[ic] time tag with adjust in February bug?

Paul Jordan paul at gishnetwork.com
Sun Mar 31 14:59:04 UTC 2013


> On 03/31/2013 04:27 PM, Paul Jordan wrote:
> >
> > Doing this today:
> >
> > A)    [time]%c[/time] <br />
> > B) - 1d: [time adjust='-1 day']%c[/time] <br />
> > C) - 1m1d: [time adjust='-1 month 1 day']%c[/time] <br />
> > D) - 2m1d: [time  adjust='-2 month 1 day']%c[/time] <br />
> >
> > gives me:
> >
> > A)                Sun Mar 31 07:04:42 2013
> > B) - 1d:        Sat Mar 30 07:04:42 2013
> > C) - 1m1d:   Sat Mar 2 06:04:42 2013
> > D) - 2m1d:   Wed Jan 30 06:04:42 2013
> >
> > Notice run C. Reaching past February seems to produce logical results,
> > and I have not noticed the system this code is running in producing
> > oddities until this, so think it has something to do with February
specifically.
> >
> > I guess it comes down to deciding what happens when minus one month
> > and one day is logically the same over a period of days. Would it be
> > Feb 27th on March 28, 29, 30, and 31?
> >
> > What is adjust="- 1 month" doing? I would think it just reduces the
> > month field date-wise, and checks the if the day number exists in that
> > result month, and if not - use the last day number available for that
result
> month.
> >
> > Does that make presumptions not in line with the intentions of that
> > facility?
> >
> 
> The difference of one hour can be explained by the Daylight Saving Time
> (DST).
> Use UTC "timezone" for calculations which shouldn't be affected by DST.


Yes, of course. I didn't think I needed to mention that, but for
completeness it is good to point that out. Now to figure out why C shows
March 2nd.

Paul



More information about the interchange-users mailing list