[camlorn]

Scandum, saying that taking away something that someone is used to is going to make them less effective is indeed true. This doesn't mean that adding it will necessarily make someone more effective, only that it made you more effective. I'd say that, in terms of Godwars2 specifically, blind people have a higher developed skill of spacial memory, for lack of a better term, than sighted peers, for the simple reason that we experience the entire world differently. I think that muds haven't changed in any meaningful way in at least 20 years, possibly even 30, and that people didn't have all these fancy status bars back then and did just fine. Even sighted players choose to play without them and there's nothing stopping a blind person from throwing together a quick sound notification--such would take me a few minutes the first time, and a few seconds after that.
Verbannon, I only just saw what you meant when i re-read the thread. Such APIs are limited to asking the screen reader to say something on behalf of the user, asking the screen reader to stop, and in one case using undocumented functions to get it to do other things. IAccessible2 may make it possible to do this in a screen reader agnostic manner via live regions, but that's hard to play with without learning a lot of the inner workings of windows first, and only NVDA to my knowledge has full support and it may not work outside of a web browser. Also, the particular feature I think might be useful there is even more finicky than just calling the screen reader directly. I'd say that with some support, namely functions to query settings and if the screen reader is talking, some callbacks, and the ability to temporarily speed up speech, something might be workable. This would only be likely to be implemented in NVDA, for the simple reason that you could go implement it yourself, and is generally a bad idea, as it begins the slippery slope of adding specific screen readers to the system requirements lists. I'd say that flushing the speech buffer is good enough that it's worked for the last 10 years, but that a better solution may exist. I just don't think that it could be too much better than what we have now, and there's no way of doing it in a screen reader independent manner. It's already a big enough problem making things accessible, as they tend to only work properly with the screen reader with which they were tested. The required APIs for this would aid the mudding community, but could very, very easily lead to numerous difficulties outside of it: nonstandard interfaces because application developers started calling directly, only working with one screen reader and not another, the list goes on for quite a bit.