[Traverso-devel] Re: some successes [was:compiling from cvs]

[Remon]

Hi Matt,

> losing track of this thread!  trying to consolidate responses here.

OK, thanks for summarizing your experiences, glad it worked out for you now! 
QJackctl behaviour under gnome indeed seems a bit weird, I've heard it from 
other people too.

Have fun using Traverso, keep in mind that the cvs version is a little bit 
unstable here and there as Nicola lovely pointed out *grmbl*, and if you have 
any questions or suggestions, feel free to post them here!

Regards,

Remon

>
> On Mon, 2007-02-05 at 17:55 +0100, Remon wrote:
> > Hi,
> >
> > > (null driver crashes too, so traverso exits).
> >
> > Does it show a message box with that information ??
> > (The Null AudioDriver stalled too, exiting application!)
> >
> > It's a fail save routine, and detects, ehm, that something is horribly
> > broken lol.
>
> that's exactly right, and I think that what was broken was not traverso,
> but some issue with the linker due to my previous botched attempts to
> modify /etc/ld.so.conf.  Anyway, undid those changes, ran ldconfig,
> deleted the old source directories before rerunning cvs and tar -xzvf,
> rebooted for good measure and recommenced compiling.  Didn't install the
> new versions, but ran them from the top source directory
> with ./traverso, after also running:
> export $LD_LIBRARY_PATH=/home/matt/src/traverso/lib/
> and found that the driver issues I reported above were gone.  yay!
>
> > I'll update the cleancompile script so that it empties the lib dir in
> > traverso, so one _really cannot_ get mixed libs!
>
> maybe this would help too, fortunately the compile time is pretty short
> so it's no huge loss to for me at least to just reload the source.
>
> anyway, wanted to say two things:
>
> -confirmed that the STATIC switch in src/libbase.pri does work, thanks
> for that!
> -switching the number of periods in the ALSA driver to 3 worked great,
> sound plays perfectly now. (well, almost perfectly.  there is an odd
> silence about 1.5 seconds into the first playback of every session, but
> that only happens once).  it really is a relief to be able to use alsa,
> esp. since qjackctl behaves so badly under my gnome (not sure whose
> fault that is, but having to hunt down the qjackctl window by closing
> down all the  other hundreds of windows on the desktop can be
> frustrating).
>
> > I've tried to make the audiobackend is stable as possible, mainly by
> > adding a simple check to see if things go like they should. There are
> > some bugs fixed in the audiobackend in cvs in respect to 0.30.0 that
> > _possibly_ could affect you. So to say it differently, if it doesn't work
> > for you now, it won't work when 0.40.0 is out too, since there is not
> > much I can do to improve things.
>
> turns out the problem was on my end entirely (see above), so that's
> solved.
>
> > Ah, one other thing, if you have a dual core (or multiprocessor system)
> > there is a crash bug in Qt 4.2 which will be hit sooner or later, which
> > is solved in Qt 4.2.3 (not out yet), one can download a snapshot, compile
> > that, and live is fine again.
> > But I admit for normal users, this is a bit of much to expect from them
> > heh. It's actually very simple to compile Qt, but oh well....
>
> well, I do have a dual-core processor, but haven't hit the bug yet,
> we'll see...
>
> > How well did the ready to run cvs binary work? It's compiled agains a Qt
> > 4.2.3 snapshot, and works just fine here :-)
>
> now that the linker issues are cleared up...  it works well, but the
> ALSA problem of course persists.  so I'm using my own binary instead.  I
> made a primitive .deb if anyone wants it (just using checkinstall); I
> have no idea whether it's of any general value.
>
> > So, as for the Qt bug, that's fixed, and we only have to wait for the
> > next Qt 4.2 incremental release.
> > As for the driver problems, it's real sad that the intel drivers still
> > seem to have so much problems, and there isn't much I can do about it,
> > you could try the position_fix=1 , I think it would be sufficient to add
> > it
> > in /etc/modprobe.d/options
> > (as: snd-hda-intel position_fix=1, but I've never tried it, don't have an
> > intel soundcard, at least, I don't use it lol)
>
> not sure if this is necessary any more and I'm hesitant to do this when
> I've got a fix in hand (don't want to break anything more globally).
> and  really I should be doing some of the stuff they PAY me for...
>
> This has all been terrifically helpful, I'm looking forward to trying
> traverso out for real now!
>
> matt
>
> > _______________________________________________
> > Traverso-devel mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/traverso-devel