User:Electus

Electus Likes to Make Paint

Do you want me to make some paint for you? I would love to. Just send me a message. I may need you to get mats for me if I don't have them :)

Electus Likes to Simulate Paint

Understanding Paint: can't do better than read GHawkins' excellent explanation http://www.atitd.org/wiki/tale6/User:GHawkins/Paint

After learning about paint from GHawkins' page I made a paint simulator. It's not quite correct yet in some small ways but pretty much gets me to a good recipe. In case others are inspired to write their own, I wanted to present how I went about constructing it. It looks like the picture below. It was written for iPad but actually it is now too big to run on my iPad so I can only run it in the iPad simulator on a mac.

[[File:]]

The reaction value tables for each character are manually generated since I am not set up to run macros. They are hard-coded into the simulator as static const arrays. Really it should let you display and edit these on a different screen. I didn't write that yet.

It can be used interactively by picking the ingredients on the date-picker wheels across the middle. To use as a recipe finder, I enter the RGB for the target color into the three text boxes at the top. I get these numbers by using a color picker against GHawkins' color list on this wiki. Then I set up which mats to exclude from consideration using the eight swithches at bottom right. This is a useful feature which allows the cheapest recipe to be found. Then choose the number of mats to consider for the recipe. Choosing 10 takes forever to run. Choosing 5 gets you pretty close, then you can extend to 10 and run again to calculate the best ingredients for the last 5. Here is one of the program's shortcomings - it should use 10 non-catlayst mats plus up to 5 catalyst mats, but it just offers 10 slots of any kind. Because it seems to get me close enough I have not gone back and fixed this.

When the 'Seek' button is pressed it runs an exhaustive brute-force search using the combined difference of generated R, G, and B from target values as a quality metric. It then displays the best match found on the wheels.

I know you can write a better one! Please share if you do :)