[pulseaudio-discuss] Segfault on startup

[Lennart Poettering]

On Mon, 22.12.08 20:51, Nix (nix at esperi.org.uk) wrote:

> On 22 Dec 2008, Lennart Poettering said:
> > Note that Ubuntu updated to libtool 2.2 but didn't really update PA's
> > lt usage accordingly. There's some major borkage in there. (i.e. the
> > "Failed to find..." msg is due to that fuckup)
> 
> What? But the needed change is only about two lines, and PA doesn't
> work at all without it. How can anyone get that wrong? An utter
> absence of testing, even of the 'does it start' variety?

The complete fix is actually not that trivial. It took me a couple of
commits to get this right. The changes from lt 1.5 to 2.2 are far from
trivial. Check git if you want to know more.

> (For a GNOME-centric distro they've really messed up integration of this
> GNOMEish component, haven't they? Not that PA is performing noticeably
> badly for me in pure KDE... I actually had it stutter once, but the
> machine was playing music over NFS *and* thrashing to death. With esd or
> arts I think I'd have been lucky to get one second's sound per minute
> under that sort of load.)

Yes, the adoption of PA in Ubuntu didn't work out that well. 

One of the biggest problems they have is that they wholeheartedly
adopted closed source drivers (nvidia, ndis, ...) and other
low-quality out-of-kernel drivers which as side effect tend to block
the CPU too often and for too long time which has a negative impact on
the scheduling latencies for programs that rely on reliable scheduling
such as PA.

But hmm, this is not the right place to discuss this and the Ubuntu
guys already know my criticism.

Lennart

-- 
Lennart Poettering                        Red Hat, Inc.
lennart [at] poettering [dot] net         ICQ# 11060553
http://0pointer.net/lennart/           GnuPG 0x1A015CC4