Subject: Re: [Boost-bugs] [Boost C++ Libraries] #10967: Timed wait points inconsistently convert relative to absolute waits
From: Boost C++ Libraries (noreply_at_[hidden])
Date: 2015-01-25 19:34:49
#10967: Timed wait points inconsistently convert relative to absolute waits
-------------------------------+---------------------
Reporter: ned14 | Owner: ned14
Type: Bugs | Status: new
Milestone: To Be Determined | Component: thread
Version: Boost 1.57.0 | Severity: Problem
Resolution: | Keywords:
-------------------------------+---------------------
Comment (by ned14):
Replying to [comment:8 viboes]:
> I don't see why we would need the 3 API. Each platform could provide
only one of them. E.g. pthread provides only wait having as parameter the
an absolute time.
They would only be internal API. Public code cannot see them.
My reason is because if you unify all the timed wait logic into a single
location for all of Thread instead of each threading primitive doing its
own thing, it makes it much easier to debug, coordinate and extend into
the future. For example, if you implement a timed_wait_for_any(future) it
is using identical timed wait code to timed_mutex.
Sure, on POSIX all three functions convert into pthread absolute timed
waits. But it is an API abstraction to help maintenance, that's all.
> I agree for fixing all this stuff if the result is alway backward
compatible (Or having the consensus that Boost.Thread can break user code
on the Boost ML).
The API I proposed is completely internal.
> If you want to work on a clean Boost.Thread implementation I suggest you
to contribute on a new completely separated version V5 for C++11 compilers
as I plan to do. Are you interested?
I'll reply to that by private email.
Niall
-- Ticket URL: <https://svn.boost.org/trac/boost/ticket/10967#comment:9> Boost C++ Libraries <http://www.boost.org/> Boost provides free peer-reviewed portable C++ source libraries.
This archive was generated by hypermail 2.1.7 : 2017-02-16 18:50:17 UTC