On Fri, Mar 22, 2002 at 11:01:24PM -0500, Michael G Schwern wrote:
> In testing LWP against bleadperl@15421 there is only one failure
> different from perl 5.6.1 (the rest are attributed to proxy
> interference).
> 
> base/date.............FAILED test 51
>         Failed 1/58 tests, 98.28% okay
> 
> I think this one was already reported to p5p.  HTTP::Date is expecting
> Time::Local to be able to deal with fractional seconds.  That's not
> guaranteed by Time::Local, 5.6.1 and back just happened to be able to
> do it.
> 
> There's two ways to handle this.
> 
> 1) Have HTTP::Date not assume Time::Local can handle fractional seconds.
>    So eliminating test 51.
> 
> 2) Shut off "use integer" in bleadperl's Time::Local to make it handle
>    fractional seconds (pass them through, really).
> 
> Given that localtime/gmtime can't handle fractional seconds, and
> Time::Local is supposed to be the mirror of those functions, I'm for
> option #1.

This has been discussed and solved at least three times :-)
Upcoming HTTP::Date will have a tweaked test.

-- 
$jhi++; # http://www.iki.fi/jhi/
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Reply via email to