Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!sdd.hp.com!spool.mu.edu!snorkelwacker.mit.edu!bloom-beacon!eru!hagbard!sunic!isgate!krafla!tryggvie
From: tryggvie@rhi.hi.is (Tryggvi Edwald)
Newsgroups: comp.os.msdos.programmer
Subject: Question on TC++ and Eckel's article on multitasking in "Comp. Lang."
Message-ID: <3028@krafla.rhi.hi.is>
Date: 7 Apr 91 22:26:57 GMT
Organization: University of Iceland
Lines: 28



Hi.

In "Computer Language", vol. 8, number 2 (February 1991) there is an article
on Multitasking, "Taking Control to Task" by Bruce Eckel.

I typed the whole thing in, in order to study multitasking and C++,
killing two birds with one stone, so to speak (not subjects to be mixed
carelessly for beginners like me-).

The code contains at least two relatively obvious typographical errors; 
near the end of tasker.hpp an ":" has become ";" in the definition of 
scheduler, and in tasker.cpp "new char" has become "newchar".

I found that the program doesn't work when compiled with the Large or Huge
memory models (the longjmp doesn't seem to longjump), while a Small model
compile produces a plausible run.  (I left the Overlay switch OFF).  

What I would like to know, is this:  WHY does the code behave so differently
depending on the memory model?  Is this something so obvious that I shouldn't
ask, and does not merit a warning in the article?

I would appreciate any explanation of this, if anyone cares to give me a slice
of his time.  Thanks.

Tryggvi Edwald, te@ejs.is, ICELAND
If one doesn't ask, how can one learn? - read the manual? Preposterous thought!