i18n@... wrote:

>
> suzmccarth wrote:
>
> > --- In qalam@yahoogroups.com, "i18n@..." <i18n@...> wrote:
> >
> > We are just discussing options, Barry. But this is likely not a
> > market driven situation.
>
> Well, OK, that is an interesting statement. Is it jsut theory then? I
> don't get it - It seems like ME has actual plans to deliver a keyboard,
> so he is not jsut talking theory. If others are, then it is no surprise
> there is nothing even approaching consensus, especially if this
> fundamental difference is not understood.

A keyboard that comes attached to a computer they can't use because it
requires English literacy, a nearly nonexistent commodity in the target
population.

> Personally, I think it is *precisely* a market situation - some people
> spot an opportunity to provide products and services to others and seek
> information on what ways such products and services could be delivered.
> That is a pure market driven situation, don't you think? It is precisely
> what inbound marketing departments in virtually all hi tech companies
> (and probably not hi-tech companies too) do routinely.
>
> > There needs to be conviction that if the
> > market has created *numerous* glyph-based keyboards for
> > Chinese, which is market driven, maybe there is something
> > behind this idea. It is the priniciple really. In a particular situation
> > it is not evident what will happen.
>
> But isn't that a restatement of the market principle? Build the first
> one, and then offer continuing evolution of features until the entire
> market is satisfied. Why worry about getting it perfect jsut right? The
> first version need only satisfy the early adopters, which we have agreed
> will be satisfied with the qwerty kb. What happens in the 2nd version
> and beyond is the interesting part, but we can't really say until we try.

As soon as it becomes associated with the tiny Anglophone ruling elite,
backs will be turned on it by the Vai-literate who are not Anglophone.
The Haiti and Uganda analogies are not so far-fetched.

> This is true for any language/writing system btw.
>
> >
> > I brought up handwriting input for Tamil a while ago - think it got
> > poohpoohed.
>
> Yeah, that was before I saw reports of the new IBM method I linked to -
> just reported a few weeks ago. It is a completely new method and I
> immediately thought of your needs when I saw it in the local paper. I
> wondered if it would be adaptable to other writing systems and what its
> learning requirements would be. Could it work in Tamil?

If it involves some sort of handwriting recognition, why wouldn't it
work for any script whatsoever? The very earliest OCR programs, I was
told, had to be laboriously trained, character by character, so
presumably they could have been trained to read, if not cursive Arabic,
then at least discreet Hebrew. But by the time I had OCR packages, they
came pretrained -- and couldn't even be taught to recognize diacritics.

I'm told that handwriting recognition actually works quite efficiently
these days!

> > I have asked kids how it compares for Chinese
> > and the answer is 'my parents like me to do it this way so they
> > know I am writing properly.' Hm. Not much fun. However, it may
> > be the best for Vai. I was wondering what squawk was all about
> > at www.yuvee.com
>
> Hmm interesting - maybe Nicholas can go to that MIT conference in a few
> weeks and report back...
>
> http://www.prfree.com/index.php?cur=index&action=preview&id=29157 has a
> detailed description...

--
Peter T. Daniels grammatim@...