Community Matters

As a strong believer in open source I might be a bit biased here, but there’s one thing that puzzles me. Why aren’t companies more eager to have people innovate on top of their products?

Look for example at Intel. Back in 1990 they published complete instructions how to program the 80386 processor and accompanying chipset. The result? A young Finish student started to play with his new system and started a little project that went on to change the world of computing – Linux. Today Intel is still the most open and active supporter of open source out there (and yes, I might be even less objective here, given that I work for Intel – on the other hand it means that I know exactly what I am talking about when I say this).

Understanding the huge success that was caused by this openness, I am completely stunned to see other companies go out of their way to prevent people from innovating on their platform. Take Motorola. The Ming is a great phone, but there is no working SDK. Or even worse the iPhone. Where Apple is actively trying to prevent people from running third party software on the phone, to the extend of bricking phones that had non-Apple software installed on them.

That has two consequences. First, many innovators (and, btw, potential customers) are scared away and will simply not buy the product. That seems dumb. And second, the most enthusiastic people who aren’t scared off by these tactics are now wasting their time hacking the firmware over and over again (and they are succeeding, mind you… regardless how many smart people are working for a company, there will always be more smart people outside that company). These smart, enthusiastic people could instead be spending their time on creating the next killer application – which likely would even more increase the market for the product.

As I said, I don’t get it.

This took longer than I thought. Inertia. It’s a bad thing. The blog, mail server and everything else was sort of working (if slowly) on the old laptop. So I never had a lot of pain that would make me jump up and make all the changes to the infrastructure that I wanted to make.

But finally after more than a month I think we are getting there. Now we have an ancient x86 laptop (sounds like a theme, doesn’t it? I think this one is an original Pentium M Banias at 1.4GHz) running as firewall/router. It cleanly separates the external network (including the wireless cloud that I have our neighbors on) from the internal network and the DMZ. I started with some very basic (and very restrictive) Shorewall rules and slowly and carefully opened those ports and connections that I really needed.

Behind that sits an upgraded Mac Mini (hooray for eBay and the online instructions that tell you how to open and expand the Mini) running an Intel Core Duo T2400 at 1.83GHz with 2GB of memory (this used to be a low-end 1.5GHz Core Solo with 512MB). And the PowerBook G4 as backup server in case I need to take the Mini off-line for something.

All three systems are running Debian Etch with the latest patches and some backported newer components (for example a 2.6.23 kernel and a newer version of SpamAssasin and Shorewall than are currently available in Etch).

And while I was at it I upgraded the second Mac Mini as well (similar specs, only the second cpu I got through eBay was a T2300 at 1.66GHz) and am running the latest pre-release of OS X Leopard on it. Nicer than Linux as a client. But I’m very happy that my whole server infrastructure is back on Linux – much easier to work with and much easier to keep up-to-date.