Topic: [ANN] [IDISNEY] Disney Coin - page 31. (Read 72996 times)

04:57 < HMC_Alph> loves->likes->tolerates
04:57 < HMC_Alph> for example
04:58 < HMC_Alph> does einstein work? :-)
05:01 < stoopkid> like this? :
05:01 < stoopkid> kb
05:01 < stoopkid> Alice loves Bob.
05:01 < stoopkid> {?x loves ?y} => {?x likes ?y}.
05:01 < stoopkid> {?x likes ?y} => {?x tolerates ?y}.
05:01 < stoopkid> fin.
05:01 < stoopkid> query
05:01 < stoopkid> ?who tolerates ?whoelse
05:01 < stoopkid> fin.
05:01 < HMC_Alph> sure, does that work?
05:01 < stoopkid> yes
05:01 < stoopkid> Alice does indeed tolerate Bob.
05:01 < HMC_Alph> einstein?
05:01 < HMC_Alph> hungry germans?
05:03 < stoopkid> einstein does not work
05:03 < stoopkid> but it compiles
05:03 < HMC_Alph> ok.  what's the current smallest failing test?
05:04 < stoopkid> besides transitivity and symmetry?
05:04 < HMC_Alph> in general
05:04 < stoopkid> A causes B. B causes C. {?x causes ?y. ?y causes ?z} => {?x causes ?z}
05:05 < stoopkid> A equals B. {?x equals ?y} => {? equals ?x}
05:05 < stoopkid> those infloop on compilation
05:06 < stoopkid> i guess symmetry would be the smallest failing test
05:06 < stoopkid> i can't think of anything that would be smaller and fail
05:08 < HMC_Alph> anything larger that is not recursive?
05:09 < HMC_Alph> or are all of the fails cases of recursion?
05:10 < stoopkid> examples/cnf failed
05:12 < stoopkid> anything i should try in particular?
05:12 < stoopkid> it works on some boolean logic i constructed
05:14 < HMC_Alph> cnf not working is interesting
05:14 < HMC_Alph> can you try to reduce that case to a small test?
05:17 -!- rszeno [[email protected]] has quit [Quit: Leaving.]
05:20 < stoopkid> ah koo7 said there might be something buggy with the lists still
05:22 < HMC_Alph> sure, so maybe some simpler list tests to try to narrow that down?
05:22 < HMC_Alph> i'm going to look over the recursion for ep check
05:26 < stoopkid> yea a simpler 2 var cnf also doesn't work
05:26 -!- sifar [~sifar@unaffiliated/sifar] has quit [Ping timeout: 272 seconds]
05:41 < stoopkid> maybe since we're on example-making we should go over the DRTs stuff as well
05:56 < HMC_Alph> well the recursion problem isn't quite ep check...
05:57 < HMC_Alph> the recursion is in compile and ep check is at runtime
05:59 < HMC_Alph> ok, so let's start with simple recursion... like your causes example
06:01 < HMC_Alph> the function that gives you the coroutine should be called by those coroutine instances...
06:02 < stoopkid> hm true
06:02 < stoopkid> i thought we fixed that, guess not
06:02 < HMC_Alph> it is not quite a recursive lambda in the c++...
06:04 < HMC_Alph> it is an outer lambda with a self capture that returns an inner lambda which captures the captured outer lambda... and which calls it when called...
06:05 < HMC_Alph> the outer lambda is the python foo function itself, the unner lambda is the actual coro function that python hides fron you
06:06 < HMC_Alph> those inner coro functions call *foo* (the outer) if they are recursive... to get a new coroutine instance... they don't call their own coro re-entry point. ;-)
06:07 < HMC_Alph> follow?
06:07 < stoopkid> yea i follow that i'm just trying to figure out where in the C++ we're going wrong
06:08 < HMC_Alph> lol
06:09 < HMC_Alph> .g lambda over lambda
06:09 < tau-bot> HMC_Alph: https://chriskohlhepp.wordpress.com/lambda-over-lambda-in-cplusplus14/
06:10 < HMC_Alph> .g c++ recursive lambda capture
06:10 < tau-bot> HMC_Alph: http://stackoverflow.com/questions/7861506/recursive-call-in-lambda-c11
06:10 < HMC_Alph> meh, no
06:11 < HMC_Alph> .g c++ lambda recursion
06:11 < tau-bot> HMC_Alph: http://stackoverflow.com/questions/14531993/can-lambda-functions-be-recursive
06:11 < HMC_Alph> sure, that
06:11 < stoopkid> i mean where in our big tangled document of C++, not C++ in general
06:12 < HMC_Alph> heh
06:12 < HMC_Alph> it would be in pred() i suppose?
06:13 < HMC_Alph> let me try something
06:15 < HMC_Alph> wait, why is compile_preds doing unification?
06:16 < HMC_Alph> it should just take all of the rules with the same pred, no predvars here..
06:17 < HMC_Alph> yah, i don't get why that is there
06:19 < stoopkid> should i just take those lines out and see what happens?
06:19 < stoopkid> doesn't look like it affects the following control flow
06:21 < HMC_Alph> well, that is only half of the problem
06:57 < stoopkid> well, i beat the compile bug
06:57 < stoopkid> but, still more bugs
06:58 < stoopkid> cause kept being put into the predQueue to be compiled, so the while loop in compile_preds just kept running
06:59 < stoopkid> i added a check to make sure it wasn't already compiled before trying to add it to the predQueue
06:59 < stoopkid> no more compile infloop
06:59 < stoopkid> unfortunately i have bad C++ so i get a run-time bug
07:00 -!- scippio [[email protected]] has joined #zennet
07:02 < stoopkid> there's a strong correlation between bugs getting solved and koo7 coming online
07:05 < stoopkid> HMC_Alph: ill push my changes for you