Death of the Hero Programmer

I went to a Technical SIG about "Open-Source in GIS" and I found this talk particularly interesting Open Source GIS Server Systems - Norman Barker, Research Systems Inc. It was almost as interesting as the free lunch. It was about Norman's experience working on open source projects, which if I am honest I had never really thought about before.

There was an interesting talk from a Norman Barker who worked on open source GIS projects and he was saying that it is a fantastic social thing to do and that a lot of chatting on forums was done. I haven't ever worked on a open source project so I can't really comment but it does sound quite interesting, lots of people all with the same source code in front of them, pair programming turned up to 11. When I think about Open Source programs you can imagine the potential power that a project could generate. There is a quote from Eric S. Raymond in his paper The Cathedral and the Bazaar, is:

"Given enough eyeballs, all bugs are shallow."

At the talk I was attending the first person presenting didn’t really believe this comment because some bugs would involve a detailed knowledge to solve it but surely that could happen on any project. I personally think that more people testing and looking at source code must find more bugs.

The talker was commenting on how open source projects have nightly builds, you don’t have to wait six months for a release you get the functionality straight away and if it doesn’t work correctly you can help fix it

This also sounded brilliant and inspiring and made me think why aren’t their more open source projects, particularly in England. It must be something to do with the money and businesses. I haven’t worked for a business, which used much open source software and certainly not one who helped develop any. In some ways I can’t blame them because it would seem in some way that all your work and the money you are paying developers for would be freely available to other companies.

One of the talks at the conference did hint at another problem. He had been working on a piece of software which collated free weather reports from various sources (called MIDAS if you are interested, I think) and he said he had to choose which open source map server to go with. It turned out that he made a bit of a beta max (old video recorder which no used because suddenly everyone used VHS) of a decision. The most popular map server three years ago has slowly become less popular and now MapServer (Minnesota university map server) has been made open source and people have invested some time on it and now it is leading the way.

This is one of the fears with open source software in business, what happens if people stop developing the product. This still can happen with commercial software as a number of people at the event were complaining about where one of the mapping software companies stopped using coldfusion and has decided to go Java (something like that I wasn’t really listening that much)

The other reason I can see why businesses wouldn’t use open source software is because if there are problems with it, you can’t guarantee anyone is going to look at and fix those problems, no one is accountable. Some people complained that the open source MapServer did struggle under heavy usage. How well tested our open source products?

I don’t know but then you think of the quote “many eyes makes shallow bugs” and think open source products must be quite well tested by enthusiast people striving to improve the product.

Open source is a bit of a catch 22 situation, the more people behind the project the better the product the more people would use it but people are waiting for better product before they support it.

Where was I rambling to with this blog entry? Oh yes I like the comment from the speaker that he thought recently that open source projects were becoming a collaborative affair with many developers working as a team and community to fix bugs and add features (it sounds good doesn’t it) and that it was death to the hero programmer like Linus Benedict Torvalds

I loved the idea of the hero programmer, what am I, a code warrior, a Java Ninja perhaps. More realistically I am probably more a code monkey or maybe if Java is really going down the pan like so many people say we are just Java Lemmings preparing to leap off the cliff.

If you are interested in the economics of open source check out this page it has some links to someone debating Linux and why another company closed off their source code. Go Java Lemmings, jump to it

this is a link to a debate about business models and open source

I don't really know much about open source projects as you can probably tell from the blog entry above, so please don't be insulted by the blog above and if you would like to comment on what really happens that would be greatly appreciated because I'm an eager learner and as they say "everydays a school day"