jump to navigation

Global warming incident 55 Millon years ago 6 January 2006

Posted by Pravin in Politics, Science & Technology.
1 comment so far

My good friend Jeff Nolan has posted a link to this article that describes a large scale global warming incident 55 million years ago called the Palaeocene/Eocene Thermal Maximum (PETM), and writes “yep, it’s all Bush’s fault …”

Jeff, you are such a partisan. I love it when you let loose one of these. 🙂

The article doesn’t let Bush, or the Europeans who are far away from the targets that they themselves set in Kyoto, off the hook. All it says is that there was a huge global warming incident 55 million years ago probably set off by tremendous volcanic eruptions or by melting glaciers in Greenland. It doesn’t say anything about the current warming cycle.

BTW, the glaciers on Greenland are melting again.

I think it would be smart to start reducing our greenhouse emissions.


Vinod Khosla on the new, new, new thing – ethanol 4 January 2006

Posted by Pravin in Economics and Startups, Science & Technology.
add a comment

This piece in Business2.0 is about what Vinod Khosla, formerly of Kleiner Perkins, is doing next. He left Kleiner late last year and is now out on his own. The article points out that Vinod likes to focus on huge problems and fund companies that attack these problems (usually piecemeal). He did this in the optical neworking space by funding Cerent, Juniper and a bunch of others. The “big problem” is our dependency on foriegn oil, and his first step is to focus on ethanol. He feels that using agricultural waste to generate ethanol, as opposed to corn, is the way to go. I don’t know enough to make any judgements, but given his record, I wouldn’t bet against him.

Software Stocks – Pretty Lousy Performance 3 January 2006

Posted by Pravin in Economics and Startups.
add a comment

Bill Burnham has a whole bunch of great pieces on the performance of the software sector in 2005. He has four posts (so far):

Software Stocks
Top 10 Worst Performing Stocks
Top 10 Best Performing Stocks
Software IPOs: 2005 Year in Review

Some interesting tidbits:

  1. Only two large cap stocks did well: SAP & ADBE
  2. The software sector, as a whole, lost 10% of its market cap
  3. The only sub-sectors that are doing well are, Search, “Linux” and SaaS

The spread sheet that Bill has on the first link above is amazing. It has every software company, and all the statistics which can update automatically. It needs an Excel plugin that Bill has provided a link to.

Lee Kuan Yew on India – with commentary 23 December 2005

Posted by Pravin in Politics.
add a comment

Atanu Dey, who runs Deeshaa.org, has written a series of three pieces on a speech by Lee Kuan Yew (former Prime Minister of Singapore). The speech was the 37th Jawaharlal Memorial Lecture and Atanu analyzes piece by piece and shows how fabian socialism and Nehru’s ineptitude set India back by decades. I really admire Atanu for his views and his forthrightness. Here are the links:

Lee Kuan Yew – Part 1

Lee Kuan Yew – Part 2
Lee Kuan Yew – Part 3

This administration is out of control 17 December 2005

Posted by Pravin in Politics.
add a comment

Given today’s revelations, we need a deep and careful investigation of what this administration is doing. They seem to have violated the constitution and have attempted to dismantle the very core of American values. The amazing thing is that the President has the gall to be upset and angry about these revelations. It is disgusting and makes my stomach turn. I am sure Lord Acton is going to be quoted heavily today in many blogs. He is the one that said, “Power tends to corrupt, and absolute power corrupts absolutely.” He also said, “Liberty is not a means to a higher political end. It is itself the highest political end.”

Population Map 12 December 2005

Posted by Pravin in Politics.
add a comment

This site has a very cool population map of the world. The size of the countries are proportional to their population. So, India is bigger than Africa and the Indian sub-continent is larger than any other country or continent. Amazing. The site is a little difficult to navigate, but still worth checking out.

Commuting in a million dollar fuel cell car 2 December 2005

Posted by Pravin in Science & Technology.
add a comment

John Spallino in Los Angeles leases a million dollar Honda fuel cell car for $500 a month. It is a test car that Honda is using to get feedback on its prototype. He commutes in the car every day and uses it just like any other car. The only difference is that he has to drive to Honda’s HQ in Torrance to fill up. NPR has a great piece on this.

Brezinski: Democrats are a lousy opposition party 2 December 2005

Posted by Pravin in Politics.
add a comment

In this interview in the American Prospect, Zbigniew Brzezinski, former National Security Advisor, points out what a lousy opposition party the Democrats are. They have been intimidated and whipped by the current administration. They have offered no reasonable alternatives and have not taken the initiative. IMHO, they have failed the American people and don’t deserve their vote.

Brezenski says:

The Democrats have a responsibility vis à vis the American people: to act as an alternative and to provide a vision of a strategy that avoids the pitfalls of what the Bush administration has created. The fact of the matter is that Democrats failed to do that during the grand debate over whether or not to go to war in Iraq. To be sure, some Democrats can rationalize their decisions by saying they gave the president contingent authority, and he pushed much further and acted unilaterally. Nonetheless, the fact is Democrats, tacitly at the very least, and explicitly in some cases, went along with a presidential decision based on a case that was dubious at best and mendacious at worst. Some leading Democrats have even acted as if they wanted to be part of the Bush cabinet, helping him prosecute the war in Iraq. L’outrance, as the French would say.

The Microsoft stack won’t work for SaaS startups 27 November 2005

Posted by Pravin in SaaS.
add a comment

I posted this earlier on my blog on blogspot.com, but I decided to switch over to WordPress.

I read Sam Ramji’s blog entry on Microsoft’s SaaS startup offering with a great deal of interest. In this posting he asks for feedback and I sent an earlier version of this posting to him in email. However, after I wrote it, I wanted to start some discussions around this, and I thought I’d create a post for this (very sparse) blog.

I work for Ketera. We serve about 25,000 users. It is mission critical hard-core stuff. Just as an example, our Spend Analysis service has analyzed over $4 trillion (yes, with a “T”) in spend across 1,500 companies. Our Procurement customers run 100’s of millions of dollars per month through our system. Our customers are a mix of very large companies and mid-market companies.

One of the key challenges that my organization faces is the rapid acceleration in the frequency of changes in business requirements that directly affects our infrastructure. For example, the first article on AJAX was written in February of this year by Jesse James Garrett, and we are working on a completely re-designed UI, based on AJAX that we will start rolling out in Q1 2006. This is a phenomenal pace. This kind of rapid innovation in the enterprise application space is something new.

Even though it is hard to keep up with this pace at times, these innovations are a good thing for companies like ours. These technologies enable us to leapfrog our competition and create features and functions that differentiate us. Our key differentiator up to now has been our business model and rapid implementation (we can usually get our customers up and running within a few weeks or a month). However, because we deliver all of our functionality through a browser, our user experience is not as rich as a “behind the firewall,” thick-client application. AJAX will help us build a UI that can compete head-to-head with thick-client apps and erode our competitors’ advantages.

Because we are a startup, competing against larger companies, we have to adopt these innovations and role them out rapidly. So, just to be clear, this is not just speed of development. It is speed of innovation. This constraint has a large impact on the type of infrastructure we choose—more importantly, this is the reason why we don’t use an MS based infrastructure (more on this below).

The other key issue for us is cost. The effect of the subscription model on a company is huge. Unlike traditional enterprise software companies, we cannot go into a customer and do a large “cashectomy” as the first order of business. Our subscription model means that almost all of our customers are not profitable to us, for some up-front period (anywhere from 3-9 months). We need to watch our expenses like a hawk during this period―that is when we are most vulnerable. Furthermore, we cannot afford any kind of “growth tax”. What I mean by that is that any kind of per user, or per CPU charge is a non-starter. We need to absolutely make sure that our margins are protected in the out years/month. We need infrastructure cost models that start off extremely low and then go down to an even lower asymptote. This is why open source is so attractive to us. Free is good. It’s really good. I suspect that almost any SaaS startup has similar issues.

Finally, I would like to highlight a few observations from my experience with Microsoft. These are all generalizations, and I am sure there are specific counter-examples for each one of these, but in the aggregate this was my perception.

  1. Microsoft components are tightly coupled and force you into an “all or nothing” approach. That is, it is difficult to insert a non-Microsoft layer in the middle of the stack. A corollary is that it is difficult to deliver a Microsoft interface on top of a non-MS component (e.g. Python, Ruby, Java). This reduces flexibility.
  2. Microsoft does not deliver as fast as open source. Open source components tend to rev their releases relatively quickly. This is connected to the speed of innovation issue above. I was less than satisfied with bug-fixes from Microsoft. If I have a bug with an open source component, I usually get a fix pretty quickly. If it is low on their priority, I can fund the bug-fixing on a case by case basis. Whereas, if my bug is not a priority for MS, I’m in deep trouble.
  3. Microsoft does not give me much visibility into the development process. I can (almost) always, download the latest CVS snapshot of an open source project.

So, given the above short-comings, plus the fact that MS’s pricing model is fundamentally, mis-aligned with our cost model, switching over to a MS stack does not make a whole lot of sense to us.