Scott Vokes
Scott Vokes
TBH, just using rlwrap (when desired) seems vastly simpler. It's probably worth looking into how other event-loop-based systems. The other I know offhand, the Tcl shell, doesn't seem to use...
And a third of the cumulative time is spent in just cd and sva... Each sample counts as 0.01 seconds. % cumulative self self total time seconds seconds calls s/call...
(The reason I broke up dv_ex is to be able to work out the dataflow, and how to apply that optimization, BTW.)
Indeed. I just sent you some code to benchmark the impact of function-point-based dispatch. It has very low (
Can you clarify what you mean by "doing valences implicitly"? Is recalculating mempool cell sizes really a significant time sink? I've never seen that stick out in profiling; usually redoing...
Nice. It went from 44340 to 6130 for me. (All tests pass with the new _vs.)
I will profile the newer version when I have a chance.
It should be possible to squeeze a lot more out of that. I'm still working up to running that on my vm, but will post a time when I get...
I figured out a better way to handle the arithmetical verbs ("scalar dyads") when setting up operator & type dispatch for my VM. If nothing else, it should greatly reduce...
"hangup"? That's a new one. I'll see if the seed suggests anything.