[Indic-computing-users] INTERVIEW: Hold your tongue...

[Frederick N. <fr...@by...>]

HOLD YOUR TONGUE...OR, WHY THE COMPUTER DOESN'T TALK IN INDIAN LINGOS

Ashish Kotamkar's dream is to unleash the power of Indian language computing
software to take the so-called IT revolution to areas it has left largely
untoched -- vast areas of rural India and non-English language users.
Working out of his company's modest office in Pune, he wants to link up with
various ICT initiatives across India, to see how he could fit in with
enabling their applications for local languages.

But though he's part of the Mithi Software Technologies (www.mithi.com), an
Indian-language technology firm in the central Indian town of Pune, Kotamkar
looks beyond business. He has an interesting insight into the the
far-reaching impact of useful solutions.

Besides the government, says Kotamkar, NGOs (voluntary, or, non-government
organisations) will be the biggest beneficiaries of local language computing
as and when it comes up in a big enough way. Excerpts from an interview with
Frederick Noronha:

-----------------------------------------------------------------------
FN: To begin, what's the regional language computing scene in India now?
-----------------------------------------------------------------------

Its absolutely fragmented, de-focussed and generally directionless. It's
still a long way away from getting into full-bloom.

-----------------------------------------------------------------------
FN: What are the factors that led to this?
-----------------------------------------------------------------------

The Indian language computing scenario is in this mess today, largely due to
the lack of a long-term vision on the part of policy-making institutions and
the short-term mindset of developers. 

Indian-language software developers, mostly font vendors, have been
concentrating on ready-markets, like publishing, and all their offerings are
tuned to the immediate needs of this market. Very few have ventured beyond
developing font-based solutions which could ride piggy-back on existing
English-oriented applications. Resultantly, we see very few applications in
local-languages that are truly home-grown, and cater to the nuances of
Indian languages and scripts. 

For instance, coping with Indian languages  requirements such as alphabets
combining, forming conjuncts and recombining in differing shapes.

Local-language computing in India has been driven by an obsession with the
wrong market -- the DTP (desk-top publishing) market -- which was never
going to be scalable. It's a dead-end. Real and important applications don't
grow out of DTP. You need to focus on the more 'serious' uses of IT.

-----------------------------------------------------------------------
FN: What do you see as the potential of using IT in regional languages?
-----------------------------------------------------------------------

Huge. The government and development sector are likely to the biggest
stakeholders in this space. They could be the biggest promoters, and
beneficiaries, of local-language computing in India. 

Governments have been using some local language solutions for some time. But
we can't say they're using these extensively, and certainly not for critical
applications like emailing. The development sector, however, is likely to 
take the lead in this.

-----------------------------------------------------------------------
FN: What local language applications do you see as having real potential?
-----------------------------------------------------------------------

There could (theoretically) be hundreds. But the major ones would focus
around three themes -- digital communication, content and collaboration.
These directions are going to drive the local language market of the future.
This is where the real, true-blue applications are going to come up.

-----------------------------------------------------------------------
FN: Which states in India are doing a fair job in local-language solutions?
-----------------------------------------------------------------------

(From here in Pune, it seems to be) Maharashtra, Tamil Nadu and Andhra
Pradesh, Karnataka.... not necessarily in this order.

Most early innovationS came up in Maharashtra. Work on local language
computing started almost three decades ago, first to enable typesetters.

(Firms like) Modular in (the western Indian Maharashtrian city of) Pune was
a pioneer in that. People here focussed a lot on designing fonts, trying to
get a wider variety of fonts, fonts which were more aesthetic. C-DAC and 
Modular captured the early markets when DOS and Unix were the prevalent
operating systems.

-----------------------------------------------------------------------
FN: Today we're caught up with a battle between ISCII-Unicode, the lack of
standards, the unavailability of Indian-language keyboards... (ISCII, or the
Indian Standard Character Interface, is the Indian equivalent of ASCII)
-----------------------------------------------------------------------

Yes. People didn't give enough thought to encoding, and didn't evolve nor
follow common standards. Data re-usability and inter-operatibility were not
seen as important issues here, unfortunately.

Since font-based solutions worked on top of existing English-oriented
applications, they threw up their own problems. To cope with this, there
were work-arounds. But since these English applications were not basically
designed to handle Indian-language situations, the font solutions had severe 
limitations in terms of processing the data (for example searching, sorting).

-----------------------------------------------------------------------
FN: So, in ways, we have major problems to cope with still....
-----------------------------------------------------------------------

Absolutely. The whole regional-language computing was geared initially to
print notices, invitation cards, etc. Now, it has moved beyond DTP. It's
moving to more of database-driven applications.

This is where computing has immense utility: in terms of distributed access,
collaboration among different people, reuse of data, etc. To do this
effectively, we have to find better ways of doing things. That's where
standards come in.

The aversion of most developers to adhere to established universal standards
like ISCII and Unicode has only added to the woes of Indian language
computing. This is probably the biggest reason why the Indian language
software industry has not seen major growth even after almost three decades
of existence. The here-and-now and standards-be-dammed approach of the
Indian language software developers has actually harmed the long-term
interests of the industry, and restricted it to minor pockets of users.

-----------------------------------------------------------------------
FN: How long would it take to find solutions, in your guestimate?
-----------------------------------------------------------------------

Solutions are already there. Standards were developed long ago. It's for
everyone to embrace them. That's not difficult. There are attempts already
happening in this space.

Interestingly, most are being driven by voluntary initiatives like
Indic-Computing. Sooner or later, the government has to get into it. Unless
the government plays a bigger role, it's not going to trickle down.

If the government sets down the framework, the local language IT sector
players will see a market opportunity, and people will start embracing
standards more readily.

-----------------------------------------------------------------------
FN: Tell us something about Mithi's work?
-----------------------------------------------------------------------

Mithi, a start-up by a couple of young computer science graduates, first
came out with a CAD application IN 1992. Though technically a good product,
it didn't fare well in the market, and was eventually sold to a German
company.

Next came DesignerNotes, a technical documentation product. We started going
around industries in Pune, and people there asked 'Why don't you enable
local language capabilities, else our shopfloor won't understand it?'

This prompted the firm to approach CDAC -- the Centre for the Development of
Advanced Computing, a Pune-based government research centre. This
interaction triggered off a formal tie-up for developing local language 
applications. In 1994 came Leap, an affordable Indian-language solution.

It's the first and still one of its kind, the largest-selling local-language
application (as against a font package) developed truly indigenously. Later
Mithi came out with Internet-based applications like the site mailjol.com
through which you can send out email in 11 Indian languages and English for
free. It remains a showcase product for us, though not a revenue source.

Mailjol gets users from over 120 countries, basically Indian expats. After
its launch, it sometimes had such a deluge of visitors that the server
crashed and we had to re-start it late night, on some occasions. That shows
that the demand exists. Absolutely.

To make it easier for people to send e-mail off-line (without being logged
on to mailjol.com, given rural India's slow connectivity) Mithi brought out
its 'Mailjol Unplugged' version. It's an off-line email client, like Eudora,
or Outlook Express. With it, you can do your e-mail composing and reading
off-line, and log-on only for a while to send and receive all mail.

We now want to change the way e-mailing is used in India. We are working 
towards taking away the complexities from the process, and are focussing on 
three crucial factors -- a low total cost of ownership, ease of deployment
and use and local-language interfaces.

We are also offering our toolkits and services to developers for developing 
world-class local language applications based on universal standards. 

Our goal is to provide solid frameworks for communication, collaboration and
content on universally-acceptable platforms, and later unify these
frameworks to enable seamless creation and sharing of digital
information.ENDS