Re: [Freetype-devel] Re: GSOC - Distance Fields

[Alexei Podtelezhnikov]

On Sat, Jun 20, 2020 at 6:32 PM Behdad Esfahbod <behdad@behdad.org> wrote:
> Systems without an FPU are vastly less common than they were 20 years ago.  
> They still exist, and is a defendable position to want FreeType to continue 
> to work on those systems, however:
>
>   - Compilers and kernels have stepped up to provide floating-point emulation 
> libraries which work transparently to the client code.  That is, introducing 
> limited use of float in FreeType is by no means an impediment to those 
> building the library for systems without FPU.  Even if that was not true, the 
> SDF module can be easily disabled,

Behdad,

The FT_Fixed underflows and overflows, which are definitely easier to
trigger, makes you think about degenerate cases in your logic and
equations. This is a huge benefit, not an obstacle. In the case of SDF
the distance always exists and is finite. Once the degenerate case is
identified using FT_Fixed, it is better to solve it using an
alternative formula or method, which always exists and is usually
simpler. This will make the code more stable and robust. IMHO, this
discipline makes FreeType indestructible.

I would love to see GPU, FPU, SIMD, compiler features, etc...
Unfortunately, there is maintenance cost associated with it. Just two
days ago 
http://git.savannah.gnu.org/cgit/freetype/freetype2.git/commit/?id=d924c5cf7e5554b22f7edfcb9e98670c4c02c3f0
and that was your contribution but the burden of fixing is on Werner.

It should have been easy to blow FreeType from the face of the Earth
with a faster GPU or SIMD rasterizer or using GPU, but for some reason
for 20 years nobody dared.

Alexei