Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site cadre.UUCP Path: utzoo!watmath!clyde!burl!ulysses!allegra!bellcore!decvax!mcnc!idis!cadre!km From: km@cadre.UUCP Newsgroups: net.lang.c,net.micro.pc Subject: Re: Lattice versus CI86 Message-ID: <128@cadre.UUCP> Date: Wed, 2-Jan-85 11:57:08 EST Article-I.D.: cadre.128 Posted: Wed Jan 2 11:57:08 1985 Date-Received: Fri, 4-Jan-85 00:16:14 EST References: <233@gumby.UUCP> <3194@alice.UUCP>, <6267@boring.UUCP> Organization: Decision Systems Lab., Univ. of Pgh. Lines: 23 Xref: watmath net.lang.c:3667 net.micro.pc:3063 Your comments are well taken. C86 certainly has its share of faults. My major point was that I had an easier time living with them than using the early version of Lattice, and that access to the library sources was helpful. My own two major gripes with C86 are the "extern" problem, which does NOT exist with Lattice when the "-x" flag is used, and the inability to use common member names in different structures with differing offsets, which also does not exist with Lattice. I have not had access to the "large" version of Lattice and cannot comment on the relative speed sacrifices with (uugh) "long" 8086 pointers. I have noticed a pronounced speed degradation with the long pointers in C86, particularly in code containing a lot of arrays. In summary, I did not mean to unilaterally "praise" C86 but to simply comment on my own experience, admittedly with compilers of quite different currentness. I would very much like to see the DeSmet compiler, and others.