I'm working on modernizing Rosetta Code's infrastructure. Starting with communications. Please accept this time-limited open invite to RC's Slack.. --Michael Mol (talk) 20:59, 30 May 2020 (UTC)

Category talk:REXX Implementations

From Rosetta Code

I want to change this page from

these are all of the REXX implementations

to

these are some of the REXX implementations,

but there doesn't seem to be a way to do that. -- Gerard Schildberger 22:18, 8 June 2012 (UTC)

Make sure you read the rest of the sentence: "on Rosetta Code". It is all of the ones that are on this site. --Mwn3d 00:10, 9 June 2012 (UTC)
No, there are other REXXes, and most of them (I haven't tried them all, of course) work with most of my REXX examples), but I didn't include the other various REXXes as one of them is proprietary (that is, it costs money to use), and others that I don't know enough about to discuss here, and one or more of them only work on *NIX systems, which I don't have). I don't mention which other REXX interpreters the REXX example(s) work on, that would entail too much testing.
Except for the Hello World program, which would probably work everywhere. I try to make sure that my REXX examples work on the three REXXes that I have on my (Windows) PC. -- Gerard Schildberger 00:23, 9 June 2012 (UTC)
Which other REXXes are there on Rosetta Code that should go in this category? They should be marked as such on their respective pages. If there aren't any others that are already on the site then the text is correct the way it is. --Mwn3d 01:30, 9 June 2012 (UTC)
I'm now confused. I don't quite fully understand about REXXes on Rosetta Code. I assume you mean the examples of REXX programs. As such, there isn't a REXX interpreter on RC, and I haven't been noting which REXX interpreters (if any) for the most part, which REXX it was written for [or runs under -- I write the REXX (example) code to conform to the REXX standard, and most REXX interpreters try very hard (well, maybe not that hard) to adhere to the REXX standard. Don't get me started on this topic. The REXX standard isn't as simple as I make it sound. There is a lot of interpretation in the Regina REXX's "standard" (which is really a documentation manual) for instance. It needs a lot of work to approach IBM's standard of documentation. I only have access to (as I mentioned previously) three REXX interpreters, and I try to use all three to test the REXX programs before posting them. None of the (earlier) REXXes noted which ones they used when I first started reading Rosetta Code, and I continued that tradition (right or wrong, it's my country ...). Oops-ay, I lied. I actually have a fourth REXX interpreter which I won't mention --- it's so buggy that I refuse to even mention the name of it -- I do have a real fear of black helicoptors. It's documentation is almost none-existant. But I really don't want to go back and mark each example for which it was written for or tested under. I note that hardly any examples (not just REXX) do this, although some do, of course. There are other REXX interpreters out there, I know but of a few. The older REXX on CMS and TSO are essentially, the same REXX. I wish dearly that I could have access to these. I consider those REXXes the gold standard. If a REXX program works there, and not elsewhere, usually the REXX that "doesn't" work that gets changed. (If not changed, there usually appears an option to make it conform.) REXX programs written on IBM mainframes will run (compile) forever, IBM is very good and keeping the older programs running. Upward compatible in IBM-land is practically a mantra, a religious creed. Kudos to the old guard. There is also one REXX compiler (I think, only on the IBM mainframes) --- I believe it is of different code. I have even less REXXes than that which I can use/test. I just looked at the "sum of squares" Rosetta Code task, and of 94 examples, only 7 mention which "compiler" they use, and some of them, like Algol, Fortran, BASIC, and a few others have versions are are somewhat incompatible, probably because of added (and/or depreciated) features (this is my assumption and probably reflects too much of me viewing the bigger picture). To top it all off, there are some people who think that ooRexx and oRexx are also REXXes, and almost none of my (classic) REXX will run as is under those object-orientated REXXes (contrary to ooREXX's claim that it'll run REXX programs). Classic REXX is the REXX that started it all (1982 or so), and that name is used to differentiate it from the object REXXes and NetRexx. It is this Classic REXX (standard) that I program the REXX examples for. I forgot to mention KEXX, the REXX that KEDIT uses (both of which I have). I haven't even brought up the differences of the REXXes that behave differently when they execute under a different operating system. Re-reading this entry, I maybe shouldn't have thrown so many stones. I may wake up tomorrow and delete the whole she-bang when I come to my senses. -- Gerard Schildberger 02:19, 9 June 2012 (UTC)
OK...well...this category is for implementations of REXX. That does not mean examples written in REXX. Take a look at Category:Lisp Implementations for an example. It's basically runtime environments, compilers, interpreters, etc. that run code from a common REXX language (possibly with a few quirks built in). This category should only list those implementations. If there is another implementation in common use out there then you can add the page and categorize it as a REXX implementation. If there are other implementations on RC that haven't been marked as implementations then they should be marked (using {{implementation|REXX}}). If they're not on RC then obviously they shouldn't be here (yet). --Mwn3d 03:11, 9 June 2012 (UTC)