Newsgroups: comp.lang.pascal
Path: utzoo!utgpu!watserv1!maytag!watstat.waterloo.edu!dmurdoch
From: dmurdoch@watstat.waterloo.edu (Duncan Murdoch)
Subject: Re: Doubts on TP 6.0 updatating being worth the while
Message-ID: <1991Jan9.140848.12962@maytag.waterloo.edu>
Sender: daemon@maytag.waterloo.edu (Admin)
Organization: University of Waterloo
References: <1991Jan8.215108.24559@uwasa.fi> <2396@bnlux0.bnl.gov>
Date: Wed, 9 Jan 91 14:08:48 GMT
Lines: 44

In article <2396@bnlux0.bnl.gov> kushmer@bnlux0.bnl.gov (christopher kushmerick) writes:
>
>Now I am sure that vision offers some powerful features. Borland
>claims that they even use it to implement the front end (Actually, I
>don't know if _I_ would use that example to promote anything).
>
>However, A: it is implemented as an OO library, and B: the source is not
>provided.

A minor correction: according to Borland, full source for TurboVision is
available, with the run-time library source code.  I think the price was
$199US.  This is really competitive with the prices for similar libraries
for C, but way out of the league of TP libraries, where Object Professional
is available with full source code for about $125US from mail-order houses.

>I havn't yet seen what is so great about the new front end. Part of this
>is no doubt due the the fact -I've said it before and I'll say it again- 
>the new color scheme does not work on a LCD screen, and lately thats 
>what Ive been using.

It's also not so hot on a monochrome screen, if you're using a mouse.  The
scroll bars use one of those block characters with half the pixels on and
half off.  The mouse cursor is just a reverse-video block; when it lands
on the scroll bar, it all but disappears.

>My advice: If you are happy with 5.0, stay there.

That's generally good advice.  One bug in 5.0 that was fixed in 5.5 is with
floating point code generation:  if you translate Fortran programs with long,
complicated expressions that are full of constants, you'll find that 5.5 (and
6.0) are much less likely to give you an error 207 from a coprocessor
stack overflow.  In fact, if "e" is of type extended, the expression 
 
  e+e+e+e+e+e+e+e+e+e

can't be evaluated in 4.0 or 5.0, but it presents no problems in 5.5 or 6.0.

I don't know of any 5.5 bugs that were fixed in 6.0, so if you don't like
TV, the only reasons to upgrade would be the inline assembler, the relaxed
rules for linking external .OBJ files, and the new debugger and protected
mode compiler if you go for the "Professional" version.

Duncan Murdoch
dmurdoch@watstat.waterloo.edu
