[Devel] Unpatented Hinting

[Tamas]

>> recently I tested the unpatented hinting method. Unlike to the given
>> Information in the mailing list archive I apparently gain the same
>> quality by unpatented hinting as with the "full" hinting method with
>> bytecode interpreter for ALL of my fonts.
>
>Note that only a few operations in the bytecode interpreter are
>patented.  Those features are disabled by default but the rest of the
>bytecode interpreter is working.
>
>> Does the quality that can be achieved with the unpatented algorithm
>> has meanwhile been improved that much?
>
>It depends on the font.  Does someone knows a good counterexample?
>That is, which glyph in which font is rendered badly with unpatented
>hinting?
>
>> I've set the following definitions for unpatented hinting:
>> 
>> #define TT_CONFIG_OPTION_BYTECODE_INTERPRETER
>> 
>> and
>> 
>> #define TT_CONFIG_OPTION_UNPATENTED_HINTING
>> 
>> in include/freetype/config/ftoption.h
>> 
>> and I pass
>> 
>> FT_PARAM_TAG_UNPATENTED_HINTING
>> 
>> to
>> 
>> FT_Open_Face()
>> 
>> Is that all or will there be needed something more for activating the
>> unpatented algorithm?
>
>This should be fine.
>
>
>     Werner

Hello Stefan,

I've just checked the two configurations and found no difference
whether I passed FT_PARAM_TAG_UNPATENTED_HINTING or not... until
I realized I forgot the FT_OPEN_PARAMS flag in FT_Open_Face(). 
Have you set this flag during your tests ?

Then, you can see important differences, especially at small font
sizes (I tested will Arial).

To be sure, you can also check the value face->unpatented_hinting
(TT_Face).

Regards,
Alexis