From: kkuzba{at}centurytel.net
To: c_echo{at}yahoogroups.com

* Author: Rob Swindell

RS> DMB>  I don't see what the fascination is with cramming
RS> DMB>  everything into a single statement.
RS> DMB>  C just isn't built properly for that.

RS>  I was simply answering a question about how to do an
RS>  assignment/test in one expression and avoid warnings.

    Certainly.  It will sometimes fail, though, even when the
 compiler allows it and it generates no warnings.  I had a bit
 of a time in Borland C++ 3.1 with assignments and compares in
 a single statement.  You would think that the order of execution
 would allow it, but it didn't, so I had to start doing the calls
 and assignments on one line, and then perform the comparison on
 the next line all by itself.  If I recall correctly, I was trying
 to open a file and test the result all in one statement, but it
 kept giving me a negative result on it.  The same thing was going
 on in Quick C 2.5 with dynamic memory allocation, I think.  That
 was fun to play around with.  The code all looked perfect and the
 executable compiled with no errors or warnings, but I couldn't
 allocate memory or open a file.  It was quite a stumbling block.

>  kkuzba{at}centurytel.net   http://home.centurytel.net/kkuzba
>  'A nasty drop, I'll bet,' said Sam.

--- SoupGate-Win32 v1.05