On Sunday, August 20, 2017 at 6:28:59 AM UTC-4, Jonas Kölker wrote: > > You've made some good observations. I haven't thought of this in terms of >> threshold reads before, but what you're describing sounds like the >> implementation of deferred goals in some MK variants. >> >> Deferred goals are resumed when the variables they depend on become >> known, which should be able to support dynamic interleaving, as in your >> interpreter/type-inferencer example. >> > > Yep, that sounds like a threshold read. What does "known" mean? Fully > ground, or just non-fresh? >
As soon as any information becomes known (so symbolo and numbero qualify for this case), a goal is resumed. If the new information turns out to be insufficient, the goal may suspend itself again right away. In a lot of cases, it will use the new information to perform some simplification before re-suspending. You could say there are implicit tripwires. > It's not working well enough to be useful yet. >> > > Interesting; what are its shortcomings? > It seems to be due to bad implementation on my part. I haven't had time to dig in to figure out why, but the scheduling of goals isn't distributed in the way we expected. Once this is fixed, I imagine there won't be any serious shortcomings. It will take a while to get back to this, but I'll let you know when I find out. -- You received this message because you are subscribed to the Google Groups "minikanren" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/minikanren. For more options, visit https://groups.google.com/d/optout.
