Re: scroll-conservatively overflow

[Kim F. Storm]

Juanma Barranquero <address@hidden> writes:

> On Sat, 17 Apr 2004 15:42:07 -0400, Richard Stallman <address@hidden> wrote:
> 
> > Perhaps we should use long long for that
> > multiplication, and if the result is bigger than the largest positive
> > integer, reduce it to the largest positive integer.
> 
> Well, that's almost what my patch does, with the difference that for
> large values of scroll-conservatively it won't produce the largest
> positive integer, but 
> 
>   int(most-positive-fixnum / FRAME_LINE_HEIGHT (f)) * FRAME_LINE_HEIGHT (f)
> 
> but the two are equivalent for all practical purposes.

Yes, IMO your patch is good enough in practice.

-- 
Kim F. Storm <address@hidden> http://www.cua.dk