Re: A Blast from the Past

From: Richard Hart <rhart@ATCDG.COM>
> And why is (1) a bad idea?  It seems rather interesting to me (that's
> why I wanted to look it up in the System/36 Assembler manual).  I do
> have a friend who is a C programmer who doesn't like them either (that
> is ,(1) defines)).  It can make the code harder to follow, I suppose.
>

Mark W was taking me to task for the same thing. Here is my reply to him:

> > >>>---> private reply to you, "off list" (for now)... <---<<<
> >
> > How can you say "1) is usually a bad idea"...? On what basis?
> >
> > I have seen some really good stuff done using true macros, mainly
> > on mainframes, usually in assembler, but some even using the
> > assembler macros to generate (PUNCH) COBOL code, etc. :-)
> >
>
> on of the things that was dropped from C/C++ in defining Java
> was the preprocessor. And for good reason. It is like the GOTO
> statement, when used correctly it is very powerful and good, but
> its potential for misuse is just too great. The same problem with
> preprocessor macros, they introduce a new nomenclature
> that has to be learned first, makes maintenance harder as you
> cannot just let your "pattern recognizer" do the work for you
> subconsciously.
>
> The assembler macros tended to become "standard", e.g.
> macro-level CICS and are then no longer "user" macros.
> In this case the above argument is less strong, although
> many macros don't "look" like assembler and then mixes
> the two styles.
>
> This is a religious war, so let me stop here....
>
>