freetype-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Devel] round, ceil, and floor functions for 16.16 fixed values


From: Antoine Leca
Subject: Re: [Devel] round, ceil, and floor functions for 16.16 fixed values
Date: Fri, 09 Mar 2001 13:00:42 +0100

Werner LEMBERG wrote:
> 
> > ... I decided to implement a fixed number rounding function ...
> >
> > For those machines that support 32 bit signed ints, this is not a
> > problem.  For those machines whose largest integer is 16 bits wide,
> > this is a problem.  I took a look at how things are down in ftcalc.c
> > when the largest integer is 16 bits wide, but I am at a loss.
> 
> I think that FT 2, contrary to FT 1, has not been tested whether it
> works with 16bit integers

At least, I never gave it a shot, although I certainly planned to do so.

> -- I'm quite sure that it doesn't work (Antoine, please correct me if
> I'm wrong) since I've never seen any 16bit related fixes.

I do not know. I do not really expect it to be completely brocken (we
know that the TT stuff is quite OK, and experience with FT1 show that
a couple of fixes like adding "L" suffix here and there just work.
Sometimes things are much more complex, such as how to deal with big
fonts whose structures exceed the 64K limit...)

 
> So don't worry, just add a remark about possible 16bit problems.

Seems the way to go. This way, when someone will try to have this to work
on a 16-bit platform (which might happen but also may not), the guy will
know where she had to search first (BTW: the first will be the type system,
much before diving into ftcalc.c ;-) so don't worry Tom).
OTOH, I have no good idea of a place to collect these kinbds of comments
(in addition to the sources themselves). In FT1 we had a docs/TODO which
collected these kind of business, but I did not notice something similar
in FT2.


Antoine



reply via email to

[Prev in Thread] Current Thread [Next in Thread]