(no subject)
i've been playing with python lately, and i have a few thoughts.
python is like lisp, but with white-space for parentheses and some semblance of syntax. i'm still trying to decide if i like it or not --- there are a lot of things that i would prefer to express in a very functional manner (i.e. lots of map / reduce / apply) but map doesn't work on numpy matrices / tensors. more's the pity, because python in general seems to be a pretty reasonable prototyping environment, and far less likely to drive people to suicide than, say, perl.
i like that python has bindings to one of ilog's rule engines, that i can play silly games with the interpreter by embedding interpreter instances within a script (handy for the same reason i like the ilog rule engine being around).
todo: see how much effort it takes to run a python interpreter inside a c++ program and expose some of the c++ program's state to the interpreter. if i can get that to work relatively easily, then i think i'm golden for my new potential endeavor.
python is like lisp, but with white-space for parentheses and some semblance of syntax. i'm still trying to decide if i like it or not --- there are a lot of things that i would prefer to express in a very functional manner (i.e. lots of map / reduce / apply) but map doesn't work on numpy matrices / tensors. more's the pity, because python in general seems to be a pretty reasonable prototyping environment, and far less likely to drive people to suicide than, say, perl.
i like that python has bindings to one of ilog's rule engines, that i can play silly games with the interpreter by embedding interpreter instances within a script (handy for the same reason i like the ilog rule engine being around).
todo: see how much effort it takes to run a python interpreter inside a c++ program and expose some of the c++ program's state to the interpreter. if i can get that to work relatively easily, then i think i'm golden for my new potential endeavor.