From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Tue Jan 10 2006 - 08:34:21 CST
On 1/10/2006 12:59 AM, Kent Karlsson wrote:
>John Hudson wrote:
>
>
>>Anto'nio Martins-Tuva'lkin wrote:
>>
>>
>>
>>>Umlaut as "E" above in this 1978 mediaeval-themed Austrian postage
>>>
>>>
>stamp:
>
>
>>>< http://www.apilch.2in.de/briefmark17.htm >.
>>>
>>>
>>This is just a glyph variant of U+00D6 and would decompose to
>>U+004F U+0308. There is a
>>
>>
>
>And I still REALLY STRONGLY dislike this too liberal interpretation of
>glyph variability.
>
>
>
I concur, but with a rationale along the lines laid out by Karl
Prentzlin in his excellent
post on this thread.
>An "e" (or "E", that would be, just barely, within allowable
>variability)
>above (or slightly "inside") a letter be encoded as U+0364,
>COMBINING LATIN SMALL LETTER E, NOT as U+0308.
>U+0308, COMBINING DIAERESIS, should always be imaged as
>"two-dots-above-like".
>
>PLEASE!
>
>
I think that it is entirely correct to make fonts that show an umlaut as
an 'e' or 'E'.
What you should not be expecting is that such fonts will behave sensibly
in the
presence of decomposition - while generically, a high end application
needs to
allow for different placement for upper and lower case diacritics, it's
not clear
that there is an equally strong generic case for a different outline for
each.
But, assuming there is, and it is supported, the font would then still
be limited
to German (or any other language where the umlaut can be spelled with an
'e').
That's precisely the limitation that Karl tried to get at with the term
'decorative'.
U+0364 is unambiguously lower case - except if you are willing to not make
a distinction - similar to decrorative fonts that have two sets of uppercase
characters (although such uses would be better handled by a case
transforming
style attribute, something that is widely available nowadays.)
>The same goes for long s: it should always be encoded as 017F, LATIN
>SMALL LETTER LONG S, NOT as 0073, LATIN SMALL LETTER S. The latter
>should always be imaged as "round-s-like", regardless of font.
>
>
>
The main reason for insisting on that is that in the general case, the
long s cannot
be placed computationally, as the round s is used in some languages in
the middle
of some words, and we all remember the minimal pairs....so I won't
repeat them here.
If that was NOT the case, then your argument would be a whole lot weaker,
like that for (or against) the use of Greek final sigma.
>>long established tradition of writing the German umlaut as a
>>small E in display lettering
>>and typography, since before use of the teo dots the umlauted
>>vowel was indicated by
>>addition of an e, either after or above. Sometimes the E is
>>above the letter as in this
>>Austrian stamp example, other times it may be within the letter.
>>
>>
>
>You're correct in that what is now encoded as COMBINING LATIN SMALL
>LETTER E is the historical origin of what is now encoded as COMBINING
>DIAERESIS. That does NOT mean that a small (or cap.) e above is an
>acceptable glyph variant of a diaeresis. Just as: even though C is the
>historic origin of G, G is not an acceptable glyph variant of C (and
>v.v).
>Likewise for V and U. Or would you encode the V in that stamp sample as
>U?
>
>
The V that is an U, yes. For decorative fonts like that, there are
slightly different rules, as
the requirements of appearance predominates, not the semantic.
A./
>I do hope not.
>
> /kent k
>
>
>
>
>
>
This archive was generated by hypermail 2.1.5 : Tue Jan 10 2006 - 08:37:04 CST