Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!uakari.primate.wisc.edu!dali.cs.montana.edu!rpi!batcomputer!munnari.oz.au!goanna!ok
From: ok@goanna.cs.rmit.oz.au (Richard A. O'Keefe)
Newsgroups: comp.lang.misc
Subject: Re: improve language by dropping ;
Message-ID: <4863@goanna.cs.rmit.oz.au>
Date: 1 Mar 91 08:15:15 GMT
References: <8507.27b91f9e@jetson.uh.edu> <8531@plains.NoDak.edu> <1875@borg.cs.unc.edu>
Organization: Comp Sci, RMIT, Melbourne, Australia
Lines: 18

BCPL's rule about treating the end of a line as a ';' token unless it is
"obvious" that the line must be continued is extremely easy to add to a
parser.  For example, I once had a "mid-processor" for C that would do
things like inserting ';' before '}' unless there was already a ';' or
'}' there, and it implemented the BCPL rule.  (It also filtered := to =.)
The scheme was
	cc -E foo.c | midprocessor >temp.c; cc temp.c
It took an afternoon, and was really tiny.  I eventually threw it away,
because it was rather like the strange case of the Dvorak keyboard.
(Hint:  people trained with the Dvorak keyboard are all of 5% faster
than people trained on QWERTY.  Big deal.)  It really wasn't much of an
improvement.

Don't forget one important thing that distinguishes C from Turing and
even from BCPL and S:  the C pre-processor.  Should one apply the BCPL
rule before or after macro expansion?  It can make a huge difference.
-- 
The purpose of advertising is to destroy the freedom of the market.