Post by anonycat on Mar 15, 2011 18:40:54 GMT -5
So about four years ago, I set out to make an interactive text parsing script that could administer games of Wheel of Fortune accurate to the then-current ruleset (the fairly tame season 24). It took a few months to work out the kinks, but eventually I got it to do what it was supposed to do.
And then just a month later, they dumped season 25 on us and I had to go rewrite a bunch of stuff. About halfway through that season, I decided that if they're going to add even more rules in the future, the best way to prepare for it would be to split many of the rules off from the core file and put them in a separate configuration, so that I can easily maintain several rulesets at once--past, present, and whatever the future brings. That guiding principle has driven the game ever since.
So I've managed to get a group of regulars into my game, who have stuck with it through the years of its existence and who bust out the occasional speedrun, but beyond that core group it doesn't really have any following. Furthermore, that group didn't originally have much of a background in Wheel at all. I'd like to see what a more dedicated group of fans thinks of my implementation.
Get to the Game Already
For those who have experience with IRC, the game can also be reached at irc.surrealchat.net #wof.
Yes, the interface is text-only, and no, there are no plans to change that. If you really want to, you could write a custom client that receives the data stream and uses it to display the game state on a graphical set like the one BigJon uses. But take this as fair warning that I consider the mechanical aspects of the rules to be more relevant than tracking down all the stage sets they've played in front of over the last 36 years. In that light, I believe my program to be unrivaled, at least among freely available WOF-likes.
While the engine is versatile enough to handle almost any element in the game's history, I keep it running in a configuration that mimics only the most recent version of the US show. Specifically, the current configuration is listed here, and my hope is that it's commented well enough that interested developers can build on it and develop their own configurations even with the base engine code not yet released. If you have ideas in that regard, go ahead and run wild with them.
THE TO-DO LIST:
-Get a bigger puzzle base (this has to be the game's biggest weakness right now--the main reason that the aforementioned group of regulars is even able to stage meaningful speedruns is that they've already been through the entire list several times)
-Along the same lines, maybe rewrite the structure of the puzzle file entirely--at the very least, the puzzle list for tossups should probably be split off into its own file like the bonus puzzle list already is
-Add support for 6P team games (the actual in-game code shouldn't be a problem, but I don't want to add this until I figure out a sensible way to allow the players to divide themselves into teams in the first place)
-Come up with a system for supporting red letter puzzles that's better than the abomination found in BigJon's game (seriously, "guess which of the multiple valid anagrams I'm thinking of?" Who ever thought that could be a good idea?) Conceivably I could treat them like any other supplement, with the drawback that they would be limited to always appearing on the same few puzzles.
-Rename some of the text events in a more predictable pattern, for ease of constructing custom layouts that don't just break the game.
-Make a definitive and distinct configuration to call my own, to stand on that much more stable ground when it comes to potential trademark issues in the future. Of course anyone can come up with a crazy new layout, but the key is I'd like to keep it as close as I can to the existing game in terms of expected payouts (minus, perhaps, the variance of the "bonusboost" space), and that much requires a touch of care.
So, if anyone's interested, just click the "game" link above, and hopefully two other board members will also be on or get on soon (a possible drawback of realism: there's no support whatsoever for 1- or 2-player games). If not, you could always try to invite the people that are always there, but I can't guarantee that a game against them will be very interesting. At the very least, just stick around for a while to watch.
If you have any feedback about how the game could be better (other than "add grafx kthx"), I'd also like to hear that--it's part of why I'm trialing the program here in the first place, with the aim of making it as nice as it can be.
And then just a month later, they dumped season 25 on us and I had to go rewrite a bunch of stuff. About halfway through that season, I decided that if they're going to add even more rules in the future, the best way to prepare for it would be to split many of the rules off from the core file and put them in a separate configuration, so that I can easily maintain several rulesets at once--past, present, and whatever the future brings. That guiding principle has driven the game ever since.
So I've managed to get a group of regulars into my game, who have stuck with it through the years of its existence and who bust out the occasional speedrun, but beyond that core group it doesn't really have any following. Furthermore, that group didn't originally have much of a background in Wheel at all. I'd like to see what a more dedicated group of fans thinks of my implementation.
Get to the Game Already
For those who have experience with IRC, the game can also be reached at irc.surrealchat.net #wof.
Yes, the interface is text-only, and no, there are no plans to change that. If you really want to, you could write a custom client that receives the data stream and uses it to display the game state on a graphical set like the one BigJon uses. But take this as fair warning that I consider the mechanical aspects of the rules to be more relevant than tracking down all the stage sets they've played in front of over the last 36 years. In that light, I believe my program to be unrivaled, at least among freely available WOF-likes.
While the engine is versatile enough to handle almost any element in the game's history, I keep it running in a configuration that mimics only the most recent version of the US show. Specifically, the current configuration is listed here, and my hope is that it's commented well enough that interested developers can build on it and develop their own configurations even with the base engine code not yet released. If you have ideas in that regard, go ahead and run wild with them.
THE TO-DO LIST:
-Get a bigger puzzle base (this has to be the game's biggest weakness right now--the main reason that the aforementioned group of regulars is even able to stage meaningful speedruns is that they've already been through the entire list several times)
-Along the same lines, maybe rewrite the structure of the puzzle file entirely--at the very least, the puzzle list for tossups should probably be split off into its own file like the bonus puzzle list already is
-Add support for 6P team games (the actual in-game code shouldn't be a problem, but I don't want to add this until I figure out a sensible way to allow the players to divide themselves into teams in the first place)
-Come up with a system for supporting red letter puzzles that's better than the abomination found in BigJon's game (seriously, "guess which of the multiple valid anagrams I'm thinking of?" Who ever thought that could be a good idea?) Conceivably I could treat them like any other supplement, with the drawback that they would be limited to always appearing on the same few puzzles.
-Rename some of the text events in a more predictable pattern, for ease of constructing custom layouts that don't just break the game.
-Make a definitive and distinct configuration to call my own, to stand on that much more stable ground when it comes to potential trademark issues in the future. Of course anyone can come up with a crazy new layout, but the key is I'd like to keep it as close as I can to the existing game in terms of expected payouts (minus, perhaps, the variance of the "bonusboost" space), and that much requires a touch of care.
So, if anyone's interested, just click the "game" link above, and hopefully two other board members will also be on or get on soon (a possible drawback of realism: there's no support whatsoever for 1- or 2-player games). If not, you could always try to invite the people that are always there, but I can't guarantee that a game against them will be very interesting. At the very least, just stick around for a while to watch.
If you have any feedback about how the game could be better (other than "add grafx kthx"), I'd also like to hear that--it's part of why I'm trialing the program here in the first place, with the aim of making it as nice as it can be.
"by hand" uncoded game this Sunday, if you can- I hear you're a great player!
