Atari AES 4.10 vs N.AES

[Konrad Kokoszkiewicz <draco@mi.com.pl>]

There's some annoying effect (noticed by some people as Anders and
Maurits) when you use NoBrain's MPEG audio decoder for Falcon under MiNT
and N.AES. Namely, if you leave the computer alone for some time, which is
rather short (about 1 min of less), the player stops playing. Apparently,
the AES gets stopped and doesn't return the control to the application via
evnt_multi(). The same effect is when using Background DVS player and
under TOSWIN2 - that last I've noticed when I ran gcc under TOSWIN.

What is more weird, it seems like this effect was depending on the mouse
position. I.e. if the AES stops, and you move the mouse to raise it, the
mp2 player may not stop again for very long time. I've tried this theory
and I've got an impression that the program remains active if the mouse
pointer is located over its mover (?). Contrary, if the mouse pointer is
positioned over the desktop, this effect usually comes very quickly.

Finally I've tried to use Atari GEM instead of N.AES and... Background
player behaves the same, but MPEG player worked fine for something like 8
hours while I was sitting at my 65XE coding, until I got bored with the
music and switched the Falcon off.

This "bug" looks for me like a hidden "feature" of the AES or MiNT kernel.
Does anyone have a clue about it?

Konrad M.Kokoszkiewicz
mail: draco@mi.com.pl
http://www.orient.uw.edu.pl/~conradus/
http://www.obta.uw.edu.pl/~draco/

** Ea natura multitudinis est,
** aut servit humiliter, aut superbe dominatur (Liv. XXIV,25)
*************************************************************
** U pospolstwa normalne jest, ze albo sluzy ono unizenie,
** albo bezczelnie sie panoszy.