Unicode 4.0 was released a few weeks ago and it enumerates
It clarifies some of problems not resolved before. For instance,
it tells us which of Mongolian variation selectors are ignorable
(actually, the other day I found that Unicode 3.0 explained which are
ZWJ/ZWNJ are not there because they have 'visual' effects on most cases, but
blank glyph list may/should keep including characters like that as it does now.
So, fontconfig blank glyph list will be a union of 'default_ignorable' and
some other characters.
Should we provide an automated way to extract the appropriate set of codepoints
from the standard Unicode file?
I've updated the list from the Unicode 4.0.1 spec.