From: James Kass (thunder-bird@earthlink.net)
Date: Mon Feb 11 2008 - 12:03:55 CST
Philippe Verdy wrote,
>> t + double-inverted-breve + dot above + s, I should think.
>
> Seems illogical. The dot above is not double. This looks like defective for
> me, the dot above is attached to no letter, so we would get either:
> * t+double-inverted-breve, then the dot would be badly positioned, and the s
> would not be below the souble-inverted-breve; or
> * t+dot-above next to a simple s, and a double inverted breve above all.
> I see no reason why the dot would move in the center between the two base
> letters. This is an unexpected deviation from the expected behavior of the
> dot above.
> How can the double inverted breve pass "over" the dot above to link with the
> s encoded after it?
>
> So I would think:
> t + double-inverted-breve + double-dot-above + s
> But there's still no such double-dot-above...
Michael's suggestion seems logical, the dot above is above the double diacritic.
It shouldn't be any different than any other stacking of marks, it could
(theoretically) be handled in OpenType with mark-to-mark positioning.
This would enable the dot to be centered with respect to the double
diacritic.
But, the canonical combining classes of the two marks are different:
234 canonical combining class for double inverted breve
230 canonical combining class for dot above
So, it looks like canonical ordering would make the dot above apply
to the "t" instead of the double diacritic.
Best regards,
James Kass
This archive was generated by hypermail 2.1.5 : Mon Feb 11 2008 - 12:06:58 CST