RE: UK

From: Emil HER©AK (emil.hersak@zg.tel.hr)
Date: Fri Nov 28 1997 - 15:22:45 EST


Thanks for the replies! This problem has been a practical one for me, since
I am trying to find some elegant why to reformulate the fonts I am usings (my
own fonts).

Naturally, I
>understand that differencing in letter form can be resolved by fancy fonts
>etc.

The form is the glyph. If you would recognize both forms as a
representation of the same thing (i.e. it means no difference which is used
in a given word) then in principle they are supposed to be unified as
characters and some process is to tell the computer whether it is to be
rendered as "oy" or as "8".

OK, but this would mean that both would have to encoded, or do I not
understand something?

No one really provides a mechanism for this. For Arabic, it is built into
the rendering engine at a high level. For simple alphabetic scripts like
Cyrillic, I don't know how it would be done. There are claims about "smart
fonts", but I never saw one. In some Tibetan software I have used you can
type a "variant" key to change its preceding character's shape (which is
not much different from having two characters in some instances since it's
not contextual. I have a lovely OCS book from 1861 here in front of me. I
seem to find "oy" used in word-initial position and "8" in other positions.
If that were a rule, then a smart font would be able to render correctly.
But experts like yourself must tell us.

Yes, this is particular to Arabic which has different forms depending on the
position of the characters. For OCS "uk" this is not the rule, but rather it is
a question of development of the letter. The form "oy" is older, and more original,
being a direct transliteration of the Greek "ou". The form "8" is a later ligature,
developed so that the "y" was placed on top of the "o". An even later habit was
to use just "y" instead of "oy", while is the case in modern Cyrillic (still quite
rare in OCS). Hence, depending on the time-period, "oy" or "8" will be used,
which of course "oy always in the capitals, if capitals are at all used (in the
earlier versions they are not).

Of course, QuickDraw GX was supposed to provide this kind of rendering, but
no one really implemented it and I hear it is being withdrawn. In favour of
what I don't know.

>but in this case the second form appears regularly in some later texts and
>not in others. Since or at least many of the other letters remain more or
>lessstandarised, it would make little sense to produce two separate type-faces
>with differing "uk" characters. It would seem to me best to provide one more
>slot in Old Cyrillic to cover this alternative form.

I would not fight strongly for a unification (which should come as no
surprise to readers of Unicode list). What do existing typefaces provide?
Are there any coded character sets with both?

Hmmm. This is the problem. I am using my own TTF typeface, which I produced
before I found out anything about Unicode. In the earliest version, when we were
using something called CROASCII in Croatia, I more or less had to position letters
as seemed logical in transliteration. I placed the regular "uk" (cyrillic "oy") on the
Latin "u" position, and positioned "8" on one of the alternative "u"-forms (I think on
one of the acute or grave accented "u" forms, since i used u-umlaut for "izhica").
Now that the Unicode positions for OCS can out, I formulated the typeface acording
the the agreed-upon positions, but I was left with "8", as well as a few alternative
forms for "jery" - which was less serious, since these were basically bi-glyphs
made from existing glyphs. Naturally OCS "uk" can be more or less produced
by doing just what the OCS scribes did, i.e. by writing two letters. In the Greek
encoding there is no special slot for "ou", the model for "uk". However, the Unicode
coding did give "uk" is own position, which is technically logical, since the letter
was perceived as one sign in OCS. This perception, I believe, probably lies at the
bottom of the invention of the ligature form "8" (while the capital remained, just as
capitals themselves was rare). Now at this point, I am wondering what to do.
Producing a special slot for "8" is beyond me, since it depends on general
agreement. I can imagine only two interim solution to getting "oy" and "8" into
the same typeface. One - place "8" in the lower case "uk" position and write
"oy" in the texts as "o"+"y". Second - place "8" somewhere in the privare use
area, just as "fi" ligature is now placed.

>2) Glagolitic. I would like to know if any work has been done so far on this
>historical script.

I am working on it with Hinko Muren in Ljubljana. Apparently John Clews is
working on it also.

Thank you for this information. I am very glad that work is being done and would
appreciate hearing something on the progress. I would, however, suggest including
some people from Croatia in the work on Glagolitic, since this script was used
for the longest period in Croatia, and hence the need for it is most apparent here.
I am somewhat included (or better said, it has been suggested to me to include
myself) in a project, which will begin shortly, to computerise the data base in the
library of the Croatian Academy. Glagolitic will definitely be encountered. The
script is also taught regularly in a summer school on the island of Krk, which
has a special position in the history of Glagolitic and Croatian vernacular literature,
due to the Bas^ka inscription.

All the best,

Emil Hers^ak,
Institute for Migration and Ethnic Studies
Zagreb, Croatia
emil.hersak@zg.tel.hr

--
Michael Everson, EGT * http://www.indigo.ie/egt
15 Port Chaeimhghein Íochtarach; Baile Átha Cliath 2; Éire (Ireland)
Gutháin:  +353 1 478-2597, +353 1 283-9396
27 Páirc an Fhéithlinn; Baile an Bhóthair; Co. Átha Cliath; Éire




This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:38 EDT