Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site cbmvax.cbmvax.cbm.UUCP
Path: utzoo!watmath!clyde!cbosgd!ukma!psuvm.bitnet!psuvax1!burdvax!bpa!cbmvax!daveh
From: daveh@cbmvax.UUCP
Newsgroups: net.lang
Subject: Re: BCPL referenes ?
Message-ID: <25@cbmvax.cbmvax.cbm.UUCP>
Date: Fri, 21-Mar-86 12:17:35 EST
Article-I.D.: cbmvax.25
Posted: Fri Mar 21 12:17:35 1986
Date-Received: Sat, 22-Mar-86 22:09:30 EST
References: <1341@osu-eddie.UUCP> <254@vu-vlsi.UUCP> <425@ucbjade.BERKELEY.EDU>
Reply-To: daveh@cbmvax.UUCP (Dave Haynie)
Organization: Commodore Technology, West Chester, PA
Lines: 12
Keywords: BCPL


BCPL on the Amiga is pretty much what you'd expect from BCPL .. a pain
in the #^%&^%@.  I haven't actually used a BCPL compiler on the Amiga,
but quite often C programs have to interface with BCPL code.  BCPL
pointers are pointers to 32-bit words, which means that to access them
in C you've got to >>2 and <<2.  They're always longword aligned
pointers from C's point of view, once converted, and this is why a good
deal of stuff on the Amiga, especially interfaces to AmigaDOS, must
be allocated as longword-aligned, using the AllocMem function from C.
My main point is that BCPL is still a DOG on byte-aligned systems.

-Dave Haynie/Commodore Technology