Realease date?

[triclops200]

When is the estimated release date for darwin bots 3... I can't wait for c# progging for dna, much easier for me to read.

[Numsgil]

Given current progress and rate, maybe Christmas 2010.  But that's not a 100% accurate estimate since I was unemployed for many months, and spent lots of that time working on Darwinbots.

You won't be able to program DNA in C#, though that's definitely a possibility in the future since the language has a built in C# compiler.  The stack based reverse polish notation language is useful because it makes mutation code very robust.  Any higher level languages would require mutation code to understand code structure, which is sort of cheating IMO.

[ikke]

Any new insights for first release date?

[Numsgil]

Chistmad 2010 is probably overly optimistic.  But we'll see. Still making slow and steady progress.

[bacillus]

Does anybody have a list of new features in DB3 that seem pretty definite ATM, or is it mostly just getting a skeleton set up and seeing where we end up?

[Numsgil]

Mostly a skeleton.  Once I have a system that can handle something like animal minimalis, I'll start adding layers of features on top of that base.

I definitely want to support multibots that move around by "swimming", though.  That's a definite.

[bacillus]

Cool, that's a third of my wish-list ticked off  
In case anyone's wondering, the other two are specialized materials, which looks like a definite as well, and hexagonal 'bricks' - these act as minerals that cluster together. With specialized glands, a bot can dissolve these for minerals (shell etc.), and regurgitate them. This would be a sort of abiotic factor in the environment to interact with that would allow bots to build shelters or hives.

[Numsgil]

I do definitely want to play with things like ant bots making nests inside shapes by removing blocks of material.  Instead of a grid (eg: something like Dig Dug) it'll probably work using shapes (this shape - union of negative shape primitives (circles, etc.)).

[bacillus]

The reason I thought of hexagons is that they are the one of the simplest shapes that leave no gaps and fit neatly, so they don't become too much of a nightmare for the physics engine. Why not squares or triangles? Lets face it-I'm a Wesnoth addict  

[Numsgil]

Hex grids are cool.  I'm totally all for them over square grids.  But in this case I'm trying to avoid all grids, since they can take up a lot of memory if you're not careful.

[bacillus]

It's a fine balance-grids would only be more effective if they are used in fairly small sims or if these objects will be used a lot. Either way, we'd need some performance testing and general playing around, and the sim would have to be flexible enough to suit both options.