Excessive locking in Speaker::fire()

[cologic]

As far as I can tell, Speaker::fire doesn't actually modify the listeners list, and given that STLport is thread-safe for read-only stuff, I fail to undertand why there are locks around all the variants of fire(). (As opposed to addListener, removeListener, and removeAllListeners or whatever they're called, which are justifiably locked.)

[arnetheduck]

another thread could remove/add a listener which would modify the structures? That's the first thing I come to think of, but probably not the last...you're welcome to investigate further, it might as well be a remnant of old code...

[cologic]

Apparently removing the locks doesn't work too well, as there are some crashes that are pretty clearly the result of their absence in BCDC++. Oh well.