Re: Impossible to even think about rewriting in RPG

John Yeung wrote:

If people don't get religious about code, or caught up in acronyms and
buzzwords, then it's not hard to envision sticking with non-/free RPG,
and even OPM. Equally, it's not hard to envision switching to /free,
and using ILE. I'm not embarrassed to say I have not fully embraced
ILE, and almost never use modules or /free.

I use modules all the time. And procedures and functions. And mixed language. And The Cycle, when it works for the problem at hand. (I have yet to find a use for secondary files, though; even in my Cycle programs [not all of which even have a primary file: some of them just set LR on when something non-database-related happens, like the user hitting F3], I find it more straightforward to set up non-primary files as full-procedural.)

My objection to /free is that (1) it doesn't really do much that can't be done *just as legibly* in standard ILE RPG format, and (2) it wasted time, essentially grafting a subset of PL/I onto RPG, that (at least IMHO) could have been better spent on things like porting PL/I to ILE (with list-directed and data-directed interactive I/O fully implemented), or making certain SQL-only fields and constructs available in native RLA, or providing easy display file support for InfoWindow II extensions to the 5250 data stream.

(Yes, I know you can indent in /free, the same as you can in most other languages. But I've learned not to trust indents to accurately show nesting, particularly when multiple programmers are in the source. It's much easier to just get a hardcopy, and do a nesting analysis with colored pens. Or use an IDE that does nesting analysis on the fly (something that could have been added to SEU!)

--
JHHL