[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
>> - MIDI driver. This is based on the old Mac port driver by Ronaldo Hideki
>> Yamada; his driver was still not complete (buggy), but I fixed it and now
>> it sounds as it should. BTW, should I backport it to the old Mac port? I
>> have no way to test it though as I have only OS X installed...
>
> Yes, try to backport it if you can, having two versions of the same code side
> by side is dangerous. And don't worry about testing, the MacOS 9 port is
> defunct.
Well, it should just be a matter of changing the #include directives to
specify other files at the beginning of qtmidi.m... The driver is pure C
despite of the .m extension (which I used for consistency with the rest of
the port, in case it'll ever be needed to access global objective-C objects
from any source). I'll do it.
> Is that really a requirement? I mean, the configure machinery is an
> excessively fragile stuff and is intended to insulate you from the subtle
> variations of Unices. Now the MacOS X port runs only on MacOS X so we
> shouldn't have to care about variations here.
Hum, ok. I use to prefer configure as "configure/make/make install" is the
most common way to build stuff in unix variants, so to better please
newcomers I thought it would have been better... That and your reasons said,
and knowing the current configure mechanism is a mess, let's leave the OSX
port with a static makefile...
>> It'd also be nice if there'd be an install-framework target
>
> Adrian Sampson posted something like that about a month ago:
> http://adrian.pygmysoftware.com/XAllegTemp.tgz (about 5k)
>
> Is that what you're talking about?
Not exactly, but even that could be useful to some extent. It'll need to be
updated to use the native port though.
> P.S. (about the exunicod example): the Unicode support of the library might
> be in better shape than I initially thought on big-endian platforms. Only
> the example itself might be little-endian oriented. Would you mind doing an
> experiment? Choose one message (message_it for example :-) and swap the
> bytes in every pair of bytes (including the last chunk that contains 4
> bytes). Does the Italian message look better?
Hum, tried it and unfortunately it didn't solve the issue... Guess we'd
better reconsider all the short * references in unicode.c
--
Angelo Mottola
a.mottola@xxxxxxxxxx