Saturday, July 01, 2006

This entry is out of considerable deliberation and after discussions with some brilliant friends who share the thoughts. The reason was because though I have the freedom (Thank God!) to write what I feel like, I am actually talking about decisions other people have made. And I respect those people, so it doesn't feel good to write this. Nevertheless, I have been having this thought and it was only adept that I put it into words.

So what is this about? I am primarily talking about business on the Internet.  Many of us have been unfortunate enough to witness the big dot-com burst at the dawn of this millenium. All of us realize why that happened. If you don't, here it is in simple words - too much money was poured into online businesses which did not have a strong enough revenue model to get profits.  A kid would understand that if you put your money into something, you should know how that would generate profit for you.  Take that and multiply it a zillion times. When too many people put in money into too many things and don't care or know about how that investment will generate profits, then too much money is in the wrong place. That causes deficits in other places, hence creating a false boom of prices, salaries and everything else. When that might seem like a boost for a while, it eventually collapses because after time there is no return and all the rich people who invested in the first place stop doing that. And the people who are yet to generate profits out of the businesses they ventured into suddenly have a reduced cash flow, increased costs and no profits. Crash!  There is very little to say after that point.

When everybody who was in business then learnt their lessons, not everybody was fortunate enough to survive the bust.  Today is a different world. Forced by the cost cutting exercises that many global companies went through during that phase, work moved to countries where "labor" was cheaper. That has resulted in the offshoring and outsourcing model that we see today - which has become so successful that it is not just for IT, but has gone further into business process outsourcing as well.

Looking back at the lessons we learnt during the bad phase, companies and organizations have to take strong decisions that have a solid revenue model backing them. It is this that brings me to the crux of this discussion - which is the path that Zoho is embarking on.  Zoho, a product of AdventNet, Inc., is launching a series of amazing Web 2.0 products that deliver a complete office suite on the Internet, with nothing more than a browser required.  Everybody I know has been awed by the amazing work done, specially on Zoho Sheet and were even surprised when I said that it was created out of a company based out of Chennai, India.  Zoho has a Word-equivalent called Zoho Writer, a PowerPoint-equivalent called Zoho Show and many other productivity tools. And Google, which recently announced Google SpreadSheet, just has an online spreadsheet program. There were friends of mine who wondered, "Why didn't Google just think of acquiring Zoho?" 

AdventNet, for people who haven't heard of them, is a remarkable organization based in Chennai. They have some of the most brilliant minds in the industry and some of the most innovative products.  They also have built amazing prototypes such as SQLOne - a product that I think can lead its way to context-sensitive search engines, if they get it right.  Just looking at www.adventnet.com awes you with the breadth of products and technical excellence and solidity of these products.  AdventNet also has had a reputation for strong senior management and for creating startups into magic companies. Also, evidently they still love startups and are an extremely confident lot in believing in what they do.  An example is Applibase or Jambav, the former founded by an previous founder of AdventNet, and the latter an amazing organization backed by AdventNet that can give you the real job satisfaction - combining the worlds of technology and social responsibility.

Given all the background, I must sound like an idiot to say Zoho is probably a huge step in the wrong direction. But I still have my convictions that somehow, AdventNet is not getting it right this time around. Zoho is no doubt out of this world, even in its current humble BETA form.  Also, you can realize by just visiting the site that AdventNet has not put up Zoho just for fun, but is serious about making money through the product.  Google might very well put their product out for free, license it to enterprises and keep it free for the general public, banking on ad revenue.  Zoho, on the other hand also looks like licensing a Professional Edition for serious users and keeping it free for the the general public.

Let us take a step back and take a look at the complexity involved in building something as comprehensive as Zoho.  They released well before Google SpreadSheets and are announcing a new product every other week.  Google is still limited-user and hence has users waiting to see what is in store.  Zoho, on the other hand, is even willing to "publish" charts for you as images hosted off their servers.  Now to have achieved all this, Zoho must have put in a huge investment to get to where they are.  Let us assume, looking at the size of the project, and having mentioned in a previous post that JavaScript skills are not easy to get, they must have had a team of senior developers working for at least a good year.  Taking a shot at an average salary of a developer (assuming a mix of senior developers, managers and juniors) at Chennai, I am assuming a figure of INR 4.5 lakhs, which is close to USD 10,000. (Oh yeah! why do you think outsourcing works?). Assuming a team size of 40 (considering the number of products) - that would work out to $400,000. Add to that an investment in servers, development software, testing tools, operating costs, etc. which might have run into something like $1,000,000 (Assuming salaries are a 30% of operating costs).

That means the whole investment in building Zoho in its current state might be around $1.4M. If you took a look at Zoho, they have miles to cross before it can be any serious competition to the likes of Microsoft Office, or even Open Office.  The reasons:

a. Bandwidth - a luxury only in countries like the USA where the "Updating" that occurs after every entry in a cell in Zoho Sheet will not matter. Organizations in India are heavily trying to conserve bandwidth given the large number of employees and hence might not really be keen towards having their major productivity tools web-based. 

b. Productivity - Which brings me to my second point.  The product's target segment is users who want to use a productivity tool such as Microsoft Office.  Notice the key word - productivity. How can productivity increase with tons of JavaScript running in a memory constrained browser window, making server roundtrips at beck and call as opposed to rich client applications utilizing system resources, hyper-threading, core duo, high power graphics cards, tons of RAM, virtually unlimited disk space, high-speed IO devices, etc?  The answer is in the question - productivity takes a hit! 

c. Connectivity - Zoho claims that it makes "global availability a reality". They should probably talk to their sales force and ask them if they would go for a customer presentation without an office product installed on their laptops and rely on Zoho Show and Zoho Virtual Drive to load up the presentation.  The Internet is not always available, well, yeah Wi-Fi, WiMax, blah blah, but no, not yet.  How could I make that last change in the costing slide when I am sitting, waiting for my presentation, just before the client walks in - and specially when I am in the customer's office!

d. Enterprise Needs - Question again. Would the AdventNet sales team use Zoho Show? (Would they?  Really?)  Or would their finance team create all of their balance sheets and tax calculations for they year on Zoho Sheet.  (Really, again?)  Or would they expect their technology teams to create 3000-page product and technical documentation on Zoho Writer?  If they ever did, I bet it is a good year away looking at the current BETAs and even then, it would be unbelievable if they did not have a single license or installation of a different office suite.

e. Reliability - How many enterprises would be willing to depend on a service provider to store data pertaining to the organization?   Even if Zoho planned a "deployment" model where they host their product in the customer network, several organizations will be bound by compliance and security restrictions that prevent using such an alternative.  Specifically, Zoho Virtual Drive makes no sense for organizations that have the least concern of protecting data and information.  And it is not just the organization - customers of the organization run audits to ensure their data is stored safely.

The only benefit, and I think AdventNet sees it as the prime advantage, is the pricing that works out far cheaper than Microsoft Office Standard Edition - almost 16 times cheaper if I consider the 500-user licensing mentioned at https://store.adventnet.com/jsp/fp.jsp?filter=10010&p1=10118 vs. the Microsoft Volume Licensing Advisor (http://www.microsoft.com/licensing/mpla/) which gave me a quote of $183,000 for 3 years.  But then, the question remains - how many 500-strong companies would consider AdventNet against the de facto leader in the office productivity market.  To add to Zoho's woes, Google has made its intentions clear of getting into the space and surely, Google knows how to attract customers even if some of their products have better competing alternatives.

Now, going back to my initial cost estimates at building Zoho, I am making another assumption that they still have another year's development to get it anywhere close to Microsoft Office (try adding a very large number in Zoho Sheet and generate charts, try working with the suite in IE 7 BETA 2, try cross-referencing values between sheets, try disabling ActiveX when using Zoho Chat - this was the most surprising - after looking at Zoho Sheet - you wonder - ActiveX!!?!, try understanding Zoho Creator, try looking for documentation, try export/import features, etc. etc.).  That would mean another $600,000 (Average 20% hike for developers and some operating costs).  Total - $2M.

And then, I have not added hosting costs for the ASP-model, marketing costs, product re-inventing, re-writing future versions, investments for support infrastructure.

If AdventNet recreated magic, they would need to get 1000 organizations buying 500-user licenses. Or many, many more individuals licensing software.  In a year. Because then, they start incurring more costs.  From the looks of where Web 2.0 is, the concerns people still have and many other factors, it looks like a very, very steep ask. 

AdventNet will, needless to say, back it up with large investments and hope to make it really big. But in that, they are trying to compete against the product planners of Microsoft Office, the inventiveness of Google and the likes of Open Office.  It will be interesting to watch if they can pull this off.

If you asked me, they are headed the wrong way. I only pray that not too many companies invest too much and trigger another collapse. Let us hope, for us and for AdventNet, that I am wrong.

posted on Saturday, July 01, 2006 2:38:14 AM (India Standard Time, UTC+05:30)  #    Comments [0] Trackback
 Wednesday, June 28, 2006
I do this only for close friends. Seema and Ashwin sure are among some of my best friends. And Arun, well, qualifies as well ;) All of them also look serious about updating their blogs. I sure hope that continues. So, here's welcoming Arun and Seema/Ash. I have also added them to my blogroll. Arun: http://chearie.wordpress.com/ Seema and Ashwin: http://tinybitoflife.blogspot.com/
posted on Wednesday, June 28, 2006 1:01:45 AM (India Standard Time, UTC+05:30)  #    Comments [1] Trackback
 Thursday, June 15, 2006

Web 2.0 has no doubt arrived. What initially looked like a spate of XML/HTTP-driven websites, now has terms like AJAX (which no doubt was a marketing genius) and Comet making its way into programmer skill-sets.

There was time, back in 1999, when I was just an year into programming, when JavaScript was dreaded.  Nobody was interested in writing more than a few lines of client-side validation. I remember writing what we called a "double dropdown", which basically meant having two JS arrays and then populating a HTML SELECT list based on the selection of the first SELECT list. Like in a country-state selection. Doing that bit was a niche skillset that a couple of us could achieve with minimal code. All because of JavaScript illiteracy.

And then I had the opportunity to work on an IE-only intranet application (oh yeah, there was Netscape 3 and 4 to support otherwise) and thankfully we persuaded the customer to go with IE 4 as the base platform. Doing that application taught me DHTML and JavaScript like never before. In fact, after that I could look at JavaScript to do inventive things. I have had the privelege to code using Remote Scripting, Design-Time Controls (VSDTC) and Visual Basic DHTML project templates - all pioneers in making it easier to code in "AJAX".  However, the best of those days was coding in Notepad.  XML HTTP, which made its debut in IE 5.0+, was something I got introduced to only in 2003.

Why all this?  To give you an indication that the Web 2.0 AJAX thing was always around. But somehow, due to the then backward nature of non-Microsoft browsers, non-standardization and the usual time for user adoption and comfort with new trends, it took a while to appear the way it has now. However, Web 2.0 is not just about AJAX. This is a key thing to remember. Check out Manoj's excellent post on Web 2.0 - a great perspective that I share too and hence am not going to repeat things here.

So, now that the background is set, let me get to what I wanted to actually talk about.

Though there are a variety of new innovative websites out there that harness Web 2.0 technologies and ideologies, it is still quite a small percentage of the larger web. One of the main reasons still remains that JavaScript, that plays a pivotal role in creating powerful interactive websites, is still a nice skillset. Excellent programmers hate coding in JavaScript (take Sid for instance - got to link to that awful post on AJAX here - Sid, place the link in the comments. Interestingly, the company Sid is part of is building something really cool on AJAX to make it easier to build mash-ups). 

Now do you smell an opportunity?  Everybody wants to be on Web 2.0, but JavaScript is a road block. And even if you love coding in JS, writing reams of code is a no-no for productivity. Of course, the larger corporations of the world did sense this and the result is a barrage of "AJAX toolkits" trying to ease the burden.

There is Microsoft ATLAS, Yahoo Developer Network, Google Web Toolkit, the Dojo toolkit, the Prototype toolkit, just to name a few. And there are tools and parsers - you will find plenty of them.

Though all of these are around, the developer view of Web 2.0 is still very confusing. The reason - each one has its own way of solving the problem. Some are even solving the wrong problem.

ASP.NET 2.0 introduced script callbacks which in itself is a great way of writing AJAX-style code. But then many of the associated controls do not have the AJAX capabilities built-in and still rely on postbacks. ATLAS aims at solving this problem - hence enabling developers to write ASP.NET code (read server-side) without having to worry about JavaScript and client hacks. Neat. But this causes two problems - one, it is tied to a server platform and two, it will take longer to mature because Microsoft has assumed that the developers know ASP.NET 2.0 already, something that is not necessarily true in all cases. Moreover, it is taking just way too long to release. I am afraid they might well miss the curve if something really better comes up.

Google Web Toolkit (GWT) works with code generation. Write in Java, generate JavaScript. Though I haven't tried GWT personally, I am pretty sure this is not a great approach. The problem is that no matter how well you write your code generation, purists will tell you that it limits the language that is being generated. If I wanted to do something one way and the tool generated in another way, it is always going to be a problem. Frankly, I had expected Google to come up with something targeting the Microsoft developer base primarily, but coding in Java will limit the user base of GWT as well.

Yahoo UI, Dojo and Prototype are all nice client libraries. In fact, I like Yahoo UI a lot for its simplicity. Dojo is really powerful.  Both of them expect you to understand how JavaScript work - which could very well mean it is for developers who are already comfortable with JavaScript.

Tools like the Microsoft Research Map Cruncher and Google Sketchup are really nifty tools that can be used to do some good innovative Web 2.0 programming. But each of them are again limited to a particular context, particular website, particular technology, etc.

What is missing is a comprehensive IDE for developing Web 2.0 applications. A platform for creating, mashing up, merging, collaborating and developing applications using innovative ideas. Sites like Codeplex should aim at not just being another source control repository, but instead creating an environment, providing the right tools and enabling developers to code collaboratively. That will be the true realization of the Web 2.0 vision.

posted on Thursday, June 15, 2006 3:54:03 PM (India Standard Time, UTC+05:30)  #    Comments [0] Trackback

Getting a little time on hand is a rare thing these days. (No) thanks to a flu that has grounded me, I have a little time now as I am recouping with dozes of green and blue pills. Should be back on my feet in a day.

However, since I had a little time, I thought I will make some good use of it and put an entry in - something that also has become a rare thing. And I already have people complaining about it - which, needless to say, is a good thing to know.

So before I make a post that has been running in my head for a long while now, let me also put in a few lines about time and the meaning of "no time for blogging", strictly in my own context.

I hate to say I have no time for blogging. Many of my friends say, how can you be that busy - it takes just a few minutes to blog an entry. But then there are other friends of mine as well who like to roll an idea in their heads for a while before it is blogged.

Weblogs have become more than a fad now. Anybody who spends considerable time online has a blog. There are different categories of blogs (and no, I am not talking about categories like tech, programming, literary, etc.). Categories created by people simply because as humans we fall into different categories ourselves.

There are the simple bloggers, that use their blog space as a hyperlink aggregator, sharing with people all links they found interesting. Many people also start blogging this way. Then there are the ones who build upon this a little, add a viewpoint or two.  And then there are the ones who get linked because what they have put in is pretty much an original post or a different viewpoint that others find interesting.

There was a time when every link I found interesting, I wanted to pounce upon it and share it with friends. That interest has waned now. I no longer want to do the "check this out.." posts.  I want to do more of the viewpoint stuff. Which is why I need more than a few minutes to do a blog post. Which is why you see the number being scanty.

I only hope I had more time on hand. Maybe I should take up a full-time blogging job - on the lines of Robert X Cringely (by the way, Cringely is not a real person).

But till I get that kind of readership, I have to just keep sneaking time and hopefully without being sick.

posted on Thursday, June 15, 2006 2:41:16 PM (India Standard Time, UTC+05:30)  #    Comments [0] Trackback
 Monday, May 29, 2006

It has been a long time since I have made any update.  Just bogged down with some tight schedules and heavy work.

There are lots of exciting things I want to write about - specially around Web 2.0. But that will have to wait for a better time.

posted on Monday, May 29, 2006 5:26:02 PM (India Standard Time, UTC+05:30)  #    Comments [1] Trackback
 Thursday, April 13, 2006

OfficeLive - the new service from Microsoft - provides three subscription models. The first model provides you with a free domain name!  It sounded quite unbelievable at first!

So I went ahead with the steps - pretty simple - you choose your domain name, login with a Passport account (I think you need a Passport account tied to a USA address) and then came the surprise part. I was asked for a credit card number. Turns out, the credit card information is just for "verification purposes".

In other words that meant to me - well, get a free domain only if you are serious about it and if you trust us - after all, we are giving it to you for free!

So I gave my credit card details (was a little disappointed to see that they actually store the information and use it if you buy any of the paid services in the future), but I don't intend to sign-up for any paid services and I can pretty well trust Microsoft not to do anything stupid like go charge your card without asking you first.

What I saw after that was amazing!  The site builder tool is fabulous! Simply awesome! I would encourage everybody to go register only to get the experience of that tool. It took me just a few minutes to select a layout and create a simple one-page site.

Of course, it said the domain registration would take 24 hours - it took lesser than that. I now own another domain - www.pandurangnayak.com - but if you visited it right away - there comes the shocker!!!  It shows me the Apache/Red Hat test page!!!

That is SO NOT COOL - I know that the domain must have been registered with some external organization who has just pointed it to some default site. And there must be some batch job that will point it right. But it is so uncool to see a Red Hat Linux / Apache test page for a Microsoft service.

If somebody is listening, please fix it before the Slashdot guys have a good laugh and scribble some nonsense.

PS: If you want to see my temporary page, visit http://temp.pandurangnayak.officelive.com If you had questions about whether that is going to be my new home - visit the site and find out. 

posted on Thursday, April 13, 2006 4:40:17 PM (India Standard Time, UTC+05:30)  #    Comments [0] Trackback
 Wednesday, April 05, 2006
Bill Gates in an amazing story-telling mode of how he works!  Nice reading.
 
 
 
 
 
posted on Wednesday, April 05, 2006 9:58:58 AM (India Standard Time, UTC+05:30)  #    Comments [4] Trackback
 Wednesday, March 29, 2006

This stuff looked so damn good that I had to have it on my page. I don't like the pro-Google attitude, but will stick with it for a few days anyway!

posted on Wednesday, March 29, 2006 6:35:17 AM (India Standard Time, UTC+05:30)  #    Comments [3] Trackback
 Saturday, March 18, 2006

Must watch:

Marc McDonald is Microsoft's first employee. Hired back in 1976. Channel 9 recently caught up with him and talked about the old days at Microsoft as well as what he's doing on the Windows Vista team (he's working on defect prevention).

Funny story. He left Microsoft in 1984 because Microsoft had gotten "too big." Microsoft's size at that point? A few hundred employees. (He came back after Microsoft bought the company he worked for).

http://channel9.msdn.com/showpost.aspx?postid=111590

posted on Saturday, March 18, 2006 10:55:21 AM (India Standard Time, UTC+05:30)  #    Comments [0] Trackback