On 16 Sep 2012, at 12:16, Gerrit Ansmann wrote: > First, there are stroked variants of ‹W› and ‹w› (see appended sketch). Since > combined characters do not qualify for unicode anymore, I first thought to > encode them using U+0337 (combining short solidus overlay). However, »LATIN > CAPITAL LETTER L WITH STROKE« and »LATIN CAPITAL LETTER L, COMBINING SHORT > SOLIDUS OVERLAY« are listed as confusable characters, which in turn confused > me. So my questions are: Do these characters qualify for a submission to > unicode?
They would. It would be interesting to know whether there were any Antiqua versions of that Sorbian orthography. > Second, there is an ‹a› with to vertically aligned dots above. Should this be > encoded as ‹a + U+0307 + U+0307› (‹ȧ̇› – ‹a› with double ‹combining dot > above›) or does it qualify for a new diacritical mark? I might find it easier to answer that if I could see an example. My initial thought is that 0307 0307 would be suitable. > Third, I noticed, that the positioning of diacritical marks on certain > letters is not straightforward. E.g., for a ‹b with acute›, the acute could > be placed either above the bowl or the vertical stem of the ‹b›. Am I > assuming correctly, that this disambiguity is not to be dealt with on the > encoding level, but on the font level, e.g., with glyph variants? Yes, unless the placement of the acute is meaningful (two different letters). In a normalized text one would most likely just decide where it goes and stick with that for all, if there is no meaningful distinction. Michael Everson * http://www.evertype.com/
