Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!m.cs.uiuc.edu!hasker
From: hasker@m.cs.uiuc.edu
Newsgroups: comp.sys.amiga
Subject: Re: TDI Modula-2
Message-ID: <7200149@m.cs.uiuc.edu>
Date: 25 Oct 89 00:11:30 GMT
References: <1461@ultb.UUCP>
Lines: 37
Nf-ID: #R:ultb.UUCP:1461:m.cs.uiuc.edu:7200149:000:1362
Nf-From: m.cs.uiuc.edu!hasker    Oct 24 09:56:00 1989


/* Written  2:36 am  Oct 23, 1989 by kxs5829@ultb.UUCP in m.cs.uiuc.edu:comp.sys.amiga */
/* ---------- "TDI Modula-2" ---------- */
Hi,
	Can anyone either post or e-mail me all known bugs in TDI
Modula-2?

			Thanks,
			Kyle
		 kxs5829@ultb.isc.rit.edu
/* End of text from m.cs.uiuc.edu:comp.sys.amiga */

Why?  Because you are thinking of buying it or because you have a specific
problem?

The problem I found with TDI M-2 is that it gives up on almost any piece of
non-trivial procedure, complaining that either some expression is too 
complicated or the procedure is too long.  For instance, I had a sequence of
if statements such as
	IF p1 & p2 & p3 THEN
	  IF p4 & p5 THEN s1 ELSE s2 ENDIF
	ELSEIF p6 & p7 & p8 THEN
	  IF p9 & p10 THEN s3 ELSE s4 ENDIF
	ENDIF
(where the p's are either tests or function calls and the s's are simple
assignments--this is reproduced from memory, so the details may not be
correct).  I needed to do this several times in the same procedure.  It
turns out I had to create a separate procedure for each of the blocks.
This is a real pain, and it happens so often that the tool is useless.
Maybe they've fixed it since, but if they have then why did it take many
years to get around to it?  My impression is that the compiler isn't really
buggy, but it has lots of unnecessary restrictions.

Rob Hasker
hasker@cs.uiuc.edu