Recent content by professorjellybean

Hooray!!!

I wish I saw this thread earlier. Is it too late to request a CLI command to do the witchess fights automatically? Also, in reply to the procedural puzzle solver issue--If done in java it should be sped up a lot with an intelligent A* search.

Can you give the property name/put it on the KoLMafia wiki?

I'm getting the same issue, and can confirm Armorer is finished. 1061 doesn't seem to be settable with any variation of choiceAdventure1061, choiceAdventure061, or choiceAdventure106, so it can't be a truncation issue. > get questM25Armorer finished

Thank you all for the pull requests and contributions!

I'll add it soon but it looks like KoL changed the specification for the page. I'll need to examine and change some output/parsing. Please KMail me, anyone, if you've made a pull request.

Thanks! I fixed and pushed yesterday. Adding some more changes now.

Granted, it is harder to solve than to generate a problem. However, it's also In addition, unlike cryptographic hash algorithms, small permutations produce predictably related results. From one almost-solution you can find full solutions, etc. We just need an efficient way to converge to a...

I find the fact that we only have arrays and TreeMaps to be extremely limiting in terms of data processing. For example, there's no way to have a variable-length list. Being able to use ordered mutable lists, sets, etc. would be extremely helpful to data processing. In addition, unordered data...

Update! Code now covers all puzzles up to #56, assuming the google doc is trustworthy. I did some quick cursory tests on dimension and displacement to verify, also, but nothing formal--so typos are still possible. Also, some puzzles appear to be non-square. However, they are still rectangular...

No problem! Please send me any additional solutions or ideas for automated solving. I have very very limited ability and coverage right now.

Oops, thank you all. Fixes made.

It works for me, but I really really really hope it works for everyone. Please let me know about any errors you encounter!

Okay, so what I'm getting is that an ideal solution probably works bottom-up rather than top-down; build possible solutions in a highly restricted way rather than spam solutions and start eliminating. Anyway, since we know the rules by now, we can always privately test solutions before...

I did a thing! This will complete the daily witchess puzzles. Get it via "svn checkout https://github.com/AllenTuring/WitchessSolver/trunk" in your gCLI. It can be launched with "witchess_solver". I typed up the puzzle solutions in ruld (right up left down) format from the communal google doc...