Re: Keyed record-level access in CL

Adam Glauser wrote:

...I still think that building a command string for QCMDEXC is considerably uglier than calling a command in CL. Particularly when you need to use MONMSG.

I know that I can use QMHRCVPM. I've done it, I've wrapped it in my own procedure to make it nicer. It works and it isn't that bad, but in this case it ends up being a bit cumbersome. That's what's great about ILE, I can mix RPG and CL letting each do what it does best.

I lean this way as well. I've seen cases where there is so much CL being run (interpreted) in RPG, along with all of the extra RPG code to retrieve, decipher and handle messages, plus the twists and turns to avoid passing actual parameter values by forcing things into character data types or other conversions, along with proper string quoting (which would often crash as soon as an unexpected embedded quote came along), that it looks like an attempt to recreate the perfectly good CL compiler.

Tom Liotta