I have the ltsp "almost completely working" at this point except for sound 
on
the workstation. The server is working fine, including OSS sound and I don't
want to break that. I am not using "local apps" on the workstation. Although 
I
may choose to do this at some point, I am hoping that it is not necessary to 
get
sound working. I need to use OSS sound from the workstation in order for
certain applications to function properly. I am using the "cmpci" sound 
module.


I am using the "aRts Driver (libartsout.so) for the server XMMS output 
plugin. I
also have a number of entries in my server "mocules.conf" file that relate 
to
sound, including an "OSS/Free" portion. I have no problems with sound on the 
server.
I can play wav files from the workstation by running "esdplay" but otherwise 
no
sound  comes from the workstation's speakers. If the same user is logged 
into
the server and workstation, I get sound from the server's speakers that is
generated from the workstation If different users are logged into the 
workstation
and server, I just get an error about the "dsp" upon booting the 
workstation. I
can not get any sound generated from the workstation.

Does anyone know how to accomplish this (OSS workstation sound without 
causing
any server problems)?

Thanks,
Rob


Message: 3
Subject: Re: [Ltsp-discuss] About Sound and LTSP 3.0
From: John McCreesh <[EMAIL PROTECTED]>
To: Nick Sklav <[EMAIL PROTECTED]>
Cc: LTSP Mailing List <[EMAIL PROTECTED]>
Date: 08 Feb 2002 20:58:13 +0000

On Fri, 2002-02-08 at 09:27, Nick Sklav wrote:
>Hello, list
>
>We have setup and ltsp Lab here in our office and even got the sound
card detected on the workstations.
>
>Now my question is they still will not play sound so im assuming that
the workstation must somehow export sound so it runs it localy and not
on the server. Here lies the problem.
>
Nick

It's the other way round.

When you boot your diskless workstation, if you watch carefully, you
should see the kernel modules for your sound card loading, followed by a
message saying something like 'accepting audio input on port 16001'.

Back on the server, let's say you have xmms installed, and some .mp3s
waiting to test. Within xmms, choose 'options' -> 'preferences' ->
'audio I/O plugins'. Under 'Output Plugin', select the 'eSound Output
Plugin'. If you now click 'configure', under 'Server' you should see
'use remote host' selected, with a hostname of 'ws001.mydomain.whatever'
with a port of 16001.

So, what is happening on the host is that when the user plays an .mp3
though xmms, the output goes to the eSound plugin, which fires it off
over the network to port 16001 on your the diskless workstation
'ws001.mydomain.whatever', where it is played.

So, through the miracles of networking, twenty different people can all
run xmms on the server and still get the sound coming out of the local
workstation they are using.

You should also note there is no security on this, so a malicious user
could redirect their music through someone else's workstation...

John

_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx


_____________________________________________________________________
Ltsp-discuss mailing list.   To un-subscribe, or change prefs, goto:
      https://lists.sourceforge.net/lists/listinfo/ltsp-discuss
For additional LTSP help,   try #ltsp channel on irc.openprojects.net

Reply via email to