RE: [boost] Pool Review: Private Implementation

From: scleary_at_[hidden]
Date: 2000-12-04 10:20:08

> [Peter Dimov]
> I think that mutex should be split into three separate classes,
> detail::win32_mutex, detail::pthread_mutex, and
> detail::single_threaded_mutex; mutex would be conditionally typedef'ed to
> one of them. This will deal more gracefully with inconsistent definitions.

Sounds fine to me.

> [Peter Dimov]
> Don't want to reopen an old debate but shouldn't lock/unlock be private
and
> accessible via guard<> only?

This is an implementation detail.  It will be hidden in a details namespace
instead.

> [Jens Maurer]
>  - "singleton" looks like a separate concept; it should be separate.
> 
>  - "mutex"/"guard" is hopefully obsoleted by the efforts of the
> boost thread synchronization library. . .
> 
>  - Likewise, "postulate" should be obsoleted by the boost
> compile-time assertions.

These will all be hidden in a details namespace, along with the lcm/gcd
functions.  They are all *intended* for obsolecence, hopefully being
replaced by future libraries.  Right now, it looks like only "postulate"
will actually be removed before pool is added, but mutex/guard should be
replaced in the future with the BTL.

        -Steve