Re: [PATCH 0/1] MR1660: mstask: Don't assume a ONCE trigger is on the current day.

[Dmitry Timoshkov]

"Sven Baars (@sbaars)"  wrote:
> On Fri Dec  2 14:16:59 2022 +0000, **** wrote:
> > Dmitry Timoshkov replied on the mailing list:
> > Sven Baars <[e-mail]> wrote:
> > > -                if (CompareFileTime(&begin_ft, &trigger_ft) <= 0 &&
> > CompareFileTime(&trigger_ft, &end_ft) < 0)
> > > +                if (CompareFileTime(&trigger_ft, &current_ft) >= 0)
> > >                  {
> > >                      if (!have_best_time ||
> > CompareFileTime(&trigger_ft, &best_ft) < 0)
> > >                      {
> > Removing the trigger end time check doesn't look right.
> > -- 
> > Dmitry.
>
> Hi Dmitry,
>
> Thanks for the reply. I don't see how begin_ft and end_ft are relevant for this trigger. It happens exactly once, at the "Start" time and "Begin" day.
>
> Having an "End" day before the "Begin" day is invalid for all triggers. I could add a patch that checks for that.

Obviosuly the trigger shouldn't fire outside of begin/end time, so changing
the trigger time bounds check doesn't look right. At best, that should be
a separate patch with a bunch of tests.