DN> On Thursday, 96/09/26, Leslie Rhorer wrote to David Noon about "Named
 DN> Pipes" as follows:

 LR>         For the VDM, you mean?  My source code isn't C.

 DN> Hi Les,

 DN> I take it you are a fellow assembler bigot?

        A bit, I suppose.  I do usually prefer Assembler, if for no other
reason than what I usually wind up writing is at least half interrupt
handlers.  However, in this case, no, it isn't.  This was such a
"canned" piece of work that believe it or not,  I used BASIC. 
Power BASIC, to be exact.  I may port to C, or C++/2, but not just now.

 DN> In that case it is just an
 DN> INT to wait or post an event semaphore.

        'Shouldn't be a problem, in any case.

LR>  DN> You should also be able to post a semaphore from 
 DN> a VDM, as well as LR>  DN> wait for one. I'll code both
waiting and posting from a VDM.  LR> LR>         This shouldn't
require posting a semaphore 
 DN> from the VDM, actually.

 DN> So the VDM doesn't need to tell the OS/2 session when it's done.

        No.  The only thing which might require some communication that
direction is the entry point if (at a later time) I want to get the two to
start at a particular point in the music, but that can easily be handled by
a command line arg.  No synch required, except of course the regular
"START" flag from the OS/2 app.


LR>  DN> My e-mail address is dwnoon{at}ibm.net. UUencoded stuff is
fine.  LR> LR>         Great.  Do you wish to see a copy of my
source, as LR> well?  I'm afraid it's rather bulky, and uncommented.

 DN> Uncommented assembler? Well, I'll take a look.

        Boy, you *ARE* brave.  I wouldn't dare ask anyone to hash through
uncommented assembler unless it was less than 100 lines.  Even then...

        In any case, the source code, plus the source for the OS/2 music
player, plus the DOS executable are on their way this very second. 
Actually, I'm sure you'll get them well before you see this.  The source
code for the BASIC program is a bit obtuse, and not very streamlined (this
was a quick hack), but it is also straightforward and quite linear.  Even
without commenting, it shouldn't give any fits.
                                                        Les

e-mail: lrhorer{at}fibrcom.com

--- EZPoint V2.2