From: Frank Ellermann (nobody@xyzzy.claranet.de)
Date: Sun May 06 2007 - 10:37:34 CDT
Peter Constable wrote:
> Shall we say that sz-lig is not a character? Clearly it's too late
> for that.
Centuries too late, and Fraktur is anyway very unpopular for some
decades now. I wonder if the BI Leipzig approach 1956 was also an
attempt to avoid the loathed "SS" on their Duden.
> The question before us is, given that sz-lig *is* encoded as a
> character, what is the best encoded representation for the
> typeform in question, "SZ-lig" (as you refer to it)?
If you use <S zwj Z> it automatically has the disambiguation from
SS as in the famous MASSE vs. MASZE case. BTW, it's hard to find
other examples justifying this effort. So even if <S zwj Z> is
displayed as SZ it's still clearly different from SS. "Standard
spelling rules" aren't applicable after a decision to ignore them.
For the hyphenation of "SZlig" <S zwj Z> a simple solution could
be "don't". And for IDNAbis if folks insist on wanting szlig it
might be possible to add it to the exceptions, without inventing
an SZlig character as justification for szlig in domain labels.
Frank
This archive was generated by hypermail 2.1.5 : Sun May 06 2007 - 10:44:18 CDT