[Openal] openal future

Yvan Vander Sanden yvan at youngmusic.org
Tue Apr 28 01:04:56 PDT 2009


2009/4/27 Jason Daly <jdaly at ist.ucf.edu>:

>  If you look, though the download links do have the newest implementations
> available, including OpenAL-Soft 1.7, which was just released last month, I
> believe.

I see you're right. Apparently, the 'recent changes' page doesn't
include everything.

> There's no reason not to trust OpenAL-Soft.  It's quite simply the best
> software implementation available.  It started with Creative's original
> software implementation for Windows, to which Chris has added quite a number
> of enhancements.
>
> As to the risk, even if Chris stops working on it, the code is open-source,
> so if you ever run into any problems with it, you're free to dive in and fix
> it.

I'm familiar with the concept of open source, yes :-) But still,
projects maintained by one man are always vulnerable. Of course I am
allowed to make changes myself, but I'm just an artist with some
limited programming skills. Making a program when you can depend on
good libraries to make the things you don't understand a lot easier is
one thing, but tackling that library itself is quite another. I would
depend on the possibility that other people also want to keep using
the library and decide to work on it. I'm using Linux for almost 10
years now, and I've seen quite a lot of abandoned software in that
time.

That said, I certainly agree with you that OpenAL-Soft is a very good
implementation and Chris does deserve a lot of credit for that.

> Have you tried the latest X-Fi drivers for Linux (the 1.0 release)?  I have
> no idea why they were shipped so silently, but they seem to have addressed
> just about every problem I ever had with them.  I downloaded them, followed
> the instructions, and they just worked (even on Fedora 10 64-bit), no quirky
> installation, no clunky recompiling, no problems loading the modules.  I was
> quite surprised, and very pleased.
>
> Don't go to the old Beta site, go to the regular Creative Labs driver
> download site.  The latest driver was released November of 2008.

Interesting. I tried that driver on Ubuntu 8.10 and I did not even
recognize my card. Felt a bit cheated because I especially checked if
there were Linux drivers, before I bought it. I'll ask you some
questions about that in a personal mail if you don't mind, because
it's not really related to this list anyway.

>> Of course OpenAL as such is an open standard, but programming can't do
>> with standards alone, it needs a library, good documentation and not
>> in the least progress.
>
> What documentation are you missing?  The core spec is pretty complete.  A
> lot of the extensions could be documented better, but the big ones (EFX, and
> MCFORMATS) are pretty well complete, I think.

Nothing really :-) I expressed myself badly. I meant that an open
standard as such doesn't guaranty a continued development. But
actually this part of my question was not that relevant after all, i
think.

> I wouldn't give up on OpenAL yet.  Just out of curiosity, what other
> solutions have you considered, and why?

Well, I thought about irrKlang because I'm also using irrlicht as a
graphics engine. They are closely related so that might be a benefit.
On the other hand, IrrKlang and irrlicht make it very easy for
developers, but at a cost of what is possible. For the graphic part I
don't mind because the visuals are nothing more than an addition to
the sound in my project. But for the sound that could be a problem. I
want to have a virtual sound environment where i can alter every
aspect to do things that are not possible in reality, as opposed to
just imitate that reality.

I also looked at FMOD. It has very interesting features (like wall
reflections!), but 2 things worry me. Firstly it's not open source.
And secondly, although there are lots of examples added to the
download, there isn't an API reference or some clearly organized
information about the library. I hate to go through dozens of examples
to find what I need, every time I want to work on my program.

In that perspective OpenAL still seems the best option. I think I just
need some reassurement before I start.

Regards,

yvan vander sanden

-- 
Copyright only exists in the imagination of those who do not have any.



More information about the Openal mailing list