Re: Compile error when using field in a subfield data structure

[Hans Boldt <boldt@xxxxxxxxxx>]

Bob Cozzi wrote:
> Hans,
> I think a free-format I/O spec is not necessary. We already have a pretty
> good D spec, why not just allow that to be used for (at least) Input specs?
> Problem solved.

You effectively have that already in V5R2, and in an improved form in 
V5R3, where you can code a data structure result on externally described 
I/O. In V5R2, you're limited to using LIKEREC data structures on record 
name I/O. But in V5R3, you can also use externally described data 
structures and LIKEDS DS's provided they can be linked back to the 
record. Also, in V5R3 you can do this on file name reads, including the 
ability to read by file name from multiple format logicals.

> ...
> As another person said in this thread, RPGIV's inconsistencies are becoming
> too difficult to teach inexperienced RPG IV programmers. This is just one
> more item that contributes to that perspective.

I don't disagree. But as long as RPG programmers want progress, we have 
to live with the inconsistencies. As you yourself pointed out, the data 
structuring enhancements in V5 are the most useful and important 
improvements to the language in a long time. But they are also a big 
source of inconsistencies.

I may be just an idealistic dreamer, but I look forward to the day when 
new RPG programmers no longer have to learn any of the old cruft in the 
language.

Cheers! Hans