I raised this issue to CSOUND-DEV hoping to hear from John ffitch, but not a
peep.

The problem is, the way Csound 5 runs kperf means that there is no detection
of the triggered end of the scheduled instrument instance in schedwhen,
which does not insert any score event for the instrument. But schedkwhen
gets around this by inserting a new score event for the scheduled event.

I think it best to replace the implementation of schedwhen with code adapted
from schedkwhen. If I don't hear from anyone today or tomorrow, I'll just do
it.

----- Original Message ----- 
From: "Art Hunkins" 
To: "Csound Developers Discussion List" 
Sent: Tuesday, July 20, 2004 2:00 PM
Subject: [CSOUND-DEV:5066] WinCsound5


> Thanks for this new version, Michael.
>
> I can verify that FLcount works, that both FLsetVal_i and its alternate
> FLsetVali work. For my purposes, FLTK seems to be in great shape now.
>
> I note (as expected) that schedkwhen works properly but that schedwhen
> doesn't (an event starts correctly, but does not turn off).
>
> I understand that ASIO is currently being explored so that multichannel is
> possible, as well as drastically better latency (the current Windows
> latency - without any tweaking - is pretty bad on my older ME system).
>
> Art Hunkins
>

Sorry, did not know I was being asked anything.  I think schedwhen was
my attempt to introduce programmable event scheduling, and it almost
worked.  Rasmus (? we do not seem to have heard from him lately?) then
wrote schedkwhen as a response -- that is to make it work.  I rather
thought that schedwhen would slowly fade away.....
==John ffitch