Re: [Devel] Freetype & Ascent

[David Turner]

Hi Antoine,

Antoine Leca wrote:
> > > [mailto:address@hidden On Behalf Of Keith Trummel
> > > Sent: Monday, June 28, 2004 4:35 PM
> > > First, I would think the check should be:
> > >
> > >        if ( face->os2.version != 0xFFFF && root->ascender == 0 )
> > >
> > > instead of:
> > >
> > >        if ( face->os2.version != 0xFFFFU && root->ascender )
>
>
> Minor nit: 0xFFFF and 0xFFFFU are exactly the same thing: unlike decimal
> constants, octal and hexadeciaml ones promote to unsigned "before" going to
> long; and under ANSI's value preserving rule, the test will be exactly the
> same, even if the version member is larger than 16 bits.

Well, you don't know the kind of crappy proprietary compilers we need to 
deal with at work, and the ugly things they do to us :-) For example, 
the following will not compile without a warning:

  short  a = 3;
  short  b = 4;
  short  c = (short)(a - b);                /* warning !!  */
  short  d = (short)((short)a - (short)b);  /* compiles ok */

yes, this is totally hideous, but we need to deal with these. I
guarantee you that 0xFFFF and 0xFFFFU are really different for the
same compiler. Grrrrrrr !!

Cheers,

- David Turner
- The FreeType Project  (www.freetype.org)


> Or course, the test against 0 IS very different!
> Note that I did not check whether it was correct or not
>
>
>
> Antoine
>
>
>
> _______________________________________________
> Devel mailing list
> address@hidden
> http://www.freetype.org/mailman/listinfo/devel