Re: Denial Mistress (GuideMe) - YOU can make it better.
Posted: Tue Apr 21, 2015 11:44 pm
Hello, this may or may not be useful, but I stumbled on this thing a while ago and found it very interesting.
http://www.xwray.com/fiftyshades
This is a random text generator in the style of 50 shades of gray. If you follow the link below to github they have a description of how it was made as well as grammar file used for the engine that generates it. I think this could be adopted to, well, any of the random style teases that are created here.
In a 10 minute look I didn't get a very clear understanding of how it works, but at a high level I can somewhat start to see it. But tweaking this out for the broad categories of texts required here, you could (not sure) possibly call this from guideme in some manner, or, run it over and over to generate texts and manually place them into the tease.
The author suggests not using the 'dada' engine she used as it is over complicated, so going to shop around to see what you should use, as this process of getting the computer to appear to be speaking to you is quite interesting (but not being as robotic as using a modfied alice bot or something like that).
I mean the results on this page alone are pretty awesome and it seems really worth it to get a 'teasing, edging' text generator going for the multiple authors on this site that might use it.
*EDIT*
Ok just looked a bit more, here is the manual for the dada engine, the first overview section is a good view on what the engine has going on.
http://dev.null.org/dadaengine/manual-1 ... .html#SEC1
Nothing insane here. The idea would be forming a sentence structure and giving variables for each part of it. So in the context of each category you could have a rule going on like (pardon the lack of correct syntax, just looking at this still)
edge_sentence: [command| ""] edge_verb [time|""]
Then you'd drill down and pad out rules for above so command could be one of or neither:
I want you to, you will, she wants you to, i demand you, you must
edge verb would one of "edge, get to the edge, get on the edge, ride the edge"
time would be one or none of things like NOW, immediately, instantly, this second, in a moment
Anyway that isn't a well divided structure but I hope you can see the potential. You just fire that through the engine in a loop 100 times and you can pick out a number of meaningful phrases.
More over this can be used all over for more complex sentences so long as you break down what each part could be and make each part follow its own set of rules. For example the command part, rather than being a list of strings could be a list of rules in turn, helping to pad out the variety.
I'm sure this could also be used all over for getting texts of other phrases. This may still be a bunch of work, but perhaps makes the thought of writing it all somewhat interesting and exciting (as you wont quite know what gets generated) and much less daunting.
http://www.xwray.com/fiftyshades
This is a random text generator in the style of 50 shades of gray. If you follow the link below to github they have a description of how it was made as well as grammar file used for the engine that generates it. I think this could be adopted to, well, any of the random style teases that are created here.
In a 10 minute look I didn't get a very clear understanding of how it works, but at a high level I can somewhat start to see it. But tweaking this out for the broad categories of texts required here, you could (not sure) possibly call this from guideme in some manner, or, run it over and over to generate texts and manually place them into the tease.
The author suggests not using the 'dada' engine she used as it is over complicated, so going to shop around to see what you should use, as this process of getting the computer to appear to be speaking to you is quite interesting (but not being as robotic as using a modfied alice bot or something like that).
I mean the results on this page alone are pretty awesome and it seems really worth it to get a 'teasing, edging' text generator going for the multiple authors on this site that might use it.
*EDIT*
Ok just looked a bit more, here is the manual for the dada engine, the first overview section is a good view on what the engine has going on.
http://dev.null.org/dadaengine/manual-1 ... .html#SEC1
Nothing insane here. The idea would be forming a sentence structure and giving variables for each part of it. So in the context of each category you could have a rule going on like (pardon the lack of correct syntax, just looking at this still)
edge_sentence: [command| ""] edge_verb [time|""]
Then you'd drill down and pad out rules for above so command could be one of or neither:
I want you to, you will, she wants you to, i demand you, you must
edge verb would one of "edge, get to the edge, get on the edge, ride the edge"
time would be one or none of things like NOW, immediately, instantly, this second, in a moment
Anyway that isn't a well divided structure but I hope you can see the potential. You just fire that through the engine in a loop 100 times and you can pick out a number of meaningful phrases.
More over this can be used all over for more complex sentences so long as you break down what each part could be and make each part follow its own set of rules. For example the command part, rather than being a list of strings could be a list of rules in turn, helping to pad out the variety.
I'm sure this could also be used all over for getting texts of other phrases. This may still be a bunch of work, but perhaps makes the thought of writing it all somewhat interesting and exciting (as you wont quite know what gets generated) and much less daunting.
