search

executors / futures

A proposal for a futures programming model for ISO C++
22 stars 7 forks source link

Fix refs and dfns #108

Closed griwes closed 6 years ago

griwes commented 6 years ago

The shared future term has been overloaded and that was making bikeshed confused - and I think it wasn't making much sense to confuse the ownership and the concept.

I'm open to other solutions to the problem this tries to fix, but @brycelelbach agrees that this general direction works.

Also removed another duplicate dfn and actually defined what a status is.

griwes commented 6 years ago

@brycelelbach @LeeHowes @dhollman please review

dhollman commented 6 years ago

This was already fixed in #105, but I like your fix better.

griwes commented 6 years ago

Ah yes, some of it has been.

105 doesn't fix status, though, plus it solves the shared future problem differently, still retaining what I see to be confusing overloading of the term (for the concept, and then for semantics, if that makes sense). In one place we have SharedFuture and ContinuableFuture, but in the other place we have unique future and shared future, with just one of those being a separate concept. I would like to fix that the way I proposed here.

Will rebase on top of #105 when that's merged.

griwes commented 6 years ago

Rebased and actually make sure that a SharedFuture is understood to be a non-uniquely owned future.

@LeeHowes sanity check on the terminology please?