GUI

[Taavi Burns]

On Sat, Nov 29, 2003 at 03:31:59PM -0500, James Knott wrote:
> "fun" thing I came across in that C homework, was the different variable 
> sizes in the Borland C++ (OS/2) I had at home and Turbo C++ (DOS), which 
> we used in class.  I'd write my program and it would run fine at home, 
> but then in class next day, I'd over run some variable.  However, the 
> same problem would have occured if I'd used Borland C++ for Windows.

I wouldn't be so sure.  Windows 3.1 was still very 16-bit, and would
probably have had a lot of 16-bit bias in it.  OS/2 Warp 3, however,
was a fully 32-bit OS, just like Linux, and so would have a 32-bit
bias.

There's currently a similar problem with Windows on 64-bit hardware,
where 'int' is still a 32-bit number; it's the only 64-bit platform
I know of where this is the case.  I suspect this was done because
setting it to a 64-bit type caused breakage due to shoddy, old,
crufty code that nobody understood anyway.  But that's purely
supposition.  It's not like I've ever peeked at the Windows
source code to find out.

-- 
taa
/*eof*/
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml