[Openal-devel] Faster buffer model, and more random stuff

[Chris Robinson]

On Tuesday 26 February 2008 03:57:51 pm Jason Daly wrote:
> > Plans with OpenAL itself, or with the development model (or both)?
>
> Not sure what you mean by "development model", but the idea is basically
> trying to address some of the needs of the game developers they've
> talked with.  I'm not sure how much I'm supposed to say on the list.
>
> I really wish you could have made it to the meeting  :-)

By development model I mean how its developed. Since I've been around, there 
hasn't been much outward noise about what OpenAL's doing or where it's going 
(though to be fair, I didn't even know about OpenAL until just a couple years 
ago..). Maybe it's just a dry spell, as even Microsoft isn't making much of 
an issue about audio developments with Vista.. but it feels more neglected 
and/or closed than I'd expect.

> What you've done with OpenAL Soft is proof positive that OpenAL is an
> open API.  The OpenAL 1.1 spec was a _very_ collaborative effort between
> people at Creative (Garin and Dan, in particular) and the list members.
> My understanding is that there's going to be some collaboration on the
> new stuff as well.  I don't know why it's being held up, but it is.

I look forward to it. :)

> Another thing to consider is that on the list, we just have a few
> voices.  Creative deals with hundreds of game sound developers, and it's
> their voices that drive hardware sales.  Because of this, the new
> features that Creative introduces come mostly from the suggestions of
> the developers, although Creative does tend to ask for feedback before
> the new features are finalized.
>
> When we peons present a new feature, there aren't any clear sales
> associated with it, so it becomes a bit harder to convince the bean
> counters that development cost should be spent to implement such a
> feature in the native drivers.  I wouldn't say it's impossible for one
> of us to get a feature implemented by Creative, but there'd have to be
> some critical amount of interest behind it.

True enough. But it also needs to be considered that even if Creative 
themselves aren't interested in implementing an extension, they'd still have 
to acknowledge it for the simple reason of enumeration values and naming 
schemes. Since AFAIK Creative is the sole controller of what's deemed as 
an "official" extension, if an extension like my proposed format-selector was 
made, we'd have to be careful with the selected values so that it's not 
something someone else is or may end up using as well. The ARB/Kronos group 
controls things like this in OpenGL, but as OpenAL is controlled by Creative, 
it's their responsibility to take care of these things.