3 ms·

such an analysis is normally a fixpoint over the control flow graph; what structure would we be taking the fixpoint over in the Sen/Nussbaum case? (non-sequitu

such an analysis is normally a fixpoint over the control flow graph; what structure would we be taking the fixpoint over in the Sen/Nussbaum case? (non-sequitur) did you see https://protocolized.summerofprotocols.com/p/strange-new-rules https://protocolized.summerofprotocols.com/p/strange-new-rul... ?

gsf_emergency_2 2y ago
(welcome back!) Didn't see it. I'm a bigger follower of geohot (these days! sorry!) but this does look far more interesting. Scheduled for an imminent read! Lacking the training/experience/talent for things like this I can now only proceed by vague analogy (work on this one and off if I don't get it quite right? Thanks!): (Fixed points are analogous to steady states?) (Even nonequilibrium,societal ones would be kind of mediocre imho, (sorry nis0s!)) it would be analogous to a diagram of (estimated likelihood(s) of progress on your favourite heuristic ) vs (memory allocated currently, or more generally, tuples of those 10 historical CA-related variables together with their estimated relevance towards progress) on every subheuristic? (I'm trying to introduce probabilities to (sigma algebraize?) the program representation)
gsf_emergency_2 2y ago
(n-s: the source of my attempted analog : https://www.cs.cmu.edu/~aldrich/courses/15-819O-13sp/resources/dataflow-frameworks.pdf https://www.cs.cmu.edu/~aldrich/courses/15-819O-13sp/resourc... A concise few pages worth drowning in! (For me)) (p-s: 2 "repulsive fixed-points" of 2 messianics of interest: PG: the anti-prig Thiel: the protocolizer as chief optimist/protocolization creditor/secretary of the politburo ) (p-p-s, from the comments: Henry VIII as designer of new kinds of windmills, would have been awesome snack if that checks out, roughly on par with Lawrence of A coming up with a new jet engine)

3 ms·

such an analysis is normally a fixpoint over the control flow graph; what structure would we be taking the fixpoint over in the Sen/Nussbaum case? (non-sequitu

by 082349872349872 2y ago

gsf_emergency_2 2y ago
(welcome back!) Didn't see it. I'm a bigger follower of geohot (these days! sorry!) but this does look far more interesting. Scheduled for an imminent read! Lacking the training/experience/talent for things like this I can now only proceed by vague analogy (work on this one and off if I don't get it quite right? Thanks!): (Fixed points are analogous to steady states?) (Even nonequilibrium,societal ones would be kind of mediocre imho, (sorry nis0s!)) it would be analogous to a diagram of (estimated likelihood(s) of progress on your favourite heuristic ) vs (memory allocated currently, or more generally, tuples of those 10 historical CA-related variables together with their estimated relevance towards progress) on every subheuristic? (I'm trying to introduce probabilities to (sigma algebraize?) the program representation)
gsf_emergency_2 2y ago
(n-s: the source of my attempted analog : https://www.cs.cmu.edu/~aldrich/courses/15-819O-13sp/resources/dataflow-frameworks.pdf https://www.cs.cmu.edu/~aldrich/courses/15-819O-13sp/resourc... A concise few pages worth drowning in! (For me)) (p-s: 2 "repulsive fixed-points" of 2 messianics of interest: PG: the anti-prig Thiel: the protocolizer as chief optimist/protocolization creditor/secretary of the politburo ) (p-p-s, from the comments: Henry VIII as designer of new kinds of windmills, would have been awesome snack if that checks out, roughly on par with Lawrence of A coming up with a new jet engine)