Mutagenesis - Day 8

gcewing on 2011/09/18 01:44 of Mutagenesis

Found a planty-looking font for the title screen, changed the background to dark green, and stomped a few last-minute bugs. Looking pretty good. Final final release made. I can stop drinking inadvisably large amounts of coffee now.

(log in to comment)

Comments

Cosmologicon on 2011/09/18 20:54:

Very interesting! I think this is a little too realistic. :) The mapping from the genotype to the phenotype is probably too complex for me grasp. I tried to do something simple and breed as small a plant as possible, but every time I crossed two small plants I just got a huge plant.



Can you explain at all what's going on in the back end so I can figure out what I'm doing?

cyhawk on 2011/09/18 22:21:

Something about the DNA is printed to the terminal whenever you cross plants. I don't understand any of it yet but I feel that might hold the key!

gcewing on 2011/09/18 22:26:

The way the genes work is quite different from real life. The DNA is an L-system, which is a kind of grammar. The height of the plant isn't something that's explicitly coded, it emerges from the way the productions interact. What tends to happen is that the plant either stops growing after two or three steps because all the productions reach a dead end, or it potentially keeps growing indefinitely. In the latter case, the only thing that stops it is a hard-coded limit on the number of growth cycles.

The result of all this is that you can't really expect to control plant height by selective breeding, because any crossing or mutation has the potential to switch between a finitely-growing and infinitely-growing L-system.

I'll try to post another message soon with more details.

Cosmologicon on 2011/09/18 22:32:

Cool, I like the choice of an L-system for the DNA. I'm not too familiar with it. Can you give any examples of things that I can expect to control with selective breeding?

I don't have to have a goal. I can just sort of explore and stuff. But still it's kind of fun to challenge myself somehow. :)

gcewing on 2011/09/19 06:06:

Can you give any examples of things that I can expect to control with selective breeding?

It's hard to be sure, because I haven't had a chance to play with it much yet. But theoretically, if you have e.g. one plant with a particular kind of leaf, and other with a particular kind of flower, there's some chance that their offspring will have the leaves from one and the flowers from another. Some of the time. Maybe. But it's also possible -- probablt even more likely -- that some other weird thing will happen.

It'll probably be less frustrating not to aim too much for any particular goal, but just play around and see what comes out.