Manufactoria
Jun. 13th, 2010 07:07 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
I'm quite enjoying the game, but it wasn't until after I'd already solved the first 28 levels that I discovered, while looking at someone else's solution, that you could make conveyor belts cross.
Good lord. That certainly makes things easier!
(no subject)
Date: 2010-06-13 03:44 pm (UTC)Speaking of making things easier, I solved all the levels and then looked at someone else's solution to see how clever people popped a character off the back of a string. Which was, um, about a quarter of the size of the way I had been doing it. Then I cleared my levels and started again with that new knowledge. Wheeee!
(no subject)
Date: 2010-06-13 09:45 pm (UTC)Are you talking about removing the last dot from a sequence? My original solution was awful, particularly since I didn't know about bridging then. But that forced me to find a non-bridging solution which was even better:
After I learned about bridging I went back and improved upon my original algorithm:
It's two pieces bigger than the other and and 4 steps slower for strings ending in red, but handles groups of the same colour faster. So overall its performance is about the same.
(no subject)
Date: 2010-06-14 05:01 am (UTC)but it boils down to this which works well and might be a little faster than your second plan if you happen to not expect empty strings:
But that first design of yours was a game-changer for me.
At the end of the day, as much fun as the conveyor belts are, the way I'd really like to conceptualize the problem is as a standard FSM with writing rules so that there is no issue of the topology of how the arrows come out of the branches. For instance, here is the (partially flawed) design for Ophanim that I came up with and is far prettier than the Manufactoria machine that eventually came out of it:
It sounds like I'd have to write that game myself, though.