Michael Fuhr <mike@fuhr.org> writes:
> On Wed, Sep 07, 2005 at 18:39:26 -0300 (ADT), Tom Lane wrote:
>> Update timezone data files to release 2005m of the zic database.
>> Among other changes, this reflects the recently passed change in USA
>> daylight savings rules.
> Looks like some horology tests need to be updated as well. Example:
Ooops, didn't occur to me to check the regression tests :-(
The comments added in the zic database point out that Congress reserved
the right to undo this change if the administration couldn't prove a
sufficient energy savings. Ignoring the point that all the transition
costs would have to be paid over again to undo the law change
(commentary suppressed by sheer force of will), should we consider
modifying the regression tests to be unaffected, rather than just
updating the expected outputs? The original test author certainly
wasn't expecting dates in March to be affected by DST.
regards, tom lane