home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!gatech!psuvax1!psuvm!lrl
- Organization: Penn State University
- Date: Tue, 15 Sep 1992 15:43:28 EDT
- From: Linda Littleton <LRL@psuvm.psu.edu>
- Message-ID: <92259.154328LRL@psuvm.psu.edu>
- Newsgroups: comp.lang.rexx
- Subject: Re: Blanks, REXX, and portability...
- References: <REXXLIST%92090916473059@DEARN>
- Lines: 45
-
- Whew! I finally finished wading through this long discussion, and
- since I attended the ANSI meeting last week, I thought I'd chime in
- with what got discussed there.
-
- The topic came up the afternoon of the first day and consumed the
- better part of that afternoon. One group member read a few quotes
- of postings from comp.lang.rexx. The discussion here definitely
- had its influence.
-
- There was unanimous agreement that tabs (and other white space
- characters) need to be acceptable in Rexx source and that tabs
- in literal strings should be considered as tabs only.
-
- For tabs in data, the programmer would have the ability to specify
- whether or not tabs should be translated to spaces. So, even within
- the same program, the programmer could sometimes have tabs translated
- and sometimes not.
-
- The programmer would specify either "do translations" or "do not do
- translations" (syntax for specifying still needs to be discussed).
- The leaning for a default was toward doing the translations; however,
- I'm having second thoughts on this, thinking that it would be better
- for the default to be the current method -- no translations.
-
- The interpretation of "do translations" (i.e. which characters get
- translated to space) might vary from implementation to implementation
- according to what is commonly accepted on that system. In the first
- standard, if the programmer specifies that translation is to be done,
- then it gets done however that implementation does it. In a later
- standard (when non-essential enhancements will be added), the user would
- be able to specify exactly which characters get translated.
-
- We discussed the possibility of having each installation specify which
- characters get translated, but decided that it's not good to have the
- same implementation work different ways at different installations.
-
- As I mentioned, the syntax is not determined, but it would be a one-time-
- per-program thing, rather than an extra parameter on each applicable
- function call, as someone suggested.
-
- Keep your comments coming. Even though the volume can be overwhelming,
- the folks at ANSI want to hear what you're saying and the message is
- getting through.
-
- -Linda
-