Re: Design for a file system library

From: mfdylan (mfdylan_at_[hidden])
Date: 2002-02-25 18:15:46

--- In boost_at_y..., Beman Dawes <bdawes_at_a...> wrote:
> At 09:58 PM 2/24/2002, Jeff Garland wrote:
> 
> 
> I guess what I'm saying is that at least two potential Boost 
libraries 
> (threads and filesystems) need some notion of time that goes beyond 
> anything now in the standard, so we really need a Boost-now and 
> standard-later solution so developers don't have to keep 
reinventing the 
> wheel.
> 

One possibility is to use a double with the same meaning as time_t 
(number of seconds since 1970, or just number of seconds).
I would think the precision of double is sufficient to allow as fine 
as granularity as you would ever want in the foreseeable future.  The 
range is surely more than enough for files and threads.
I know quite a few MS Office products, and think even SQL server use 
floating point numbers to represent time, although they use 1.0 = 1 
hour instead of 1 second.

Dylan