Re: How about a typology for input methods

--- In qalam@yahoogroups.com, Andrew Dunbar <hippietrail@...> wrote:

> --- suzmccarth <suzmccarth@...> wrote: > --- In
> > I just assumed that you were thinking of Devanagari,
> > the word Hindi to be precise, with its preceding and
> > following i.
> > Tamil has preceding e, ee, and ai but also o, oo,
> > and au have parts that both precede and follow, so
> > half the vowels are out of sequence. They are full
> > size characters, not diacritics. Lots of reshaping
> > too.
>
> It's not size but function which make the vowel signs
> of Tamil and the other Indic languages diacritics.
> You cannot write them without joining them to a
> consonant in a meaningful way.

Thank you for explaining this. In appearance a lot of the vowels in
Devanagari and Punjabi are smaller and look attached. In Tamil some
do, but many don't. But I see that they are joined at the editing
level. Thanks.

>
> > > whether visual or phonetic sequence is used, that
> > > has nothing to do with Chinese-like IME's.
> >
> > The new syllablic IME looks like Chinese IME, sort
> > of.
> > >
> > > Tamil has only about 50 basic signs (letters,
> > > matras, etc.), so it makes perfectly sense to
> > > assign one letter per key, as it is for English
> > > or any other alphabet-like keyboard.
> >
> > Well, more like 30 something. But the Tamil call
> > the aksharas their alphabet, so 247 units. These
> > are the visually constant and distinct units. A
> > phoneme, without support of a constant visual
> > support is too abstract to be processed by everyone.
> > Some people in any language community are never
> > able to manipulate phonemes easily, members of qalam
> > aside, as I am so often reminded.
> >
> > > > No, Microsoft was asked explicitly to provide an
> > > > IME in order to enable input of each visually
> > > > distinct akshara and Microsoft refused.
> > >
> > > Correctly so, IMHO. It's sound like a silly
> > > request
> >
> > silly to whom, westerners, come on...
> >
> > > > now some research centres are moving to
> > > > handwriting input and speech inut because they
> > > > are so dissatisfied with trying to input in
> > > > order of phonetic sequence.
> > >
> > > That sounds like moving from a bicycle to a space
> > > shuttle because one is dissatisfied with the
> > > height of the saddle. Isn't it easier to fix the
> > > height of the saddle?
> >
> > I totally agree with you here. But it seems that
> > there is a bit of a crisis - how can the less
> > literate keyboard with so many issues to be
> > resolved. It is the lack of visual input that
> > has precipitated this crisis.
>
> If you think this is a crisis, just try to imagine
> what a crisis it would be if Unicode had this visible
> order. Most of the normal kinds of data processing
> which is trivial for most languages, becomes extremely
> complicated if you make the internal representation
> match an illogical visual order. Sorting is the most
> obvious one. Basically, every program would have to
> implement these functions seperately for Tamil which
> now work for most languages without any extra
> handling.
>
> The input problem might not yet be solved to
> everybody's satisfaction, and not enough fonts, OSes,
> and rendering systems (Uniscribe etc) yet work
> properly with the Unicode encoding. But, once they are
> the other, harder problems will not even exist.
>
> The data processing and text processing problems may
> not be relevent to your needs at all but if Tamil is
> to "just work" in *serious software* without a Tamil
> expert reprogramming each application to work around
> visual encoding, it would make Tamil a 2nd class
> language for computing. That would be a much worse
> problem than the one being faced now.

Well, I am a very big fan of Unicode. I have just started to work
with someone in the accessibility domain who has encouraged me never
to post text as images again so I look to a future of posting
mulitingual text and including the language attributes. I will be
promoting Unicode for multilingual webpages at a conference soon.

However, I will have to present to other teachers and these teachers
will some of them speak Chinese, Punjabi, Tamil, etc. I can't be
using clumsy terminology.

>
> > Not that input in visual sequence is the answer. I
> > don't think there is a consensus yet on what to do.
> > >
> > > > FOR ME, typology and input method ARE related.
>
> Perhaps but not on a 1:1 basis. Thai and Devanagari
> and Tamil writing systems are related typologically
> but
> their methods of encoding in Unicode are not.

Yes, I can see that there are lots of other issues that may take
priority and sometimes historical reasons seem to predominate.

>
> > > I might agree, here.
> >
> > You agree that it is for me - but should not be a
> > norm in the industry?
> > >
> > > > The way I think about it, I see Tamil as having
> > > > syllabic characteristcs and then I can look for
> > > > the syllabic IME.
> > >
> > > I definitely disagree, here.
> >
> > Well, it happened. I looked for the syllabic IME -
> > it does exist.
> > What's to disagree?
> >
> > > On the other hand, it's not a problem for an
> > > operating system to ship with three or more
> > > different Tamil keyboard drivers, e.g., "Visual
> > > sequence", "Phonetic sequence", and even a
> > > crazy "Syllabic IME".
> >
> > Haven't seen visual sequence for Unicode,
>
> What does this mean? You confused input,
> representation, and fonts so often that it's very
> difficult to guess which you're talking about some-
> times.

There were several different code standards for Tamil, with
different types of keyboards as well, and many of these were
identified by font name. That is just how there were labelled.
There are so many different input methods for Tamil, so many
different encodings, different keyboards, fonts, input sequence,
transliterations, etc. and some of these have been associated with
each other.

I think Marco, whose quote that was, is really familiar with Indic
editing and supports a variety of input methods. - just like in
Chinese the different IME's have completely different functions and
charactersistics, very complicated. For Chinese, different people
use different methods for different reasons.

>
> > I think it might be hard to implement coding, I have
> > no idea here, maybe you could tell me how possible
> > this is - or maybe they just can't agree on this.
>
> It's possible, it's just a very bad idea from which
> Tamil users would suffer for longer than they have
> been suffering so far over input woes.
>
> > However, the syllabic IME has the ITRANS
> > transliteration built in and that is a disaster for
> > Tamil.
>
> It would be trivial to change the transliteration
> compared to working with a visual encoding. Think
> hours
> of work versus years of work.

Yes, now that my Uniscribe file has been found I certainly have no
interest in changing the coding. I just want to promote a better
understanding of how Tamil can be input by the less literate.

Suzanne

> Andrew Dunbar.
>
> > Suzanne
> >
> > >
> > > _ Marco
> >
> >
>
> =====
> http://linguaphile.sf.net/cgi-bin/translator.pl

http://www.abisource.com

>
>
>
>
>
> ___________________________________________________________ALL-NEW

Yahoo! Messenger - sooooo many all-new ways to express yourself
http://uk.messenger.yahoo.com