Interview - PHP's Creator, Rasmus Lerdorf

SP: What would you say to young developers who are considering starting an Open Source project themselves?

RL: I am not sure it is possible in that sense. Sort of like sitting around watching your phone, trying to will it to ring. It always rings when you are in the shower or at some other inconvenient time.

I don't think you really sit down and decide to start an open source project. This "movement" is a bit of a myth that people like to glamorize and assign all sorts of unrealistic characteristics to. Nobody is going to join your project if all you have is a cool idea. Everyone has cool ideas.

People will pool around your efforts if you build something that is useful enough that they find it easier to take your code and extend it a little bit to solve their problem. If your stuff is half-baked then people are likely to dismiss what you have done, and just solve their problem themselves or by using something else out there.

So to start an open source project, first assume that you are completely on your own and solve some problem that has been bugging you for a while. That means months and months of work to get something that actually works and solves the problem. At that point you can start considering whether you might want to give up control and let others join your effort.

The key phrase here is "give up control". Starting an open source project does not mean you suddenly get a staff of programmers you can boss around. In fact, to get it off the ground you will have to be very receptive to suggestions from early adopters and do everything you can to make your tool more useful to a broader audience. And then you give it all away and let people do just about whatever they want with your code. Now you have started an open source project.

SP: What other open source initiatives have you got planned? Given infinite time (and 100 extra pairs of hands) what would you love do?

RL: Well, I didn't plan PHP. I think in terms of solving problems, not in terms of software projects. I actually hate programming, but I love solving problems. So what problems would I like to solve if I had lots of time and resources?

Having recently become a father I had grand plans to build a smart-crib. Live video, audio and various other gadgets that could be viewed and controlled remotely to let remote friends and family interact with the kid. Of course, once the baby actually arrived finding enough time to just sit and read a book is nearly impossible, never mind building a smart-crib!

Another really cool project would be the ultimate distribution building system. One where I could specify that I had a device with an 80486 processor, a certain type of NIC, and whatever other hardware specifics along with the type of media and ram, and this thing would crank out a small Linux distribution tailored exactly for my device.

Then, to extend that, be able to say that I want it to act as a firewall, mp3 server, browsing station or whatever. Basically get to the point where you can take almost any hardware device and stick Linux on it and have it be useful. I have a lot of devices around the house that I know I can put to better use simply by putting more powerful software on them, and I'd like to find a way to do this without having to spend 6 months on each one.

The SitePoint community and I would like to take this opportunity to thank Rasmus for his time and detailed answers to the inquiring minds in our community. He may no longer be the controlling force behind PHP, but he definitely shines as a member of the team that is working together all over the world to take PHP to the next level.