Re: Double Macrons on "gh"...

From: William Overington (
Date: Sat Aug 31 2002 - 05:33:30 EDT

Kenneth Whistler wrote as follows.

>In practice, fonts might simply choose to have ligatures for
>the entire sequence, to avoid complications of calculating
>the accent positions dynamically.
>For more examples, just look in dictionary pronunciation guides.

An interesting problem which may arise is that the Unicode Consortium will
not be specifying particular ligatures to include in fonts and that font
designers may not have available from any public source a list of such
ligatures for which to prepare the glyphs to include in a font. This could
then result in a muddle in the future when end users are trying to use such
ligatures in a document and find that for some key ligatures which they wish
to use that the implementation in some fonts is by default action rather
than special glyph, which default action may, for some requested ligatures,
result in a typographically awful display.

This issue first came to my attention in the matter of the ligatures for the
romanization of Cyrillic names and unknown words, where special ligatures
would be desirable due to the need to have U+FE20 and U+FE21 act in both TS
and iu ligatures.

I wonder if, for the guidance of font designers, there should be a list of
desirable ligatures for which font designers might choose to prepare
specific glyphs for inclusion in an advanced format font, the list prepared
by consultation between the various dictionary publishers, libraries and so
on. Such a list, while not obligatory for anyone to use, would nevertheless
be a useful collected guide which font designers could use so that fonts
could be designed so as have individual glyphs for all of the ligatures on
the list. The list could include the specific Unicode sequence to access
each ligature.

It may be that there would need to be more than one list, so as to provide
for various specialised areas of activity without making a general list too

Do you think that such a published list or lists would be useful?

William Overington

31 August 2002

This archive was generated by hypermail 2.1.2 : Sat Aug 31 2002 - 06:28:44 EDT