A use for integer quotients

[Gareth McCaughan]

Stephen Horne wrote:

[GvR:]
> >Well, so far the only arguments I've heard come down to
> >
> >(1) int/int *ought* to return an int;
> ...
> >IMO, argument (1) is misguided.  I've heard all the variations many
> >times by now and none are convincing to me; they are either wrong,
> >like "that's how mathematicians define it" or "that's how all
> >right-thinking programming languages define it", or they miss the
> >point, like "it's easy to explain to newbies that int/int returns
> >int".
> 
> Please read my 'Steve Summary' message - it addresses all these and
> more.
> 
> If you can prove that the set of mathematicians only includes those
> who are dealing with the currently interesting discrete-mathematics
> problems, or you can prove that magically warping discrete measures
> into continuous measures is useful for typical bread-and-butter
> programming tasks, I'd be interested to hear it.
> 
> I define a mathematician as someone who applies mathematics, and who
> understands the mathematical principles being applied. I don't limit
> it to people working in field theory or some such, however interesting
> it may be and however pervasive it may be in current
> college/university study and specialist application of discrete
> mathematics. Any narrowing of that definition should be explicit or at
> least for day-to-day practical reasons.
> 
> I define practical as the normal case, not the first-day newbies
> expectations and not specialist fields.

I am a professional applied mathematician. I was trained as
a pure mathematician. 1/2 is not 0. I have never knowingly
met a mathematician or any persuasion who thought that 1/2
was 0.

By the way, any definition of "mathematician" that excludes
pure mathematicians is spectacularly broken.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
.sig under construc