He has also authored two books: CGI Programming with Perl and CGI Programming On the World Wide Web. He spends the rest of his time consulting for internet start-ups, writing open-source applications or technical articles, and hold your breath, competing in Track and Field events. Here in Delhi as a speaker in the India Internet World, Gundavaram talked to Economic Times in an exclusive interview. Excerpts:

What are the main trends in the wireless internet market?

Let me take you back in time a little. Companies have been toying around the idea of a Wireless-Internet for a long time now; witness the birth of Phone.com (nee Unwired Planet) way back in 1994. But, the wireless revolution as we know it today started to really pick-up steam in the summer of 1997.

Phone.com released the second generation of their wireless application suite, including UP.Browser and UP.Link Server. UP.Link Server (or gateway) provides network operators such as AT&T Wireless, GTE and Sprint, a solution to Net-enable their networks by performing the needed protocol and format conversions.

UP.Browser, on the other hand, is a microbrowser that is embedded into wireless phones, allowing users to receive wireless content and applications. Several months later, Phone.com , along with Nokia, Motorola and Ericsson, formed the Wireless Application Protocol Forum to develop an independent standard for the Wireless-Internet.

In the early months of 1998, the Forum ratified version 1.0 of the WAP specification. They had succeeded in developing a protocol for high latency, low bandwidth environments, capable of working uniformly across a range of fairly extreme situations.

Around the same time, NTT DoCoMo in Japan released a similar technology called i-Mode. Unlike WAP, which requires developers to use the severely limited language WML for display, i-Mode capable phones contain the compact NetFront microbrowser, capable of rendering cHTML (compact HTML), a subset of HTML versions 2.0, 3.2 and 4.0. The Wireless-Internet was officially born!

Past couple of years have seen a rapid proliferation of wireless content, mostly throughout Europe and Asia, but also in the USA. According to an NTT colleague of mine, there are nearly 10,000 i-Mode capable websites and 6 million users in Japan alone.

The numbers for WAP, however, are not as thrilling, mainly due to the fact that application developers need to redesign their content using WML. That brings us to today. So far this year, we have already seen much-improved phones, capable of displaying more than 3-4 lines of text, new development tools for implementing wireless applications, and a large dose of healthy excitement.

i-Mode will continue to explode in Japan, and WAP will catch on in a very big away throughout the rest of the world. Predictions differ, but there definitely seems to be consensus that by 2003, more users will access the Web from wireless phones than from their wired counterparts.

How do you see WAP/WML evolving in the coming years? How will it co-exist with 3G?

That's an interesting question, and one that's on the minds of many people. WAP, in its current state, suffers from a host of problems. First, and foremost, WAP requires the use of WML, as opposed to a subset of HTML, such as cHTML. This forces developers to spend a significant amount of time to redesign their applications to render WML.

Second, the protocol is flexible enough to allow a range of implementations on different handsets. While you may think this is a positive feature, it can actually be a nightmare for content providers. This, so-called flexibility, results in a dramatic increase in the number of interfaces that web developers have to design for.

Remember the days of trying to figure out if your content would render properly in Netscape Navigator and Internet Explorer? We face a similar situation with WAP, but multiplied several times over. Ideally, the solution to both of these problems is to create the content in XML, and use XSLT to format it for delivery based on the handset.

Third, there is an incompatibility between WAP's security implementation (WTLS, WAP Transport Layer Security) and that of the Web. In other words, secure data coming from the caller using WTLS must be decrypted by the WAP gateway, and then re-encrypted in either SSL or TLS before being passed to the intended destination. This means that, for a split second, the transaction is insecure.

Having said all of that, however, the advances that we have seen up to this point in the wireless revolution are nothing short of miraculous. The WAP Forum is working hard to not only resolve these issues in WAP 2.0, but also add new features, such as support for XHTML, an XML compliant version of HTML, and multimedia extensions (i.e. streaming audio and video).

So, when GPRS (2.5th generation) and IMT2000 (3rd generation) networks finally roll out in 2002 and 2003, WAP will be well-positioned to excel.

To sum it up, we can compare the early excitement generated over WAP to the release of the Mosaic Web browser in the early 1990s. Just as Mosaic slowly matured into Netscape Navigator, WAP will also transform into a lean, mean protocol capable of handling text and multimedia over both low and high bandwidth networks. In fact, it will probably bear little resemblance to what we see now.

Has XML been over-hyped, or are there unforeseen challenges which developers seem to have missed?

The quick and short answer is, a resounding NO! But, to truly answer this question, we must ask ourselves, "What is XML and what can we do with it?" XML is a meta language that allows us to create our very own language, with a set of custom-defined tags to add some much-needed intelligence and structure to our data.

Now, compare this to HTML, which simply contains a fixed set of tags for describing the 'display' of data; no intelligence, no structure! For example, XML allows us to create one set of documents and transform them on-the-fly for display on a range of devices, from a Web browser to a wireless phone, exchange data between set of distinct data sources on a platform neutral and 'self-documented' form communicate with distributed, heterogeneous applications using a unified messaging format.

As you can clearly see, XML's generic (and easy to use) syntax allows us to use it in situations that go far beyond the typical delivery models used by internet and Web applications. This is very exciting, and adds support to the fact that XML is not all hype, and is definitely here to stay!

However, by using XML in this manner, we are aggressively pushing the envelope on existing tools and infrastructures that were not built to deal with flexible content. Unfortunately, this will ultimately lead to other challenges.

Let's take the case of XML-RPC, and particularly Simple Object Access Protocol, which allows us to remotely interact with application code. These RPC mechanisms use HTTP as a full-blown peer-to-peer communications protocol, which totally goes against its purpose; HTTP is a publishing protocol with very limited two-way communications ability.

Of course, it's true that the HTTP protocol is well understood and widely deployed, but it is also inefficient, and not the best choice for exchanging short disconnected messages.Here are some of the other issues and challenges that come to my mind:

• MIME identifiers- MIME provides only two layers of identifiers (i.e.text/html, image/jpeg), but XML typically needs three or more.
• Performance - XML is platform neutral, which means we can't take advantage of platform-specific tricks, such as compression.
• XML is incredibly verbose standards - Will we ever agree on that!

What are the three most successful instances of harnessing XML which you have come across in the business world?

The number of small internet startups and big corporations using XML have grown tremendously over the last year, mainly due to the introduction of a large number of tools and applications, and the `now. factor.

It is now hip to use XML. I have several success stories in mind, namely Ask Jeeves, Food.com , and our very own lifeguru.com . All three of these companies use XML to exchange content and information with their respective provider partners.

By standardizing on XML, they achieve economies of scale for content integration with multiple sources, and the ability to separate content from presentation.

What are the key misconceptions you notice in the way companies are approaching XML?

There are many misconceptions in the way companies are approaching XML, but, for the sake of simplicity, I'll break them down into two main categories: a lack of understanding of XML's capabilities, and an overzealous attitude towards the use of XML.

As we discussed earlier, XML can be used for a variety of tasks: to standardize data exchange, facilitate e-commerce, and provide a consistent interface to enterprise data. However, there are companies who continue to look at XML as a markup language, one that will replace HTML.

By using a combination of XML and Extensible Stylesheet Language, it can become a replacement in a very loose sense. But, obviously, these companies are missing the big picture. One can hope that the arrival of XHTML, a clean combination of XML and HTML, will force these companies to understand more clearly what XML can and cannot do.

On the other hand, I've also seen a pure overzealous attitude by companies that have adopted XML, the 'now' factor that I alluded to earlier. They seem to view XML as a panacea, a solution for nearly every data management issue.

I will give you an interesting example. I have seen companies store their customer profile information in individual XML documents, where a standard relational database management system, like Oracle or Sybase, would be the much more efficient and logical solution.

Of course, when they have the need to exchange this data with their partners, they can always extract it from the database, and format it using their XML vocabulary, or syntax.

Which are the new platforms and tools that are emerging for XML publishing? What kind of standardisation movement is emerging along various B2B vertical areas?

Within the last year, an incredible amount of commercial and open-source XML tools and applications have become available. To see just what I mean, point your browser to: www.freshmeat.net and enter the keyword `xml.. You'll see numerous-mostly open-source- applications for developing and managing XML content, ranging from editors and content managers to parsers and application servers. Here is a brief list of some of the more interesting and useful commercial tools:

• XML Spy- www.xmlspy.com
• EXml- www.cuesoft.com
• XML Pro- www.vervet.com
• XML Authority- www.extensibility.com
• Visual XML- www.bluestone.com
• Astoria- www.chrystal.com
• Data Harmony- www.dataharmony.com
• Adept/Epic- www.arbortext.com
• QuickSilver- www.interleaf.com

Talking about the standardization, RFCs, working drafts, recommendations, specifications ... we're seeing more and more of these every day. For example, there are about a dozen active working drafts related to XML alone, ranging from XML Schemas and XML Path Language to XSL Transformations.

Though this is a positive sign that illustrates the excitement in XML and other meta technology, it also begs the question, `do we really need of all this?. As you may have guessed, it's near impossible to answer this question in general terms.

But, I can say it's going to be very difficult to bring about a consensus on the use of any particular technology, since developers will have many options available to them.

Unfortunately, standardization of vocabularies is not moving as well as expected. There is a high level of fragmentation and a great number of parallel efforts. A survey taken in January of this year shows that there are 124 different XML business vocabularies in use, development, or planning.

Unfortunately, vertical market vocabularies account for the largest number of entries in the survey, at a very significant 77. See http://www.xml.com/pub/2000/02/23/ebiz/index.html for more details on the survey.

There is one effort, however, that is very promising, and that is ebXML, a joint initiative of UN/CEFACT and OASIS. This standard will allow companies in different industries to share data such as inventory information or requests for quotes under a common Web messaging format.an XML-based, cross-industry standard similar to EDI. But, even without ebXML, by next year, the Gartner Group predicts that 70 percent of all B2B transactions will be done using XML.

Tell us, how did MechanicNet come into being and what exactly does it do?

Our vision is to become the internet foundation for the automotive repair and maintenance industry; the automotive aftermarket. We deliver applications to enable electronic communications, interactions, and transactions between mechanics, their customers and supply chain partners.

The company was founded in November 1999, and went online this February. Since then, our team has grown to an actual size of 15, and a virtual size of more than 200, achieved through relationships with industry giants. For example, our alliance relationship with ALLDATA LLC (a subsidiary of AutoZone) and Mighty Distributing have given us a surrogate national sales and support staff to quickly sell our services to the 300,000 mechanic shops throughout USA.

Our core products today include Web hosting and design, e-mail hosting, internet dialup connection, customer relationship management applications, online scheduling and various other demand driving consumer tools.

What is your vision of what the Net can offer a country like India?

I am absolutely thrilled to see the Net playing such a crucial role in the development of our country, from quicker and more efficient communication among common folk through the use of e-mail, to an explosion in the number of IT jobs, to widespread electronic commerce.

I never imagined, for example, that there would come a time when I would be able to communicate with my cousin studying in Pune on a regular basis, with such little effort. However, I'm most excited at the prospect of a burgeoning e-commerce revolution, in a country such as ours, where there is so much diversity, geographically and culturally.

Imagine being able to order an exquisite rug from a dealer in Jaipur, or intricate folk art from a local artist in Kerala, all from the convenience of your home in a metropolitan city. In fact, it is this type of localised and specialised e-commerce that is still highly successful in the US, despite the so-called experts clamouring that B2C is effectively dead.

Of course, for such a revolution to take place in our country, there must be an improvement in utilities, infrastructure and transport mechanisms. So, the question is not whether it will happen, but when.

Any plans of bringing MechanicNet to India and becoming part of this revolution?

Specific to my company, I think India could present us with a huge market opportunity via its exploding and aging vehicle population.auto repair and parts procurement will become a tremendous industry. Digital interaction and transactions will enable a deep penetration and broad reach to the customers that will come to depend on their functioning vehicles, even more so in the coming years.

As most broken down cars today sit in wait for parts and repair information, MechanicNet will deliver the empowering information and transactions required to get the cars back on the road. In a country like ours, the Net, and more specifically, mechanicnet, may be essential to the functioning fleet.