Re: Scoping of procedure-level variables

[James Lampert <jamesl@xxxxxxxxxxxxxxxxx>]

Barbara Morris wrote:

> When you say your recursion experiments suggest dynamic, you're talking 
> about the storage type for the variables (see Joe Pluta's original post 
> in this thread).  You probably had variables in automatic storage, which 
> meant you got new storage for each recursion.  If you had coded STATIC, 
> you would have seen different results (same variable for all recursions).

Actually, given that ILE RPG supports neither inline true inner blocks 
(i.e., blocks with local variables in inline code), nor nested procedure 
definitions, most of the difference between dynamic and lexical scoping 
falls away (at least unless you've got something that looks like the 
test case in the Wikipedia article).

But you are correct (and it goes to show just how much programming 
language design theory I've forgotten), in that I was asking the wrong 
question. The issue with the program in question was really a matter of 
local scope overriding global. (To restate what had happened, the 
programmer in question had intended DS definitions in an INCLUDE to 
overlay file fields, but because he'd put the include in at procedure 
level instead of module level, they were *overriding* rather than 
*overlaying* the file fields.)