At 11:07 +0000 97/11/18, Pablo R. Azero wrote:
>...All syncronization and communication has to be done
>through AppleEvents, and it is very complicated. Can be done, but
>it needs an effort which no one apparently is willing to spend in a
>dying OS. And, after all, we like to program in Haskell and we don't
>like all those C obscenities ;-)
The RhapOS is supposed to have full UNIX support, so the work done will
surely go down the drain. A funny thing though, is that the old MacOS has
an advantage over UNIX and Mach in real-time multimedia applications,
because the latter OS's may stop processes at an embarrassing moment.
>> > >Command-. does not seem to interrupt long IO sequences.
>
>If you use my port, Control-C works for most cases. I tryed now[1..] in
>Hans port,
>and I have to kill the process to stop the
>program.
>
>> But maybe there's a kludge we can use in the meantime.
>>
>> For example, the evaluation part of Hugs calls the macro allowBreak
>> on a regular basis to test whether the evaluation has been interrupted.
>> This is required to get round problems in DOS signal handling but isn't
>> needed on real operating systems so the macro can be set to nothing.
>>
>> On HugsForWindows (a GUI version of Hugs that runs on Windows 3.1, WIN95 and
>> Win-NT), the macro is redefined to test the event queue and dispatch any
>> pending events.
>>
>> Are either of these tricks applicable?
>
>I didn't have to make any change to get Control-C work. Probably becauseUnix
>interrupts are also emulated in Codewarrior.
Interrupts are generated also with my compiler (Symantec); the problem is
that the MacOS is not too good at picking them up. Perhaps Metrowerks have
a somewhat better implementation in this respect.
Hans Aberg
* Email: Hans Aberg <mailto:[EMAIL PROTECTED]>
* AMS member listing: <http://www.ams.org/cml/>
