[freespace2] freespace2 OSX audio problems

[Jonas Laurio]

Ok thanks. I reset the file to the original and started over again. 
There are two lines in audiostr.cpp that I had to replace the 
errorcheck with geterror. The first I said before was line 694, and I 
replaced that one, compiled, ran binary, checked log because same 
problems w/ no voice or music in specified parts... log reported err on 
line 682 - right after the genbuffers _ not so good I'm thinking, but I 
did the same for the * of it and now the log spews gobs of one specific 
line in ds.cpp:
STUB: ds3d_update_listener at src/sound/ds.cpp, line 3825, thread 905

Yeah I still do get that corruption at the end of the gl strings output 
in the log. It must be a bug in the program like you said, glxinfo 
doesn't output garbage like that :)

Thanks for the update file in a regular tarball, I appreciate it!

-d

On Apr 2, 2005, at 5:24 AM, Taylor Richards wrote:

> On Sat, 2005-04-02 at 02:23 -0500, Jonas Laurio wrote:
>> Commenting out the 'offending' lines in audiostr.cpp does not help,
>> (more show up after the initial ones are removed). After I do that,
>> then all sorts of openal error check lines in src/sound/ds.cpp fly off
>> about the data they get (because they don't get it since those lines
>> are gone in the other >?).
>
> Ah, yeah sorry.  When you comment out the OpenAL_ErrorCheck() lines you
> need to put "alGetError();" there instead to clear the old error
> message.  Otherwise the previous error shows up on the next error check
> which isn't what you want since they'll just pile up.
>
>> Anyhow the ones I noticed that initially complained starting the
>> problem are the ones using the m_al_multichannel_id. Can you test it 
>> on
>> a NON_multichannel setup over there, or does this even matter? If
>> openal returns multichannel is not available could this be causing the
>> issue, or am I just overthinking this...
>
> m_al_multichannel_id is just the OpenAL source id.  I'm already going 
> to
> change the name to make it a little more intuitive since I had to look
> at that a couple extra times as well.  It should be ok as is though,
> just make sure to clear the old errors like I mention above and try
> again.  Hopefully you'll get better results now.
>
> And just to note that I seem to be getting some strange error values 
> too
> on a few commands with OpenAL CVS.  The command looks to get executed
> correctly but still throws out an error like it didn't.  I haven't
> looked at this any further though.  You may be seeing a variation of 
> the
> same thing.
>
> Oh the OpenGL extension list thing I mentioned is likely a bug in the
> code.  I just noticed, and fixed, something similar in FS2_Open where 
> it
> reads past the string into random memory if the last character is also 
> a
> token (ie. " ").  When the extension string has a trailing space it
> spits out garbage.  Easy fix though and it will be going in soon.  If
> you still keep getting the garbage in the log then let me know since
> that's not a good thing to leave messed up like that.
>
>
> Taylor
>
> -- 
> Taylor Richards <taylor at icculus.org>
>