anything mildly
>>>>>>>>> processor intensive like loading images into a web browser, the
>>>>>>>>> music skips.
>>>>>>>>>
>>>>>>>>> I had previously had this trouble with xmms on this machine and
>>>>>>>>> found I could get around it by writing a hack to launch xmms
>>>>>>>>> from a perl script that automatically "niced" the process to
>>>>>>>>> give it kernel priority. Has anyone experimented with nicing
>>>>>>>>> softsqueeze/slimserver? Should I be renicing softsqueeze or
>>>>>>>>> slimserver (they are running on the same machine)? I've
>>>>>>>>> experimented with both and neither seemed to work (I could be
>>>>>>>>> doing something wrong). Also, is there another way to
>>>>>>>>> workaround this issue?
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> Ben,
>>>>>>>>
>>>>>>>> I don't think it will make any difference - the code streaming
>>>>>>>> music is in the same thread as music scanning / web server, etc.
>>>>>>>> so giving it higher priority won't help.
>>>>>>>>
>>>>>>>> If you're running slimserver from the /etc/init.d/slimserver
>>>>>>>> script then you could add a different nice value immediately
>>>>>>>> after the "daemon" line, e.g.:
>>>>>>>>
>>>>>>>> daemon +5 --user $SLIMSERVER_USER $SLIMSERVER_BIN $SLIMSERVER_ARGS
>>>>>>>>
>>>>>>>> R.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Hi;
>>>>>>>
>>>>>>> I used to have the same problem, on my setup there's an
>>>>>>> "/etc/conf.d/slimserver " file where some slimserver daemon stuff
>>>>>>> gets configured, look for the "SLIMSERVER_OPTS" line and add the
>>>>>>> priority... (mine cut&pasted below) It should definately help,
>>>>>>> (I don't have any issues at all anymore) though I guess it
>>>>>>> depends how fast your machine is.
>>>>>>>
>>>>>>> SLIMSERVER_OPTS="--daemon --prefsfile=$SLIMSERVER_CFG
>>>>>>> --logfile=$SLIMSERVER_LOG --priority=-5
>>>>>>>
>>>>>>> (Negative values mean higher priority too btw, making it
>>>>>>> positive will make things worse) The only downside is that it'll
>>>>>>> do the library rescans at a higher priority as well, which you
>>>>>>> don't want. Maybe that's something the software guys could look
>>>>>>> into?
>>>>>>>
>>>>>>> Chris
>>>>>>>

rtitmuss
2005-01-24, 15:02
Hmm. To be honest I am not sure what to suggest from here (other than
using xmms again :( ). If the java sound buffer is being kept at least
75% full then really the music should not skip. On Windows I had found
a similar problem when loading the Sourceforge site, then the java
sound buffer was not constant and could get very low. This was fixed a
couple of versions ago, and it does not look like the same problem
here.

I'll have a think about this, and let you know if I come up with some
other ideas.

Regards,
Richard


On 24 Jan 2005, at 21:21, Ben Klaas wrote:

> it seems to be quite stable at the same percentage on my system.
>
> 160631 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
> 160897 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
> written=63992,open:true
> 160920 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
> 161273 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
> written=63992,open:true
>
> as far as the "loading images from a browser" comment, I just gave
> that as an example of the type of activity that would cause
> softsqueeze to skip...it will skip when anything remotely
> processor-intensive takes place...reloading nytimes.com has been what
> I've been using to test this issue, but I can reproduce the
> softsqueeze skips by just running a perl script that does some number
> crunching.
>
> Richard Titmuss wrote:
>> OK, that all looks fine. The audiopci mixer should be better than the
>> java sound engine. Could you enable the 'Java Sound (Verbose)' debug
>> in Softsqueeze and play a few tracks. When you look back in the
>> Softsqueeze console it will show the buffer how full the java sound
>> audio buffer is before a write, you need to look for 'available=xxx'
>> on the end of the line. Playing back now on my mac this settles at
>> 24% available, e.g.
>> 15010 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>> 15326 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>> written=63992,open:true
>> 15375 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>> 15686 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>> written=63992,open:true
>> 15731 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>> Does yours stop at a constant number, or go up and down? It would be
>> useful to see this log during a skin.
>> Also looking back at your original post I notice that you said that
>> the music skips when loading images into a web browser. Is this
>> images from any site, or from the slimserver?
>> Regards,
>> Richard

Ben Klaas
2005-01-24, 15:12
Richard-

no problem. I really appreciate the time you spend to help people
running softsqueeze. It's a cool app, and FWIW I run it without issue on
my windows laptop.

#!/ben

Richard Titmuss wrote:
> Hmm. To be honest I am not sure what to suggest from here (other than
> using xmms again :( ). If the java sound buffer is being kept at least
> 75% full then really the music should not skip. On Windows I had found a
> similar problem when loading the Sourceforge site, then the java sound
> buffer was not constant and could get very low. This was fixed a couple
> of versions ago, and it does not look like the same problem here.
>
> I'll have a think about this, and let you know if I come up with some
> other ideas.
>
> Regards,
> Richard
>
>
> On 24 Jan 2005, at 21:21, Ben Klaas wrote:
>
>> it seems to be quite stable at the same percentage on my system.
>>
>> 160631 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>> 160897 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
>> written=63992,open:true
>> 160920 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>> 161273 [Player Thread-2] DEBUG javasound.verbose - playFrame: bytes
>> written=63992,open:true
>>
>> as far as the "loading images from a browser" comment, I just gave
>> that as an example of the type of activity that would cause
>> softsqueeze to skip...it will skip when anything remotely
>> processor-intensive takes place...reloading nytimes.com has been what
>> I've been using to test this issue, but I can reproduce the
>> softsqueeze skips by just running a perl script that does some number
>> crunching.
>>
>> Richard Titmuss wrote:
>>
>>> OK, that all looks fine. The audiopci mixer should be better than the
>>> java sound engine. Could you enable the 'Java Sound (Verbose)' debug
>>> in Softsqueeze and play a few tracks. When you look back in the
>>> Softsqueeze console it will show the buffer how full the java sound
>>> audio buffer is before a write, you need to look for 'available=xxx'
>>> on the end of the line. Playing back now on my mac this settles at
>>> 24% available, e.g.
>>> 15010 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>>> 15326 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>>> written=63992,open:true
>>> 15375 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>>> 15686 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>>> written=63992,open:true
>>> 15731 [Player Thread-1] DEBUG javasound.verbose - playFrame: bytes
>>> read=63992 bufLen=63992 fillLen=63992 available=24.9937504529953%
>>> Does yours stop at a constant number, or go up and down? It would be
>>> useful to see this log during a skin.
>>> Also looking back at your original post I notice that you said that
>>> the music skips when loading images into a web browser. Is this
>>> images from any site, or from the slimserver?
>>> Regards,
>>> Richard
>
>
>

Jules Taplin
2005-01-24, 15:45
Hi Daniel.

We'll have to see. But... even if it's just a face-lift, it'll probably
need a re-work. Hopefully it won't be too much of a nightmare. My guess
is that at least the dedicated station parsers will still function.

Of course... if they really loved me... they'd ditch the current
approach, implement an XML-based schema for audio metadata, a
web-services interface for stream control, and throw out Real Audio, in
favour of a decent open codec. But I'm not at all sure they do. ;)


-- Jules


Daniel Cohen wrote:

> I've just learned that the BBC is redesigning its Player interface, to
> be launched in the next couple of days (The Independent, Media
> section, Monday 24th January).
>
> Will AlienBBC cope with this, or will it need a reworking?

Bill Burns
2005-01-24, 16:23
Daniel Cohen wrote:
> I've just learned that the BBC is redesigning its Player interface, to
> be launched in the next couple of days (The Independent, Media section,
> Monday 24th January).

Here's the writeup from the BBC site:
http://www.bbc.co.uk/pressoffice/pressreleases/stories/2005/01_january/24/player.shtml

This week, the BBC re-launches its internet Radio Player to make almost
every BBC Radio programme available live and on-demand for seven days
after broadcast, creating a massive, ever-changing library of music,
talk shows, dramas and documentaries.

Latest figures show more than 10 million hours of BBC radio is consumed
online per month and, from 25 January, the new Radio Player will offer
500 extra hours of programming and offer a range of new features.

The re-launch puts audiences in control of their listening, allowing
them to listen at convenient times, control their schedules and
fast-forward through programmes while exposing them to new shows - and
perhaps whole networks - they otherwise wouldn't come across.

The new Player provides all the programmes and benefits of the current
model - which makes many of the BBC's radio programmes available online
- but includes over 80 more programmes, making virtually all of the
BBC's national radio output available on-demand via the internet.

Newly available programmes include: Colin & Edith, Jo Whiley and Vernon
Kay (Radio 1); Steve Wright, Sarah Kennedy and Ken Bruce (Radio 2);
Morning on 3 and Performance on 3 (Radio 3); Ace & Invisible and Rampage
(1Xtra); Night Train (6 Music); Midday News (Five Live); and Drive with
Nikki Bedi and Breakfast with Gagan Grewal (Asian Network).

The new Radio Player also features live streaming of every one of the
BBC's English local radio and national stations: Radio Scotland and
Radio Nan Gaidheal; Radio Wales and Radio Cymru; Radio Ulster and Radio
Foyle; all 40 of the BBC's local radio stations.

The improvements follow 30,000 emails received from listeners about the
Radio Player since it first launched in June 2002 and record online
listening figures.

In addition to the wealth of new programming, the new Radio Player makes
it even easier for listeners to navigate, with lists of the most popular
programmes and links to allow listeners to click through to any
programme related to their favourite genre.

For the first time it will also be possible for listeners to stop a
programme at any point, switch off the computer and then resume
listening from that point at any time during the seven days the
programme is available.

It will also be possible to explore the other content on offer without
interrupting the listening experience.

The new Radio Player enhances the live listening experience by providing
details of what programme is on now and next for each station and is
integrated with the BBC's News, Sport and Weather Players, providing
one-click access to a huge range of audio and video from the BBC's News
and Sport divisions.

--
Bill Burns
Long Island NY USA
mailto:billb (AT) ftldesign (DOT) com

John Gorst
2005-01-25, 12:38
Richard Titmuss wrote:
> Hmm. To be honest I am not sure what to suggest from here (other than
> using xmms again :( ). If the java sound buffer is being kept at least
> 75% full then really the music should not skip. On Windows I had found a
> similar problem when loading the Sourceforge site, then the java sound
> buffer was not constant and could get very low. This was fixed a couple
> of versions ago, and it does not look like the same problem here.
>
> I'll have a think about this, and let you know if I come up with some
> other ideas.
>
> Regards,
> Richard

I get similar issues using Linux, same verson of java but the older 1.14
version of softsqueeze on Mandrake 10. Are the newer versions of
softsqueeze not being bundled with 5.4.1 nightly anymore?

I am also still getting the bug where when I start/change tracks there
is a large pause and cpu useage goes up to 100%.

rtitmuss
2005-01-25, 14:00
John Gorst wrote:

> I get similar issues using Linux, same verson of java but the older
> 1.14 version of softsqueeze on Mandrake 10. Are the newer versions of
> softsqueeze not being bundled with 5.4.1 nightly anymore?
>
> I am also still getting the bug where when I start/change tracks there
> is a large pause and cpu useage goes up to 100%.

The newest versions of Softsqueeze are not included with the 5.4.1
nightlies at the moment. The slimserver 5.4 releases are only for safe
bug fixes; the recent changes I have made to Softsqueeze I think will
improve the audio playback on lower spec systems, but this is yet to be
fully proven. It all works fine for my on my PCs, but then the older
versions did as well.

It is probably worth you trying the latest version and the Java MP3
Plugin. I don't think it will help with the music skipping, but might
fix the 100% cpu problems.

Regards,
Richard