Re: Porting Audio Games To HTML5/Chrome
This should probably be in development, but I've done some looking into this and happened to glance at general for a change.
The short answer is no, but soon.
The long answer:
Webaudio, itself, only exists fully in Chrome and is still a moving specification to my knowledge-the HRTF panning capabilities have been in Firefox for maybe a couple months, MDN is filled with "this is deprecated in the latest spec" warnings and tons of incompleteness, etc. Porting specifically for Chrome may be possible, but is not worth any developer's time: the number of blind people actually using Chrome as a primary, secondary, or even tertiary browser is quite small, and this would be a significant effort-especially because you'd be effectively downloading the game every single time you played it, which necessitates some clever coding and a server to make it not have a minute or more start-up time for lots of people. Most Webaudio demos do not work
outside Chrome; it is incredibly hit or miss as to whether a sample is going to actually do what the author says, usually failing by doing nothing at all (at least in my experience).
As for Webaudio as a hypothetically complete and stable API, though, it actually looks nice if minimal. The shortcoming is that it doesn't actually provide anything more advanced than a basic biquad filter. Admittedly, that's a building block of everything more advanced, but the point stands-you still need to code everything more advanced. The limit of the environmental capabilities without extra middleware is limited to something called convolution reverb, which requires pre-recorded data that few here besides myself could make an attempt at without a very detailed tutorial or some higher end software. The irony is that this group is probably not the programmers, it's probably the musicians who own something like Soundforge.
But Webaudio isn't all b
ad. The interface itself is quite nice and, once every browser offers it, I suspect the ecosystem will explode. People are rightly complaining about bugs at the moment, sure, but I'm actually moving libaudioverse to basically match their spec as closely as possible. The interface itself is clean even if the implementations tend to be lacking right now, and someone just has to basically do Libaudioverse, the web version. This will not happen from me, but it will happen-it's actually pretty easy, probably a hundred times easier than Libaudioverse or a similar effort-and there's enough sighted people that care.
The only open question I have about Webaudio at the moment is to do with performance-some of the stuff I know about _javascript_ and how these architectures tend to work is raising red flags. But I have no excuse for this question, as I can profile easily enough.
As for the bigger issue, that is interface. In theory, mov
ing to the web gives a lot. In practice, Aria will probably be a horribly broken mess for at least 2 more years. When Aria actually starts fully working, however, we're looking at desktop app style UI in your browser if you use it properly. This means that you could do Shades of Doom, but actually build all the menus and everything in such a way that you can use spelling functionality, etc. Perhaps the biggest gain we will see here is that HTML tables work, writing _javascript_ code to move the focus with keystrokes is easy, there are all those wonderful form controls, and consequently you have a very convenient way to make level editors and mapping stuff-not to mention using them for displaying stats about inventory items, including allowing sorting. But at the moment Aria works very inconsistently across the 20 or so possible combinations-in some cases not at all-and so we can't yet gain anything from it.
if I had to put a time estimate o
n this being technically possible, it would be about 6 months ago. If I had to put a time estimate on this being worthwhile in that it would actually reach and be usable for a significant number of users without a great deal of technical knowledge while actually providing improvements over the current approach, 3 to 5 years. A lot of that is needing screen reader improvements, and NVDA is the only one that actually moves super quickly on Aria bugs as far as I can see.
_______________________________________________ Audiogames-reflector mailing list Audiogames-reflector@sabahattin-gucukoglu.com https://sabahattin-gucukoglu.com/cgi-bin/mailman/listinfo/audiogames-reflector
