>> [...] we are talking about the possibility to shift glyphs by a
>> call to `FT_Set_Transform', where the shift value is in the range
>> [0;1] (or rather [-0.5;0.5]), and this call takes place *before*
>> hinting, which is not possible current.
>
> Rotation forfeits hinting, but you suggest that shifts would not.
Yes, the auto-hinter could be extended to cover this case. On the
other hand, today we have a tendency to avoid hinting along the
x axis, so we are probably talking about a hypothetical issue. Please
remember that the origin of the whole discussion is an innocent but
rather cryptic comment in the source code, and what to do with it.
> Therefore the advance, rounded or not, is not sufficient to control
> spacing.
It would be sufficient, since the hypothetical `fractional hinting' is
`strong' hinting, which modifies the advance width also so that...
> You need the bearing deltas anyway.
... you don't need deltas at all, AFAICS.
Werner
_______________________________________________
Freetype-devel mailing list
[email protected]
https://lists.nongnu.org/mailman/listinfo/freetype-devel
