John Hudson wrote,
>
> Ligatures are not generally entered via a single keystroke: they are glyph
> representation of multiple characters that are usually entered
> individually. The notion of handling surrogate pairs via ligature
> substitutions was discussed a couple of months ago on the OpenType list,
> and the view of font developers seems pretty unanimous that this is a Bad
> Idea. Handling surrogate pairs is something that should be happen in
> character space, not in glyph space, which means that it properly belongs
> in something like Uniscribe, not in the glyph substitutions tables of
> individual fonts. The font support should be limited to providing the
> correct cmap table format to support supplementary plane characters.
>
This is correct for fonts, but I think Michael Kaplan was referring
to the ligature table in a keyboard driver. In the keyboard driver,
(IIRC) the ligature table is used to split a single keystroke into two
or more characters which are then passed along to the system. This
is perhaps the best approach for building a keyboard driver for
non-BMP characters.
Best regards,
James Kass.
This archive was generated by hypermail 2.1.2 : Sun Dec 16 2001 - 22:32:40 EST