Greetings!

At the last update to musixflx, I too was running a machine without a math
co-processor, so (on the inspiration of Werner Icking) several contributions
were made to a general archive containing pre-compiled executables for a
variety of  musixflx installations.  This archive can be downloaded from:

  ftp://ftp.gmd.de/music/musixtex/musixflx.zip

In particular, find the musixflx.exe compiled for DOS-16 without a math
coprocessor.  That should run on ANY PC.  I know by experience that it runs
on an old XT.

Another alternative (which once worked for me) is to download a coprocessor
emulater (EM87) and load that before running musixflx.  (Such things can be
found by searching http://www.simtel.net/simtel.net/ (The World-Wide Simtel
Archive).  One SHAREWARE example is:

  http://www.simtel.net/pub/simtelnet/msdos/emulate/em87v1_3.zip

Now, with that problem handled, be aware (encouraged?, ... it's not your
fault) that the Borland C++ compiler
(http://www.borland.com/bcppbuilder/freecompiler/cppc55steps.html) cannot
compile the musixflx.c code... This compiler is great for C++ Code, but has
a notorious "unstability" toward compiling pure C code under Windows 95.
(Borland has not admitted there is a problem yet, although the chatter about
it has been consistent and frequent.  See the newsgroup:
borland.public.cpp.commandlinetools).

Another Free compiler is the older Borland Turbo C 2.01, which can be
obtained from:

  http://community.borland.com/museum/borland/software/TC201.zip

This is about as "PURE DOS" as you can go... The problem is that it is
limited by the old 64K data models... To compile musixflx.c successfully you
have to declare all the data arrays as External and put them in another
separate C module, then compile it as two C files, linked together.  This
allows Turbo C to assign the data to its own area in the memory map.  (Hint
thanks to Dirk Laurie.)

If you really want a nice FREE C Compiler (cannot do C++), you might also
try LCC-WIN:

  ftp://ftp.cs.virginia.edu/pub/lcc-win32/lccwin32.exe

This one can compile musixflx.c, too, but has an add-on Fortran to C package
available, that some have used successfully to recompile PMX as a C program.
(This is a very capable package for budding Windows developers using C.)

At last, there is the old standby, DJGPP, the MSDOS Port of the GNU C
compiler...

   http://www.delorie.com/djgpp/

However, this installation is really a major commitment, with lots of
notorious setup problems, and it only makes DOS programs that run under on
386 or better anyway.

Finally, I find it curious that any version of Musixflx.exe does not run
under Windows 95.  The problems we had in the past were typical when we
tried to run version of Musixflx compiled for Extended Memory models such as
Windows 95 on DOS based machines.  I am wondering if you (Juraj) may still
have some other problem with the Windows installation.. (I hate answers like
that :-)).... Anyway, the DOS-16 version in the archives should certainly
work for you... That is about as old as it goes.

Hope this helped.

Joel Hunsberger
[EMAIL PROTECTED]

-----Original Message-----
From: taupin <[EMAIL PROTECTED]>
To: mutex list <[EMAIL PROTECTED]>
Date: Thursday, March 23, 2000 5:17 AM
Subject: [Fwd: musixflx]


>Has somebody a version of musixflx.exe running on a 486 with NO
>math-coprocessor?
>
>[EMAIL PROTECTED] wrote:
>>
>> Dear Mr. Tauplin
>>
>> I have downloaded your (I hope the newest) version of MusiXTeX T.98. I
worked
>> with an older version of MusicTeX from April 1994 untill now. So I was in
fact
>> surprised, when I had found the new version on Internet. The huge amount
of
>> possibilities offerd by the MusiXTeX is amazing. Namely, the sloping slur
archs
>> and some new macros to help a typesetter to write a text under choir
scores.
>> Everithing is good and I am satisfied. BUT...
>> Due to the absence of the "glue" macro \temps you decided to calculate
the
>> right measures of bars by a special program "musixflx.exe". The program I
>> downloaded doesn't want to run in my computer. I have got an older 486
without
>> any numerical coprocessor and your program writes down the error message:
"Can't
>> find numerical coprocessor." And is terminated. So I tried to compile the
>> source code of the program by using a Borland C++ compiler. I am not too
skilly
>> in programming, so that is my problem. The compiler is not able to
compile it,
>> because there are some function error messages in the message window. I
don't
>> know, what to do.
>>
>> In according to my problems alredy described, I would like to ask you to
send me
>> either some helpfull advice or, and it would be better, the programm
copiled
>> without using a numerical coprocessor. I work under MS Windows 95.
>>
>> I would be very glad if I had a complete and working MusiXTeX.
>>
>> Yours sincerely
>>
>> Juraj Chlpik
>
>--
>
>------------------------------------------------------------------------
>  Daniel Taupin, Physique des Solides, Univ. Paris-Sud, 91405 ORSAY
>  E-mail= mailto:[EMAIL PROTECTED]
>  T�l: (33)1.69.15.60.79, Fax: (33)1.69.15.60.86, home/fax:
>(33)1.60.10.26.44
>
>------------------------------------------------------------------------
>

Reply via email to