On Tue, 11 Feb 2003, Peter J. Acklam wrote:

> Nit-picking perhaps, but I think the language in the docs needs to
> some rewording.  There is too much confusion about what the epoch
> is.  Common practice says the epoch is _point_in_time_.  Keeping
> that in mind, the following makes little sense
>
>    $epoch_time  = $dt->epoch;     # may return undef for non-epoch times
>
> What is a "non-epoch time"?  And why "may" it return undef?  Will
> it or will it not return undef?  I'd say something like "returns
> undef if output overflows".
>
>    * day_of_year, doy
>        Returns the day of the year. Analogous to the yday attribute of
>        gmtime (or localtime) except that it works outside of the epoch.
>
> What is a time "outside of the epoch"?  The epoch is not a time
> span.  And this
>
>    Using epochs is not recommended, since they have such a limited
>    range, at least on 32-bit machines.
>
> "Using epochs"?  I'd say "Representing time as seconds since the
> epoch is not recommended ..."
>
> Any opinions?

I've changed some of the wording to try to make it clearer that what I'm
recommending against is using your system's representation of seconds
since the epoch.


-dave

/*=======================
House Absolute Consulting
www.houseabsolute.com
=======================*/

Reply via email to