Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site uwmacc.UUCP
Path: utzoo!decvax!ucbvax!ucdavis!lll-crg!seismo!uwvax!uwmacc!oyster
From: oyster@uwmacc.UUCP (Vicious Oyster)
Newsgroups: net.micro.atari
Subject: C for the 8-bit Atari
Message-ID: <1661@uwmacc.UUCP>
Date: Wed, 6-Nov-85 12:54:00 EST
Article-I.D.: uwmacc.1661
Posted: Wed Nov  6 12:54:00 1985
Date-Received: Fri, 8-Nov-85 02:09:39 EST
Distribution: na
Organization: UWisconsin-Madison Academic Comp Center
Lines: 26

<>



   Thanks for the info on where to find the Atari Technical Users Notes.

   As for your Deep Blue C problems... well, I recently purchased Deep Blue C,
and frankly, it's lacking in several respects.  First and foremost, it is *not*
a true compiler.  It "compiles" into pseudo-code, which is subsequently 
translated for execution at runtime.  My admittedly non-exhaustive testing has
shown that it's maybe 20% faster than interpreted Basic, and the generated
command files take up a lot more disk space (I don't have the figures with me).
Another large problem (why the .COM files are so large) is that the linker
doesn't use libraries; if a file is listed in the .LNK file, the entire
thing gets thrown in, even if you're only using one of the functions.
   As chairman of the local Atari User's Group languages SIG, I have just
embarked upon the odyssey of attempting to write a "true" Small-C compiler,
complete with relocating linker featuring library management.  Now, it may be
that I come to the conclusion the J. H. Palevich did (i.e. it's not
possible, given diskspace and memory constraints), but I'll learn a lot
trying.  If you have any suggestions, comments, etc., let me know.

 - Joel Plutchak
   {allegra,ihnp4,seismo}!uwvax!uwmacc!oyster