[XeTeX] : Re: wrong uccode of ß

Adam Twardoch list.adam at twardoch.com
Mon Apr 14 16:26:21 CEST 2008

Ulrike Fischer wrote:
> Replacing \char223 (the uppercase version of
> the ß) by two distinct S  may look similar, but it changes the
> orthography and so the search now fails.
Yes, it is an interesting discussion. The question is whether the 
sequence of letters is more important than casing or less. Setting 
all-uppercase text is done through two kinds of transformations: you 
either enter the mixed-case characters into the backing store and use 
some sort of visual transformation (e.g. an OpenType Layout feature) so 
the resulting glyphs are all-uppercase, or you enter all-uppercase 
characters directly.

It's basically the question whether all-uppercase text is part of the 
content or part of the form. In many cases, I think it is part of the 
form (all-uppercase titles etc.), and then it is useful to store the 
mixed-case backing store because it preserves the semantic distinction 
(e.g. between common words and proper names). Sometimes, however, it is 
better to keep all-uppercase characters (the most obvious situation is 

A problem with digraph glyphs is e.g. that if you letterspace the text 
(which is often the case when you set all-uppercase text), the digraphs 
would typically stick together, so ou get something like:


While it might be orthographically desirable, it simply does not work 

I, for my part, consider the current German orthography imperfect in 
that regard. Since 1996, the lowercase letter ß is simply a single 
letter (e.g. because the rule of doubling consonants after a short vowel 
has been clarified). A situation where uppercasing a single letter 
produces two letters is anachronic, counterintuitive and confusing. I 
believe that in 50 years or so, the majority of users will adopt the 
uppercase ß letter — if type designers will provide it in their fonts.



Adam Twardoch
| Language Typography Unicode Fonts OpenType
| twardoch.com | silesian.com | fontlab.net
| Marchlewskistr. 24 | D-10243 Berlin | Germany |
| mobile +4917620114051 | Berlin +493021237390 |

More information about the XeTeX mailing list