At 16:30 -0400 2005-08-22, Peter T. Daniels wrote:

> > >Do you really not see that you're requiring the
>> >billions of non-roman-users in the world to
>> >change their lifelong habits -- and their
>> >centuries or millennia of cultural tradition --
>> >for your convenience?
>>
>> I didn't invent Pinyin, or Pinyin-romanized input
>
>For the umpteenth time, pinyin isn't used by anyone after third or
>fourth grade. It's for foreigners.

Nevertheless I did not invent it and I am not requiring them to use it.

> > for Chinese. I didn't invent Roma-ji input for
>> Japanese. The Japanese did that. I didn't devise
>
>Japan had romaji before it had roman-input computers.

Nevertheless some Japanese use Latin-input on
their computers, and are not forced to do so by
me or any other Latin conspiracy trying to crush
their cultural tradition for reasons of
convenience.

> > the QWERTY-based keyboards for Devanagari or
>> Arabic or Hebrew. Apple did that.
>
>Hebrew is irrelevant; aside from its direction it presents no special
>challenges to the typist. (Israelis react almost violently to the
>inclusion even of the sin/shin dot in Modern Hebrew typesetting.)

Nevertheless, Apple ships both QWERTY-based as
well as non-QWERTY-based keyboard drivers for
Hebrew.

>The Arabic keyboard bears no relation to QWERTY, but Apple's WorldScript
>I took care of all the contextual forms automatically as well as the
>most common ligatures. (It left out several characters needed for fully
>pointed text, however, such as dagger alif and hamzatu l-wasl.)

Nevertheless, Apple ships both QWERTY-based as
well as non-QWERTY-based keyboard drivers for
Arabic.

> > I did devise QWERTY-based keyboards for Inuktitut
>> and Cherokee, alongside other keyboards. Both
>> have their uses.
>>
>> I have devised a QWERTY-based Vai keyboard which
>> seems to work very well. A non-QWERTY-based Vai
>
>Seems to whom? How many Vai people with computers have checked it out?

It seems to me to work very well, in that with
it, I am able to type Vai text accurately with it
without much effort or difficult. I am developing
this software, and Vai isn't formally encoded
yet, and so no one else has "checked it out" yet.

> > keyboard would be hard pressed to give users
>> access to 294 characters plus digits and ASCII
>> punctuation without deadkeys (since 48 x 4 =
>> 192). Moreover, Vai speakers (who number 105,000)
>> live in a country whose official language is
>> English, so it's not as though the Latin script
>> is unknown to them. A QWERTY-based keyboard for
>> Vai would hardly be a curse upon them.
>
>It would hardly be optimal.

Oh come on, Peter. Follow the argument.

I say that a QWERTY-based keyboard for syllabic
input (where s + a = sa) will enable Vai users to
type all 294 Vai characters plus all the digits
plus all the ASCII punctuation (needed for URLs
and internet access) plus typographic punctuation
(like smart quotes) on the hardware which they
*will* be using, that is, US/UK QWERTY-engraved
47- or 48-key keyboards. Into the bargain, the
Vais live in an English-speaking country, and any
Vai who uses a computer will already be literate
in English, or will have to become so if he or
she wishes to use any software in the foreseeable
future. (One would like to see basic software
like Firefox localized into Vai, but it would be
unrealistic to hope that much more will ever be
localized and maintained for this market.)

Would you like to suggest another configuration
in which 340 characters can be squeezed into a
keyboard layout based on another model, with 192
positions available in 48 x 4? To get more, you
will have to sacrifice some of those positions
for dead keys. That's doable, but what will the
layout principles be?

>What is so offensive is your assumption that the Western way is the best
>way -- and that you refuse to recognize that it isn't.

I haven't said this. I have taken pains to point
out to you that I devised *both* QWERTY and
non-QWERTY keyboard layouts for both Cherokee and
Inuktitut, and that I recognize that there are
Cherokees and Inuit who have first learned to
type English and by whom QWERTY is considered
easier to learn than another layout.

> > >Shift was devised for a particular quirk of
>> >contemporary roman, cyrillic, and greek.
>>
>> Shift was long used in other traditions. For
>
>You have a curiously short notion of "long."

These things are relative.

> > Arabic shaping fragments, for instance. The 1962
>> Standard Hindi Typewriter used shift states to
>> access half-forms in some instances (MA, M-) and
>> different letters in others (U, TTA; DA, DDA).
>
>This is not 1962. In 2005, computers can be expected to do all the
>automatic shape-changes automatically.

What I was pointing out was that the shift state
was used in India long ago, by Indians, and was
not a quirk imposed upon India.

> > >Option isn't used for any ordinary English
> > >characters, and Shift-Option is an immense
> > >imposition on the typist.
>>
>> Façade, naïve, résumé. All English words
>> correctly spelled with diacritical marks. In
>> Ireland we use option for áéíóú and shift-option
>> for ÁÉÍÓÚ and this causes no particular hardship.
>> (I say this as one who typesets books in Irish
>> regularly.)
>
>All those words are correctly spelled without the diacritics.

Not in good typography, they aren't. Consult
Bringhurst or any other work on good typography.

And Irish typists do not feel imposed upon by
having to type Shift-Option for ÁÉÍÓÚ. (I repeat
this because it refuted your point that
"Shift-Option is an immense imposition on the
typist".)
--
Michael Everson * http://www.evertype.com