home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!howland.reston.ans.net!paladin.american.edu!gatech!pitt.edu!pitt!willett!ForthNet
- From: ForthNet@willett.pgh.pa.us (ForthNet articles from GEnie)
- Newsgroups: comp.lang.forth
- Subject: Environmental Queries
- Message-ID: <4188.UUL1.3#5129@willett.pgh.pa.us>
- Date: 25 Dec 92 10:09:00 GMT
- Organization: EIEI-U
- Lines: 76
-
- Category 10, Topic 33
- Message 17 Sat Dec 12, 1992
- B.RODRIGUEZ2 [Brad] at 17:27 EST
-
- During the second public review period of ANS Forth, the TC has
- satisfactorily addressed -- or at least convinced me to
- postpone -- most of my concerns. But although some aspects of
- ENVIRONMENT? were clarified, its operation remains unchanged.
- My reply to the TC's response follows:
- ----------------------------------------------------------------
-
- Re. Pp.3-4: S" should be interpretive in CORE
- Pp.5-6: ENVIRONMENT? should take text from the input stream
-
- I do NOT accept the Committee's decision. I do not feel that the
- committee has adequately addressed my concern, or has justified
- its decision on technical or compatibility grounds.
-
- The Committee's argument seems to be: "This is only a problem for
- cross compilers. We are not addressing cross compilers.
- Therefore, this is not a problem."
-
- The Committee has indicated that cross compilers _will_ be
- addressed in a future edition of the standard. Your reply states
- that "...the TC considers it highly desirable to address this
- issue [cross-compilers]." If we can identify a problem _now_ that
- will plague the next edition of ANS Forth, why not fix it _now_?
-
- Your rationale note A.6.1.1345 makes clear that ENVIRONMENT? can
- only be portably used in user-defined words. But the committee
- has not yet taken a stand on whether execution of user-defined
- words will be allowed or necessary in a cross-compiler
- environment! And, by your own statement, the Committee "...would
- be remiss in undertaking such a task without prior public
- announcement, and without having solicited specific input from
- vendors of cross-compilers."
-
- Enshrining this usage of ENVIRONMENT? will, in effect, mandate
- one of the major decisions which pertain to cross-compilers -- a
- decision which should be reserved until cross-compilers are
- properly deliberated.
-
- At no time has the Technical Committee indicated why it is
- necessary for ENVIRONMENT? to take a string argument from the
- stack. No technical justification has been offered.
- ENVIRONMENT? is a new word, so there is no question of
- compatibility with previous versions.
-
- ENVIRONMENT? is also risky in that its portable usage is not
- obvious. Many ANS Forth users, failing to note section
- A.6.1.1345, will surely employ interpretive S" in their
- "portable" Forth code.
-
- The proposed word ENVIRONMENTAL is technically superior, in that
- its portable use is obvious, and is compatible with Forth cross-
- compilers.
-
- The time to fix this problem is now, before a troublesome and
- potentially incompatible practice has been established. If a
- problem can be foreseen, and simply fixed, to refuse to
- acknowledge or correct the problem is negligent.
-
- ----------------------------------------------------------------
- I'm posting this for further discussion. It is my fear that
- ANS Forth will be made forever impractical for metacompilers;
- so far, the TC has not been reassuring.
-
- Brad Rodriguez
- B.RODRIGUEZ2 on GEnie | bradford@maccs.dcss.mcmaster.ca
- So many projects....so little time.
-
- -----
- This message came from GEnie via willett. You *cannot* reply to the author
- using e-mail. Please post a follow-up article, or use any instructions
- the author may have included (USMail addresses, telephone #, etc.).
- Report problems to: dwp@willett.pgh.pa.us
-