Re: Difficulties of Tamil input; polylingual computing.

--- In qalam@yahoogroups.com, "Nicholas Bodley" <nbodley@...> wrote:

>
> Perhaps this is better asked privately to Suzanne McC., but I'm

wondering

> whether other Indic scripts have similar problems, so that when a
> satisfactory method of keying in Tamil is finally defined, it

might also

> serve to help with other scripts.

I think someone else could answer this better. However, I do use
Punjabi. It is different from Tamil in a variety of ways. There is
only one 'full-size' vowel that precedes the consonant, not 6, like
in Tamil. So in the India, the ISCII code standard input was
in "logical order' except for 'i'. That favoured Indic languages,
but not Tamil, Dravidian. Tamil used the TSCII code standard and
input in order of visual sequnce, with all 6 vowels preceding the
consonant in inut order. Then Unicode adopted the ISCII model and
idea for 'logical order' neccesary for collation and data storage.
Now windows ships with a code conversion tool for ISCII but not
TSCII.

The problem is that some people seem to think that inut in 'logical
order' can be done by everyone. Unfortunately not. Without
experience with an alphabet which segments and sequences this
abstract manipulation of phonemes is not as easy as it seems for
many people, qalamites apart, of course.

On the other points, yes, I plan try some other multlingual editors
soon. Thanks for mentioning them again. Obviously I need a little
help in that dipartment. By the way download aboriginal serif from
Alan Wood, maybe you already have, and visit the earth words at
http://www.hotpeachpages.net/lang/crees/text.html

Suzanne

>
> ===
>
> Mentions of Uniscribe remind me of how I set up a DOS machine to

operate

> with any of the ISO8859 character sets (up through 10), although

only one

> at a time. Codepage 819 offers total Latin-1 compatibility,

although the

> nice box-drawing graphics like those starting at U+2500 are

sacrificed.

> Doing so did require installing a few custom-modified commands;

the work

> was done by Kosta Kostis.
>
> Earlier, while reading the Early Music list, and being repeatedly
> disappointed at the mangling of European names in ASCII, I tried
> installing Codepage 850, and was unhappily surprised and

disappointed. Web

> searching led me to Alan Flavell's essays about the quest for the

elusive

> Codepage 819, and the adventure began. Apparently, Codepage 819

was

> created for IBM's minicomputers, maybe the AS/400.
>
> Later on, I became hooked on installing all the foreign-language

support

> (although not fonts!) I could get from Microsoft; I might yet

install Win

> 2K Pro (if it's not too great a security risk), and will likely

enjoy

> more, or easier, access to other scripts and languages.
>
> As a consequence of politely pestering Opera Software to extend

beyond

> Latin-1, I learned about RichEdit and various Uniscribes, as well

as WGL4,

> Core Fonts, and Arial Unicode (which, btw, was recently still

available at

> an obscure Web site). I've long known about, and read material by,

Jukka

> Korpela, Alan Wood, Roman Czyborra, Markus Kuhn, and others

equally worthy

> of mention. Along the way, I came across SC Unipad and Yudit, both

very

> nice polylingual editors.
>
> In general, it seems that both entering and properly rendering

South and

> Southeast Asian scripts is the last major language category to

become

> workable for many people.
>
> --
> Nicholas Bodley /*|*\ Waltham, Mass.
> Bitstream Vera: very good-looking screen (and more?) font [family]
> Opera 7.5 (Build 3778), using M2