[pulseaudio-discuss] Assertion 'usec <= ((pa_usec_t) 1000000ULL)*60ULL*60ULL' , failed at pulsecore/rtpoll.c:516

[Timo Harmonen]

Hi,

Well, that assumption is easy to make, and at least on the video
driver side it has been a good rule of thumb. And then of course
try to buy really popular hardware (here I failed).

I can try reporting this to alsa, but fixing it by myself might
be too much asked :).

One "workaround" would be to simply buy some other hardware. Is
there anywhere list of "pulseaudio friendly" hardware / alsa
drivers? It's not that clear just by looking alsa soundcard lists.

- Timo

> Lennart Poettering wrote:
> Heh, I never understood why people think that old, bitrotten code could
have less bugs than new, maintained code.
 >
> snd_pcm_avail()/snd_pcm_rewind()/snd_pcm_delay() and frends tend to be
broken on old drivers/old kernels, since PA is the first client that
actively makes use of them. Parts of APIs that aren't used are
> broken. That's true for ALSA as it is true for anything else.
>
> Also, developers don't like to focus on age old hw, so if bugs exist in
old drivers it's a hard time to convince anyone to fix
> them. Especially because commercal companies usually ask them to work on
the new drivers, not the old ones.
>
> Also, just trying to work around problems is never doing any good for
helping the bugs to get fixed properly. Please report the bug against
ALSA, and be insistive to get the bug fixed. Even better, try to fix it
yourself!
>
> Lennart
>