Hi.

This is Ray Dillinger's homepage. And it isn't on Facebook or Blogspot or any of the thousand other places that want people to put their homepages on it. This is hosted on a machine that's at my house.

Because I'm jobhunting, I guess I will cave in to the pressure to have a Linkedin page, and inflict the inevitable spam on people who want to visit it. But first I'm going to post all the relevant information here, so you don't have to be advertised at in order to get it. So, Here's a link to my Resumé in HTML format.

I'm working on a roguelike game called Neohack. I'm taking several "roads less taken" in terms of design with it, mostly to modularize the game content to a degree greater than previous roguelike games have done. So, it's got a data-directed component architecture using reified "event" and "response" closures rather than the more usual nest of linear control structures and OO methods. I think it's interesting and fun, but I'm only making progress slowly. I'm implementing it in C (even though that means doing some hacky things in the makefile to implement data-directed initializer callbacks, because C doesn't natively have reflective capabilities), just because I like the old-fashioned hardcore-ness of C. It hasn't been released yet as a game, because there's no real content as yet, but here is a link to the official project page, where you can download the latest dev snapshot.

Roguebasin is a wiki devoted to roguelike game development. It's a neat place, with articles about every aspect of game development and links to all the games you need to waste time for the rest of your life. I've contributed several articles there. In particular, FastLOS and the Spiral Path FOV algorithm are both algorithms I developed. FastLOS in particular is one I'm pleased with because although it's a heuristic, its unreliable cases can be easily detected and handled conventionally and it is at this time the most efficient known method for determining line-of-sight in all other cases. It does, however, require some heavy preprocessing and a conventional field-of-view implementation is needed to do the preprocessing. At this time I know of no games that use it. It seems that once there is a conventional field-of-view implementation, most people would rather just use it directly rather than implement the preprocessing procedure needed for FastLOS (and, somewhat embarrassingly, that includes me. FastLOS is not currently implemented even in Neohack).

I really like the Reprap project. Not so much for the plastic extruders, although those are nifty, as for the hands-on experience with CNC and robotics and the general idea that ordinary citizens ought to have access to useful and worthwhile productive capabilities. The idea that you can go from design to object using an automatic process is really cool. The fact that a whole lot of people are collaborating to create that process is even cooler.

Of course, in order to see the collaborative side of the project, you have to go where everybody's collaborating: the reprap wiki ! Personally I think the project is far too focused on plastic extruders and additive manufacturing - I think that it should encompass manufacturing in general. Extruders are nice, but so are CNC mills and lathes and welders and laser cutters and arbor presses and injection molding rigs and so on. Enabling people to make their own things does not mean developing one productive tool; it means developing as many as we can.

And of course, if you're going to include a link to reprap, you've got to include a link to the Thingiverse -- a repository of public online plans and diagrams for things anybody can build with the proper tools. It is here that most people start seeing the possibilities involved in having their own productive capacity. Want something a bit different? Modify the plans before building!

I have a lot more interests and projects, and I'm in the process of organizing and documenting them for a web presence. As they come online, I'll add links to them here. When there are enough of them that organization is an issue, I'll organize them a bit.