pgl@cup.portal.com (Peter G Ludemann) (12/04/89)
The recent discussion about ISO's WG17 has been too much in the
spirit of alt.flames. There may be a number of irreconcilable
camps but they do not need to be ARMED camps.
-----
Many people, including myself, have been too busy to participate
actively in the standardisation effort. The active committee
members deserve our thanks. We may not like them, respect them,
nor agree with them; but their long hours will benefit all Prolog
users and vendors.
I know how difficult standardisation is: I was responsible
for implementing a new ANSI-C library while the standard
was still being written. I had many long discussions with
our company's ANSI representative. What at first seemed to be an
obvious defect in the draft standard turned out to be much more
difficult to resolve in the context of other computer systems
(and in 20 years, I've seen a variety of systems).
-----
So, with the intention of getting a useful Prolog Standard:
- Remember that that this is comp.lang.PROLOG, not
comp.lang.EDINBURGH nor comp.lang.OKEEFE. [I have greatly
benefited from Richard O'Keefe's many contributions to this
newsgroup and I know that he is a nice guy in person, but his
writing style is often brutally abrasive. This may not bother
him or me but it does bother some people. I want people to be
unafraid of contributing even if they are not as eloquent or
(gasp!) correct as Richard.]
- Specific deficiencies in WG17's draft should be criticised in a
professional manner, with the intent of improving the Standard,
not with the intent of proving how correct one was in 1984
and how stupid the Draft Standard's authors are.
Each deficiency should be presented in one article, with
criticisms and suggested remedies. Please don't discuss a
bunch of problems in one article -- it might be fun for the
writer to talk about everything at once, but it's hard on the
readers.
- Avoid meta-discussions and editorial snide remarks:
- We can't change ISO's structure.
- 2500 years of philosophers haven't produced a universally
acceptable definition of "democrarcy".
- Don't presume someone from Company X is proposing something
for some devious business reason -- Company X might have
much more integrity than you had thought.
- Some compromises are inevitable. Parts of Edinburgh Prolog's
design seem (to me) to have been dreamed up at 2 a.m. during an
all-night hacking session. But if the majority thinks that
these decisions are best, then I shall give in with good grace.
In most cases, a bad decision is better than no decision
(and, possibly, no Standard).
- Before blasting someone in public, why not send a private note?
I know that this may be difficult (e.g., my e-mail system has so
far refused all my attempts to send to OZ.AU).
- Some things can't be resolved in this newsgroup. That's why the
standardisation committee has meetings.
- Re-read the Usenet style guidelines from time to time.
I especially like the advice:
Before you post an article, think a minute; decide whether
or not you are upset, angry, or [intoxicated]. If you are,
wait until you calm down (or [become sober]) before deciding
to post something. Then think about whether or not you really
want to post it. You will be amazed what waiting a day or
even a few hours can do for your perspective.
-----
Hoping for a speedy resolution of the Prolog Standard,
- Peter Ludemann pgl@cup.portal.com
...!sun!portal!cup!pgl
--- my opinions are my sole responsibility, of course ---
Direct all flames to me, not to the newsgroup.jeff@aiai.ed.ac.uk (Jeff Dalton) (01/04/90)
In article <24701@cup.portal.com> pgl@cup.portal.com (Peter G Ludemann) writes: >Many people, including myself, have been too busy to participate >actively in the standardisation effort. The active committee >members deserve our thanks. We may not like them, respect them, >nor agree with them; but their long hours will benefit all Prolog >users and vendors. But that's one of the points in dispute: will the standard make things better or worse? It's certainly *possible* for it to make things worse. I used to think it would be easy to make a Prolog standard. The language was fairly small and well-defined and there was already a large degree of compatibility among implementations. Moreover, most of the incompatibility between the Edinburgh syntax Prologs and the rest was just a matter of syntax. Richard's approach to standardization was designed to make those differences less significant. However, I'm disappointed to find that Prolog isn't easy to standardize because people don't seem to be able to agree on what the language should be like. The problem seems to be that too many people aren't sufficiently happy with the language as it is and so want to change it in various ways. Or else they want some interesting features from their version of Prolog to be incorporated in the standard, even though the Prolog community has not decided to move in that direction on their own. I hope I'm wrong here, but that is how it often looks. >- Remember that that this is comp.lang.PROLOG, not > comp.lang.EDINBURGH nor comp.lang.OKEEFE. Richard has been somewhat unreasonable in this discussion, but so have some of the people on the other side. (Remember this has been going on for a couple of years now.) On the whole, I think Richard has been a big plus for this newsgroup. > - 2500 years of philosophers haven't produced a universally > acceptable definition of "democracy". The standardization process is certainly less democratic then we might like. For example, it is biased in favor of people who can afford to attend meetings. It's not necessary to fight over definitions of "democracy" to make a point of this sort. > - Some things can't be resolved in this newsgroup. That's why the > standardisation committee has meetings. But it would be good if it were possible to participate in the process, at least to some extent, by e-mail. The various groups concerned with Lisp standardization have benefited immensely from this.