Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!rutgers!seismo!mimsy!eneevax!todd
From: todd@eneevax.UUCP
Newsgroups: comp.lang.prolog
Subject: Re: C-Prolog Bug Fixes Wanted
Message-ID: <831@eneevax.UUCP>
Date: Tue, 14-Apr-87 18:35:22 EST
Article-I.D.: eneevax.831
Posted: Tue Apr 14 18:35:22 1987
Date-Received: Wed, 15-Apr-87 23:42:13 EST
References: <4855@robin.cs.nott.ac.uk> <1636@megaron.arizona.edu>
Reply-To: todd@eneevax.umd.edu (Todd Aven)
Organization: the Softwear Sweatshop
Lines: 41
Keywords: they *sell* this to VMS shops?
Summary: Far too many #ifdef's


>In article <4855@robin.cs.nott.ac.uk>, abc@cs.nott.ac.uk (Andy Cheese) writes:
>> 
>> I am compiling a list of bug fixes for C-Prolog 1.5
>> I would be very grateful if you have improved or altered/fixed
>> it all for your improvements.

OK, I hacked at it too. Read on...

>The following changes were made to C-Prolog+ by Chris Janton to get it to
>run under version 4.4 of VMS.

[lot's of diffs removed to save on phone bills for everyone :-) ]

Looks like Chris made the substantive changes I wanted to work on
'when I got the time'. I made superficial changes, but in the course of fixing
things I found that there were a lot of unnecessary '#ifdef VMS' lines
floating around. So what I'll do is this... I'll incorporate Chris'
modifications to my code (fixes bugs I haven't even discovered yet),
then I'll DIFF against the distribution with SUM/SLP output to make
changing easier. Then, if Edinburgh incorporates that into their 
distribution it might be worth money :-). No, I shouldn't really
rag on them so hard, it really is a nice interpreter. I just don't
think they should go out of their way to make VMS look so screwy.

>        copied times.h from Eunice distribution to current directory

I fixed the timing code directly, instead of bringing in a
foreign include file. My philosophy is that if it compiles on
my VAX with C 2.2, then it will compile on *your* VAX. If you
start bringing in extra files, things get muddy (and lost).
Thus, it's a personal preference. But at least my way I get
the same times reported for building the initial system on the
Sun and the VAX, so it can't be all bad.

So, I modified the makefile (only slightly) since VMS command lines
are a little different, I used pl/all instead of pl/vmsall, and it runs.

Regards,
Todd Aven
the Softwear Sweatshop ('home' of completely distributed computing :-)