Why not the parts of Concurrency TS that go in C ++ 17?

According to Michael Wong, Concurrency TS does not go, despite the full, apparently

Although there is implementation experience, it was just approved and too fresh to be voted for adding in C ++ 17.

My favorite suggestion was originally in N3327 , but I first read it in N3857 / N3784 and expected it for C ++ 14. it futureshas and implements . then () in Boost since 2013, and Microsoft implemented their form in PPL, so any problems would have been affected, discussed and fixed. The asynchronous version .get()is a necessity and seems to have been resolved since the beginning of 2014, when N3865 (“Additional improvements for std :: future ') spoke as if .then()given. This means that even from 5 years since N3327 first raised.then() because of the “streamlining” of any well-thought-out, discussed and implemented proposal, it is blocked by discussions that concern it.

Now it will be until 2020, before it sets off.

Is Concurrency TS a whole or void offer? And if so, why?

+2
source share
1 answer

Now it will be until 2020, before it sets off.

This is shipping now. Not in the C ++ standard itself.

You should not think of TS as some kind of fictional document. This is the real thing, and many vendors are implementing TS. For example, Microsoft and lib ++ support FileSystem TS systems.

With the full version of C ++ 17, the compiler and standard library developers fully perform these functions. Not to mention TS concepts, etc. Therefore, it is not surprising that Concurrency TS is not their highest priority.

, Concurrency ++ 17, , SG1 ( Concurrency) Parallelism TS , Concurrency TS. , -, , Concurrency TS .

+3

Source: https://habr.com/ru/post/1677937/


All Articles