Re: 64-bit issues in date functions?
Joe Hohertz (jhohertz@golden.net)
Mon, 21 Jun 1999 13:36:39 -0400 (EDT)
Thank you for your insightful response.
-----------------------------------------------------------------------------
/\ Joe Hohertz - Senior Systems Administrator
/ \ jhohertz@golden.net http://home.golden.net/~jhohertz
/____\ Golden Triangle Online
-----[The opinions expressed by me do not necessarily match those of GT]-----
On 20 Jun 1999, Gisle Aas wrote:
> Joe Hohertz <jhohertz@golden.net> writes:
>
> > I used v5.43. I'm not sure it's a 64-bit issue, but it works fine on BSDI
> > and Linux/i386, but fails for Linux/alpha. I'm even *less* sure it's a
> > 64 bit issue as I see no XS code is used, meaning a 64bit issue, if it
> > exists AT ALL, would be deeper into perl than this module. I think. :)
> >
> > ------------- In 'make test'
> >
> > base/date...........FAILED tests 32-33, 38
> > Failed 3/60 tests, 95.00% okay
> >
> > .
> > .
> > .
> >
> > Failed Test Status Wstat Total Fail Failed List of failed
> > -------------------------------------------------------------------------------
> > base/date.t 60 3 5.00% 32-33, 38
> > Failed 1/20 test scripts, 95.00% okay. 3/272 subtests failed, 98.90% okay.
> >
> > -------------
> >
> > So we're not doing *TOO* badly by this.
> >
> > Any idea where to start? I looked in t/base/date.t to see if I could
> > figure out what went, but I have no idea how 'make test' assigns numbers
> > in the date.t file. Just figured it out. :)
>
> >
> > All the failures (I think!) lie in the 'try some out of bounds and garbage
> > dates' bit. So I wonder if I should worry all that much?
>
> The test script is buggy. If should not assume that we can't
> representet time before 1970 or after 2038.
>
> > What do you think? :)
>
> I'll zap those tests.
>
> Regards,
> Gisle
>