Re: Does anyone use indicators anymore?

James H H Lampert wrote:

Well, I did say _modern_ RPG.

FOR PITY'S SAKE! The primary elements of RPG that 
distinguish it from all other languages are (1) its source 
format, and (2) The Cycle.

You get rid of those, and why in blazes should you still 
call it RPG?

Same reason you call java "Java" ... it's a name.

I haven't used the RPG cycle in 15+ years.  The applications I develop
just aren't conducive to it.  Of course it's also been a very long time
since I've written an actual _report_ in RPG.

But occasionally, I run into problems (as with 
my earlier project involving reading an 
externally-described file into a data structure) in which 
the "ER" condition is equivalent to the "NF" condition, 
and the simplest way to keep the program from locking up 
is to stick an indicator in the appropriate column for the 
ER and NF conditions, and use it instead of %FOUND. 

Wouldn't the (e) extender and %error bif be easier to understand than
the ER indicator?

Likewise, it is not inconceivable that even 
%FOUND(<filename>) might get polluted by the time I 
actually need to know whether a record was found, and 
sticking an indicator in NF is more convenient than 
defining a variable to hold the %FOUND result.

Maybe ... but if it comes down to that, I would far prefer to have a
named boolean variable to hold that information so it's obvious what I'm
checking.

And of course, you may want to take some action depending 
on whether an arithmetic operation produced a positive, 
negative, or zero result; indicators save you from having 
to do comparisons on the result.

I would rather sacrifice a few cpu cycles for the sake of readability.

david