From kde-devel  Mon May 31 18:06:23 2004
From: Robin Atwood <robin.atwood () attglobal ! net>
Date: Mon, 31 May 2004 18:06:23 +0000
To: kde-devel
Subject: Re: arts degradation?
Message-Id: <200405311906.43354.robin.atwood () attglobal ! net>
X-MARC-Message: https://marc.info/?l=kde-devel&m=108602696924364
MIME-Version: 1
Content-Type: multipart/mixed; boundary="--===============0778767466=="


--===============0778767466==
Content-Type: multipart/signed; protocol="application/pgp-signature";
	micalg=pgp-sha1; boundary="Boundary-02=_zQ3uAZl+6aZ96dq";
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit


--Boundary-02=_zQ3uAZl+6aZ96dq
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Monday 31 May 2004 17:32, Allan Sandfeld Jensen wrote:
> On Monday 31 May 2004 17:48, Robin Atwood wrote:
> > I have just upgraded my hardware to an ASUS P4P800SE with a 3.2GB P4. I
> > mpved the disk with my old KDE 3.3.3 system to the new machine and (aft=
er
> > fun-and-games with the kernel) got everything booted up. I installed the
> > (pay-for) OSS sound drivers that support the Intel ICH5 chipset. The
> > sound test worked fine but when I logged onto KDE, arts started issuing
> > those CPU overload messages I saw so much of when testing FC2-T3+ALSA .
> >
> > I eventually found that selecting Threaded OSS on the KCC sound hardware
> > dialog gave me sound in applications: KSCD, KMIX, Noatun and Xine all
> > work but there are no system event sounds! I then installed my old sound
> > card (ES1371) and reconfigured OSS. Exactly the same result: only TOSS
> > doesn't hang arts but there are no system sounds. Anyone throw some lig=
ht
> > on these strange results. This worked perfectly on the old machine with
> > the same software.
>
> Well the TOSS vs. OSS issue, is that the basic OSS drivers relies on the
> sound-driver to produce events when it is ready to recieve sounddata. If
> for some reason the kernel-driver is broken and just reports 'ready' all
> the time, this will result in an infinite loop (I am ready -> how much da=
ta
> can I send you? -> nothing really, just kiddin') . Instead TOSS asks at
> timed interval how much data the driver can recieve, sends that and goes
> back to sleep.
>
> The TOSS solution is by far the safest, and should really be made default.
> However the only driver which is known to be reproducible broken in this
> way is the OSS-emulation in ALSA, and here it is better to just use the
> ALSA driver.
>
> So basically: Use ALSA in the kernel and use the ALSA driver in aRts.
>
> `Allan

Hmmm, rebuild my kernel yet again with all the ALSA stuff. :-((
The weird thing is that artsplay works, I am wondering if this is a corrupt=
=20
configuration problem.
=2DRobin.
=2D-=20
=2D---------------------------------------------------------------------
Robin Atwood.

"This tag line left intentionally blank"
=2D---------------------------------------------------------------------









--Boundary-02=_zQ3uAZl+6aZ96dq
Content-Type: application/pgp-signature
Content-Description: signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQBAu3QzCUw9DP/G9zcRAqldAJ9e5BfcOxekOFqw6X3/idTovxw57wCfbujD
psFGzozDe+f8sEWEs95WP8o=
=FjEn
-----END PGP SIGNATURE-----

--Boundary-02=_zQ3uAZl+6aZ96dq--

--===============0778767466==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

--===============0778767466==--