lesscode.org

I just had a chance to finally listen to the latest Gillmor Gang…

UPDATE: I goofed and ended up listening to the June 2004 episode. It’s still an excellent episode–and one I don’t remember hearing last year–but is not nearly as timely as I had thought.

… with Mitchell Kertzman, a venture capitalist with Hummer Winblad Venture Partners. He talks about an investment that’s in progress. I’m just going to dump out a whole segment from him because it’s great stuff:

We are actually in the process of making an investment in a company where open source is a key part of their plan in two ways. The first is, that they are building on top of the so called “open source commodity stack” — that’s everything from x86 processors, to Linux OS, to the whole LAMP aggregation of core software stack.

So I think one of the huge breakthroughs–I’ve been running software companies before I did this for 30 years–and in my experience in running software companies that deal in enterprise software you may have a product that you sell to an enterprise customer but you had to drag around this expensive platform with you.

So if I want to sell my software to a customer, I might be selling them my software for a few hundred thousand dollars but in order to run it they might have to buy some expensive Sun servers and some Oracle databases and some application servers and the cost of that stack meant that you were pulling a boat anchor around with you. The commodity stack, with its performance and quality these days, means that software companies don’t have to do that anymore.

You don’t have to spend .5 million dollars of your first round of financing to buy the infrastructure you need to build and test your products. And it also means that you can sell your software to customers without the incremental costs.

The company we’re looking at, and all companies we’re looking at now are taking advantage of the lower cost to market of the commodity stack.

I think that’s a huge development in software..

The Money understands why less code is important, that’s a Good Thing, even if you don’t think you need it.

Mitchell goes on to talk about why Java isn’t looking as attractive as it once did when the hype machine was in full swing and the world was a different place:

When I first started doing this, everyone was doing J2EE backends. … We still have companies that are doing J2EE backends but the questions that are being asked lately are very interesting. When Java was first being developed, the portability of Java was a huge part of its attractiveness. The write-once, deploy-anywhere was extremely compelling. The interesting thing, if you stop and look now is that Linux essentially has made that, ummmm, almost moot. There are now really two environments: there is .NET and there is Linux. And so the portability of Java, since, according to Microsoft, it doesn’t have a place in the .NET environment, therefore if you’re non-Microsoft you really have one environment that you’re deploying to (if you look into the future). So it does raise the question of why you need to carry this JVM around with you if you don’t need portability.

And then the other interesting question is that, Java was useful when, uhh, if you consider that multi-tier or distributed application architectures used to communicate essentially for years by communicating binary data. Today, applications are really communicating by passing text back and forth, you know, XML or HTML, and Java isn’t particularly good at text processing so there’s an interesting development in [looking] at the scripting languages–at the P part of LAMP - PHP, Python, Perl, which are very good at text handling. This company we’re investing in is taking advantage of that.

So I think we’re in an interesting period of evaluation on Java.

Ahh man, this whole damn episode is amazing. I’m not transcribing the whole thing, just go listen to it.