RE: [AD] Windows MIDI_autodetect bug?

["Ben Davis" <ben@xxxxxxxxxx>]

> > Maybe DIGMID should always be tacked on to the end of the list.
> If the user
> > wants to select DIGMID using the config file or the setup
> program, he/she
> > should be able to.
>
> Can't he/she?  If not, that is definitely wrong.

Last time I tried using DIGMID with a Windows program (MingW32, but probably
a pre-MSVC-compiled DLL), it reverted to OPL3 or something. The same
program, compiled in DJGPP, using the same config file, used DIGMID. This
wasn't the latest CVS though - probably WIP 3.9.32 or before.

I don't really have the time and patience to set the latest CVS version up
with MingW32 and try again. Maybe someone who's in Windows port development
at the moment could try it. Use the setup program to choose DIGMID, make
sure you have a patches file, and run the MIDI example. Try a DJGPP version
to check that the patches are OK and so on, and then try a Windows version.
If the Windows version *reverts to a different driver* (rather than
failing), please tell us!

I believe, if the digital sound fails to install, MIDI will fail rather than
revert to something else. Tell me if I'm wrong.

Ben Davis