Am Mittwoch, 30. Januar 2002 um 00:39 schrieb Philipp Reichmuth: PR> ... for example, in German hyphenation the consonant PR> cluster "ck" gets hyphenated as "k-k" under some circumstances. This PR> is a rule as well, but still it is a clear case where putting it into PR> the encoding by means of a hypothetical "UNUSUAL HYPHENATION SELECTOR" PR> would be a bit inappropriate.
This is a complete algorithmic decision. "Some circumstances" is practically identical to "using old (i.e. pre-1998) ortography" (at least I don't know a German compound word which first part ends in -c and which second part starts with k-). The new orthography hyphenates before the -ck. (Thus, the decision how to hyphenate "ck" is for the whole text, not for the individual position, and does not need to be marked there.) PR> I think most of these cases, including PR> the Fraktur problem, deal with _typesetting_ rules and should thus be PR> left to _typesetting_ software, i.e. the now-famous "higher level PR> protocol". The question is, are typesetting rules "part of the script"? (I mean rules in the sense of obligatory regulations, not guidelines). If yes, (in my opinion) the plain text must carry the information that is needed to follow them. If no, their execution can be left to higher level protocols (which then have to decide whether a word is a foreign word [to be set in Roman letters] or a name [to be set in Fraktur letters], such at least according to German typesetting rules). PR> Would this mean much of an advantage over selecting a different font PR> for the respective character by means of markup? The advantage is that you can encode text to be displayed correctly (i.e. according to the obligatory typesetting rules) in Fraktur as plain text. You even can display this text correctly in Fraktur or Roman without change (as you can encode a Serbocroatian plain text to be displayed in Latin or Cyrillic correctly without change). Fraktur and Roman are "script variants", not "font variants". Both "script variants" have a lot of fonts, but they are not fonts themselves. If you regard the typesetting rules as "part of the script", you can look at Fraktur as a script variant which has four cases: "upper/lower for foreign words" and "upper/lower for the rest". The former accidentily happen to look like the two cases of the Roman script variant; thus you can use a Roman font for these two cases and another "real Fraktur letter" font for the other two. Cases could be left to higher level protocols, but for good reasons they are not. -- Karl Pentzlin AC&S Analysis Consulting & Software GmbH M�nchen, Germany
