[milters] Archive

Lists Index Date Thread Search

Article: 710
From: Kevin Rosenberg
Date: 2005-08-27 16:10:38 -0400
Subject: Re: libsnert/milter-ahead on AMD64

Removal...........: milters-request@milter.info?subject=remove
More information..: http://www.milter.info/#Support
--------------------------------------------------------

Anthony Howe wrote:
> > So, from a portability standpoint these days, one can count on int
> > being 32-bits and that long varies across platforms. That's quite
> 
> No. int has NO defined size. It can be the sizeof (int) can be equal to 
> either sizeof (short) or sizeof (long) or some where in between, which 
> prior to 68 bits systems never happened.

Yes, I agree that int does not have a defined size (like long, but
unlike char and short). I was speaking pragmatically with respect to
current platforms with 32-bit and 64-bit cpus.

> An int is defined to represent the architecture's native or optimal 
> machine word. So in the days of CP/M Z80 an int was a short, then later 
> with 80x86 and M68000 systems compilers would define int equivalent to a 
> long.

That's right. My first system was a Z80 and I programmed assembly on
that system. I didn't start with C until I had an 8088 system.

> Many people don't understand it and complain that its "wrong", but I 
> borrowed from Java directory tree structures, since LibSnert is also 
> used by some of my Java JNI code and so had to merge into that 
> development environment too.

I don't use Java much, but I recognize the similarity. It works well
for doing "tar xzf" of multiple tarballs and having everything end-up
in the right place.

-- 
Kevin Rosenberg
kevin@rosenberg.net

Lists Index Date Thread Search