Re: [boost] [review] status of Boost.Conversion?

Subject: Re: [boost] [review] status of Boost.Conversion?
From: Nathan Ridge (zeratul976_at_[hidden])
Date: 2012-06-22 21:34:15

> > Hi,
> >
> > I am wondering, what is the status of the Boost.Conversion library?
> Hi,
>
> are you interested on the library? the review status?

I am interested in the library. At the moment I am particularly interested
in the ptime <--> chrono conversions (there isn't functionality for that
elsewhere in Boost, is there?), but the library as a whole seems
interesting and useful.

By the way, I have a question about the ptime <--> chrono conversions.

In this stackoverflow question [1], the implementation of these conversions
is suggested as a way to implement ptime <--> std::chrono conversions, and
Howard Hinnant points out in a comment to the accepted answer that this
implementation silently returns incorrect results if the clock's epoch is
not the same as ptime's.

I am wondering, is your implementation susceptible to this problem, or can
you somehow assume more about the epoch of a boost::chrono clock than you 
could about the epoch of an std::chrono clock?

In case your implementation does have this problem, Howard also suggests a
solution in another comment.

Regards,
Nate

[1] http://stackoverflow.com/questions/4910373/interoperability-between-boostdate-time-and-stdchrono