For example, X Tamil uses Bala's Madurai transliteration
scheme, and understands Madurai transliterated text if
it is enclosed within \bt and \et and understands
Dr. Malten's transilerated text if it is enclosed
with \BT and \et.
I have registered the Tamil font that is used in the X
Tamil softwares (see libtamil-1.2.tar.gz at ftp.stolaf.edu
in pub/tamil directory) with the X consortium. It is offically
called wn.tamil.1993--University of Washington Tamil 1993.
This wn.tamil font is the closest thing that we
have as a standard at least in the X world. A number of
X programs exists that support wn.tamil: Tamil Xrn,
Tamil Xedit, Tamil Xgopher (by Badri), Tamil Xmailtool
(by Badri), and Tamil Xtalk. Besides these X window related softwares,
there exists other softwares for printing Tamil
of laser printer like Tamil lpr and Tamil TeX utility (by Ravi Sundaram).
In wn.tamil, characters 0-127 are same as ASCII and thus, wn.tamil
fonts can be used for normal ASCII display. Characters
128-255 are used to encode Tamil characters and thus, the
same wn.tamil fonts can be used to display Tamil and English characters,
which is quite essential if we need to make use
of the millions of lines of software that has already
been written to support English language.
The Tamil X editor program that comes with the X Tamil library
assumes ASCII keyboard like the ones used in North America
and India, and thus, Tamil key entry has to be different from Tamil
keyboard. libtamil uses a scheme which is very easy to remember:
Type ESC to get into Tamil mode.
to get \btk \ettype kk, to get \btka \ettype ka, to get \btkA \ettype kA,
to get \btkow \ettype kw and so on.
Type ESC again to get out of Tamil mode to type English.
As you can guess, each Tamil letter requires 2 ascii key strokes.
The first letter is mnemonic to the mey and the second is
mnemonic to the uyir that make up a uyir-mey letter. If
the letter is pure mey or uyir then you type that mey mnemonic
or the uyir mnemonic twice respectively. For example, to
get \bt_t \ettype tt, to get \bto \ettype oo, and to get \btO \ettype oO.
Therefore, in conclusion, we should adopt the following approach:
(a) Communication:
Translate Tamil text using a well known transiliteration scheme
like Bala's Madurai, Dr. Malten's scheme or Dr. Srinivasan's ADAMI
(enclosed within proper escape sequences of course) and transmit
the transliterated text.
The receiver will decode the transliteration for display
or can read it directly if the user does not have one.
(b) Font encoding
We neither have resources not have time to rewrite all the
softwares that were written to support ASCII text. So, any
scheme that is used for Tamil font encoding should leave
ASCII intact so that we can use ASCII software without
changing much. wn.tamil 1993 fits the bill in this respect.
(c) Key Board Mapping.
If you are using a Tamil keyboard, then the point is mute.
If you are using an ASCII keyboard, which is the norm, then
any key board mapping should be very easy to remember and
type. An ASCII keyboard is used mainly for English and you
only use it very rarely to type in Tamil. Thus, there is
a need for mnemonics. The key board mapping used by
the X softwares in libtamil are very easy to remember and
very easy to type. Please take a look at it (there is
PostScript file in the distribution which descibes keyboard
mapping) and see for yourself how easy it is.
-Sekar
I'm writing a bit off the subject indicated by subject line.
I any standard you apply to etexts, there should be possibility to
-see the translitterated version,
-see the tamil symbol version i.e. the text you read normally,
-print it on paper.
TeX is good for these.
Writing the text:
-with english keyboard editor by typing TeX codes,
-with english editor by typing other translitteration format,
-a converter to TeX format is needed, if we would like to print the
tamil text to paper,
-to view the translitteration there may be separate program; people
in this group have used such program to read newsarticles -- how
it is related to TeX?
-with tamil keyboard,
-output might be 8 bit ascii codes; writer may view the text with tamil
symbols directly
-output might be translitterated code; writer may view the text with
-tamil symbols; recall the note about tamil-newsreader,
-english ascii text,
Reading the text:
-printing it to paper; this is best done with TeX,
-it is needed converters from translitterated code and from 8 bit code
to TeX,
-viewing the tamil text,
-as translitterated text
-I have seen a program which converts some translitteration code
to 2x2 matrices of 7 bit ascii characters representing the tamil symbols,
-people use tamil newsreader on graphics terminal,
etc.
I'm getting confused on what converters it is needed and which afterall
are possible to apply...
In newsgroups, the 8 bit code is no no; translitterated version is ok,
since it is readable without any fancy programs.
Juhana Kouhia
I like the idea of standardization of keyboard. Right now it is difficult
to use fonts from many sources because of the lack of a standard. I guess
if you distribute one font free and more people use it,
it may become a defacto standard!! I personally don't like transliteration
schemes. With windows fonts, you can make use of the power of windows
programs. I think the majority of the people who use tamil wordprocessing use
only PCs. So there is a need to standardize the keyboard for PCs.
If you want a standard, you may have to get the TN govt. involved in
this process. They can bring together people from the industry and
universities and they have the resources to accomplish this goal.
Prabaharan
>
> K. Kalyanasundaram
# In wn.tamil, characters 0-127 are same as ASCII and thus, wn.tamil
# fonts can be used for normal ASCII display. Characters
# 128-255 are used to encode Tamil characters and thus, the
# same wn.tamil fonts can be used to display Tamil and English characters,
# which is quite essential if we need to make use
# of the millions of lines of software that has already
# been written to support English language.
Yes. This I see as the most important contribution of libtamil.
However I do not think this is implementable with other operating
system platforms like msdos or macOS. Since in commercial word
processing products one will not have access to the source code, there
is no way to intercept the function calls and modify them the way it
is done in libtamil.
# The Tamil X editor program that comes with the X Tamil library
# assumes ASCII keyboard like the ones used in North America
# and India, and thus, Tamil key entry has to be different from Tamil
# keyboard. libtamil uses a scheme which is very easy to remember:
# Type ESC to get into Tamil mode.
# to get \btk \ettype kk, to get \btka \ettype ka, to get \btkA \ettype kA,
# to get \btkow \ettype kw and so on.
# Type ESC again to get out of Tamil mode to type English.
# As you can guess, each Tamil letter requires 2 ascii key strokes.
# The first letter is mnemonic to the mey and the second is
# mnemonic to the uyir that make up a uyir-mey letter. If
# the letter is pure mey or uyir then you type that mey mnemonic
# or the uyir mnemonic twice respectively. For example, to
# get \bt_t \ettype tt, to get \bto \ettype oo, and to get \btO \ettype oO.
The same way, the keyboard assignment of X-TAMIL (in spite of being
the most intuitive and the easiest) will not be implementable in
msdos/ms windows or mac word processors simply because each keystroke
involving normal keyboard characters will produce a separate
character. Thus with respect to these platforms, one should evolve
another set of keyboard mappings but one can use the same
transliteration scheme. The user need not even bother about the
transliteration scheme involved and that will be left to the various
client programs.
# Therefore, in conclusion, we should adopt the following approach:
# (a) Communication:
# Translate Tamil text using a well known transiliteration scheme
# like Bala's Madurai, Dr. Malten's scheme or Dr. Srinivasan's ADAMI
# (enclosed within proper escape sequences of course) and transmit
# the transliterated text.
# The receiver will decode the transliteration for display
# or can read it directly if the user does not have one.
Correct.
# (b) Font encoding
# We neither have resources not have time to rewrite all the
# softwares that were written to support ASCII text. So, any
# scheme that is used for Tamil font encoding should leave
# ASCII intact so that we can use ASCII software without
# changing much. wn.tamil 1993 fits the bill in this respect.
A very good idea. However this is where one would have problems in
case of commercial WYSIWYG word processors. There must be a way to
switch to the tamil mode and start typing with the following keyboad
mapping that you suggest------
||
||
\/
# (c) Key Board Mapping.
# If you are using a Tamil keyboard, then the point is mute.
# If you are using an ASCII keyboard, which is the norm, then
# any key board mapping should be very easy to remember and
# type. An ASCII keyboard is used mainly for English and you
# only use it very rarely to type in Tamil. Thus, there is
# a need for mnemonics. The key board mapping used by
# the X softwares in libtamil are very easy to remember and
# very easy to type. Please take a look at it (there is
# PostScript file in the distribution which descibes keyboard
# mapping) and see for yourself how easy it is.
I am not sure about this but is it possible to add a new mode to the
existing commercial word processors? For example one can start typing
equations in microsoft word in macs (that is the only one I have used)
using a certain keyboard combination and entering the equation mode.
If it is not possible to do it easily, then it is not worth it. There
must be some other way of having a normal english mode and a tamil
mode within the escape sequence.
--badri.
# -Sekar
#
--
--------------------------------------------------
S.Badrinarayanan
Graduate Student
Department of Mechanical and Aerospace Engineering
Cornell University
--------------------------------------------------
I am used to typing on the standard Tamizh typewriter keyboard at high speeds,
and I find it very annoying to transliterate everything before I do the actual
typing. For example, it is so much more convenient to type "miyfl;y Xir" than
to type "alaikadal Osai" (after transliteration). However, my use of Tamizh
typewriting is limited to email messages, at best.
Therefore, I'd be interested in knowing if there are programs that
perform an "on-the-fly" transliteration from the standard typing to the ASCII
version (for example, to generate valid input for "madurai").
Thanks,
--prth (that's the output of typing "Siva" on
the QWERTY keyboard, imagining that it is a Tamizh keyboard!)
The above three are very important in electronic communication
that normally uses ASCII.
: Writing the text:
: -with english keyboard editor by typing TeX codes,
: -with english editor by typing other translitteration format,
: -a converter to TeX format is needed, if we would like to print the
: tamil text to paper,
: -to view the translitteration there may be separate program; people
: in this group have used such program to read newsarticles -- how
: it is related to TeX?
What ever form is chosen must be easy to use and intuitively obvious.
That is, a user should be able to guess the ASCII key sequence for a
Tamil letter. Such is the case with Madurai transliteration scheme:
o You type two characters for every Tamil letter.
o The Tamil text is saved in the Madurai transliterated form
to be sent over the internet.
o Converters like tamil-lpr or m2t can be used to print the
Tamil text on a PostScript laser printer.
: -with tamil keyboard,
: -output might be 8 bit ascii codes; writer may view the text with tamil
: symbols directly
With Tamil keyboard, there already exists Tamil typewriter
standard and thus, there is no need for another standard.
If you are talking about standardizing the character
encodings, you have got a point.
: Reading the text:
: -printing it to paper; this is best done with TeX,
: -it is needed converters from translitterated code and from 8 bit code
: to TeX,
Or you can directly convert the transliterated text into
PostScript for direct printing. With the advent of WYSWYG
text processing utlities like Frame, not very many people
will be using TeX. In such cases, direct printing will be
very helpful. Such an utility already exists as part of
libtamil.
: I'm getting confused on what converters it is needed and which afterall
: are possible to apply...
:
: In newsgroups, the 8 bit code is no no; translitterated version is ok,
: since it is readable without any fancy programs.
Key board mapping
|
|
|
V
Tamil Text in Tamil Fonts
/ \ ^
/ \ \
/ \ \
V V \
PostScript Transliterated Text
As you can see from the figure, you need only four converters.
Such a scheme is successfully used in libtamil.
-\btsEkar.
\et
| Or you can directly convert the transliterated text into
| PostScript for direct printing. With the advent of WYSWYG
| text processing utlities like Frame, not very many people
| will be using TeX.
1. Cost of software, access, bandwidth.
2. Accessibility (Put the couch potatoes in this class)
WYSIWYG still has a long long way to go. ASCII just won't die,
certainly not in the near future.
Sridhar,
who will hopefully be making the transition from MS Word to TeX shortly.
--
Sridhar Venkataraman ASU, Tempe, Arizona USA sri...@asuvax.eas.asu.edu
<A HREF="http://enuxsa.eas.asu.edu:8080/~sridhar/places/">India page</A>