AR update

[Kalvesmaki, Joel]

Greetings,

Build 15 of Athena Ruby is now ready for download:
http://www.kalvesmaki.com/athenaruby/

In addition to adding monotonic codepoint, the new build incorporates preliminary kerning. This is implemented as OpenType class-based GPOS kerning. At present it only supports default letter forms and related left/right shapes that can be classed with the default forms; kerning for the exceptionally shaped variants has yet to be developed.

The added codepoints mappings are:

0x0384 oxia (spacing; left positioning option for caps)
0x0385 dialytikaoxia (spacing; left positioning option for caps)
0x0386 Alphaoxia
0x0387 anoteleia
0x0388 Epsilonoxi
0x0389 Etaoxia
0x038A Iotaoxia
0x038C Omicronoxia
0x038E Upsilonoxia
0x038F Omegaoxia
0x0390 Iotadialytikaoxia
0x03AA Iotadialytika
0x03AB Upsilondialytika
0x03AC Alphaoxia
0x03AD Epsilonoxi
0x03AE Etaoxia
0x03AF Iotaoxia
0x03B0 Upsilondialytikaoxia

The documentation continues apace. The database of characters is much improved, but much remains to be done. There are now test pages that show the use of character variants that do not use the Private Use Area. I've had success with Firefox and Chrome, but not others.

Do pass on to me any questions or comments you might have.

Best wishes,

jk
--
Joel Kalvesmaki
Editor in Byzantine Studies
Dumbarton Oaks
1703 32nd St. NW
Washington, DC 20007
(202) 339-6435

[Edward C. D. Hopkins]

Dear Joel,

I am not proficient in InDesign (using CS6), so correct me or make
recommendations where necessary.

I was surprised to see several missing characters in the font. Despite the
fact the focus is on Byzantine texts, I suspect you will need additional
glyphs for numismatic inscriptions. Using BabelMap, I loaded Athena Ruby
Version 0.80 build 015 and found the following absent:
U+03DC GREEK LETTER DIGAMMA
U+03F7 GREEK CAPITAL LETTER SHO
U+03FA GREEK CAPITAL LETTER SAN
----------------------------------------
Listed on the Athena Ruby glyphs web page, but glyphs not found in the font
(http://www.kalvesmaki.com/athenaruby/ARdatabase.html):
U+03DE GREEK LETTER KOPPA
U+03DF GREEK SMALL LETTER KOPPA
----------------------------------------
Why are the Greek small letterforms (such as U+03DF GREEK SMALL LETTER KOPPA
and others) rendered in the same size as the capital letters?
----------------------------------------
The InDesign document opens with the error message: "This document contains
a link to a source that is missing. You can find or relink the missing link
using the Links panel."
I think this may refer to the doaks_logo.tiff. Can you correct the link (or
embed the item) and provide it to me?
----------------------------------------
In the InDesign document, you have used fonts that I suspect many beta
testers haven't purchased. InDesign complains they are missing when I open
the document. Can you change these to more commonly available fonts?
Garamond Premier Pro Semibold Caption
Garamond Premier Pro Medium Italic Caption
Garamond Premier Pro Italic Caption

I am very interested to try the variants and will report my results.

Cheers,

Chris Hopkins
www.sylloge.org

--

[Kalvesmaki, Joel]

Dear Edward,

Thank you for your good observations and questions.

The font is supposed to cover inscriptions (particularly coins, seals, and
weights) dated 325 through 1453. This should explain the absence of the
digamma, sho, and san (but if you know of any examples from this period,
please do let me know). On the principle that the term minuscule is
anachronistic for this period, the font has only uncial letterforms, bound
to both upper- and lowercase keystrokes. We considered binding them only
to the uppercase forms, but that seemed to needlessly invite complaints at
the expense of a questionable principle. I've clarified our manual to make
this dating issue more apparent.

The Koppa issue is interesting, and I can't replicate your difficulty.
What software and OS are you using?

Thanks for the suggestions on the InDesign files--I've made some
adjustments. Let me know if they don't work.
http://www.kalvesmaki.com/athenaruby/

>--
>
>
>

[Edward C. D. Hopkins]

Dear Joel,

Thank you for your response. I want to first acknowledge the huge amount of
good work you have done on this project and that my comments are meant to be
constructive criticism. That said, let me nibble around the edges....

I'm a bit better informed now that I have read the AR manual -- I missed it
first time through.

Reference the missing digamma, sho, and san: as a long time developer, I ask
you to reconsider. (1) The font will inevitably be used by more than
Byzantine scholars (plus Byzantine scholars who also like the font and work
in other eras). (2) It is much easier to add the three missing archaic Greek
letters now than to address their absence should they be found necessary
after the development cycle is complete. Unless the cost-benefit ratio
prohibits, include them in the initial release.

I understand your reasoning on the small letterforms. But have you
considered that you are not following the Unicode specification in the sense
that the small letterforms are not Unicode compliant? The font will be used
by many who will not know of the narrow definition you have adopted. I think
you will have more complaints that the "letterforms are wrong" instead of
missing. Your "questionable principle" is the Unicode specification and I
recommend you remove the non-compliant letterforms.

I was wrong in stating my concerns about the missing "lightning bolt" koppa;
it is not listed on your AR database page, but...

Allow me to revise. My concern is that modern koppa is missing from Athena
Ruby and scholars may have created other texts that use the U+03DE GREEK
LETTER KOPPA in their inscriptions. When they apply Athena Ruby font to
their inscriptional texts, or convert from another font, there will be a
missing character (or unfortunate operating system substitution). As you
know, some versions of some fonts mistakenly contain an archaic koppa
character at the modern koppa position (among them Aisa Unicode, Arial
Unicode MS, Georgia Greek, Minion Pro, Palatino Linotype). I suggest you
have a letterform at the U+03DE code point.

I downloaded your revised InDesign document. It still opens with the error

message: "This document contains a link to a source that is missing. You can

find or relink the missing link using the Links panel." Checking the links
panel, the doaks_logo.tif has a status of "missing", yet I do see the logo
in the InDesign document. Keep in mind that I know little about InDesign.
Can the logo be embedded instead of defined as linked?

The InDesign document still complains fonts are missing when I open the
document.
Helvetica:
- Default font substituted for missing font
Helvetica Oblique:
- Default font substituted for missing font
Times Italic:
- Default font substituted for missing font
My recommendation to use common fonts obviously doesn't help those using
InDesign on vanilla Windows 7 computers. The Windows 7 op system does
provide font substitutes, so the document is certainly legible. That small
issue aside, all the Athena Ruby glyphs in the InDesign document exactly
match the PDF document. Great!

An issue on which I need some education: I checked all my installed Unicode
fonts and each has a full set of glyphs in the U+0020 to 007E code points,
except Athena Ruby. My understanding is that all Unicode fonts should
contain the basic ASCII character set. Some quick experimentation showed
that operating system substitutions are being used for common characters
contained in inscriptions such as superscript numerals (footnotes),
parentheses, curly brackets, hyphen, question mark, low line, etc. There
also some marks available in BMP General Punctuation that are useful for
inscriptions, but not present in AR.

[Kalvesmaki, Joel]

Dear Chris,

Thanks for your acknowledgement and suggestions.

The cost-benefit ratio you mention was a significant factor in our
decisions. (Quality font development is not trivial.) That explains the
missing pre-Byzantine characters, not just the three you've pointed out
but also many of the characters in U+10140..U+1018F. If characters
couldn't be attested in Byzantine inscriptions (on hard surfaces or in
hard materials), or if they didn't correspond to our publishing history,
we didn't spend the resources. Thus the missing extra ASCII
characters--that would have been quite a significant extra investment. And
the only modern punctuation marks we included were those that had been
used in diplomatic transcriptions in our prior print publications.

But, that said, we plan to release the font under a CC-BY-SA license,
which allows you or anyone else to add to the set. DO will add new
characters to the font only as new publications needs emerge. I keep track
of requests for new characters, but only those that have attestation in
the 4-15th c. Mediterranean, and preferably with a reference to a
publication with a clear photo or line drawing.

Your idea of mapping of GIDs 153-155 also to U+03DE and U+03DF seems like
a good one. Let me check on that.

I'm not certain about your reasoning on Unicode's specification of glyph
shape in lowercase letters. As I understand it, Unicode does not dictate
glyph design. Their saying "characters, not glyphs" seems to motivate the
"may" and "generally" of the first paragraph of §4.2 on Character
Properties: http://www.unicode.org/versions/Unicode6.1.0/ch04.pdf. But
maybe I'm missing something. Could you point me to a position paper from
Unicode on this?

On the InDesign file, let's give that one another try. I've uploaded a
version that uses simply Verdana and Georgia, and I've embedded the file
now. Feel free to change and play with the file as you like.

Also, please note that much still remains to improve the database (e.g.,
references, ligature variants, etc.; the head comment of the XML file has
my to-do list). That said, I've just posted a new version that includes
quite a bit of new metadata on the symbols.
http://www.kalvesmaki.com/athenaruby/

Thanks for providing your feedback.

>--
>
>
>