I've long maintained that Rexx's (arguable) failure (I'm a huge fan, so I resist that term, but the popularity of first Perl and now Python kind of make it true) is due to two things:
1) It's IBM and too many folks therefore concluded it must be bad. 2) There were few decent/public examples of how to write function packages for it. I did so on CMS a few times, and it wasn't as easy as one might like; and they seem rare to nonexistent on z/OS (yes, yes, I know y'all are about to cite a bunch of examples, but relatively speaking, they seem rare). RXIUCV, RXSOCKETS, RXLDEV on CMS were popular, but CPAN and whatever the Python equivalent are beat Rexx by orders of magnitude. Revisionist history is always easy, but: if IBM had at least made it easier and/or just provided some simple examples of function packages, maybe we'd be having a different conversation now...though I still doubt it, since IBM never tried to make it open source or even promoted it on non-IBM platforms. I guess that last might should be in the list above as #3. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
