Re: [O] Yearly repeats on the agenda

[Nick Dokos]

SW <address@hidden> wrote:

> SW <sabrewolfy <at> gmail.com> writes:
> 
> > This
> > 
> > *** New Year's Day
> > <2011-01-01 +1y>
> > 
> > does *not* include the timestamp in the agenda, yes.
> > 
> > However, timestamps are *not* included in the agenda from other entries 
> > which
> > *do* have timestamps in the headline.
> > 
> > I've tested with repeating timestamps, timestamps with times, timestamps
> > repeating with last year as the start date, and I cannot replicate this. 
> > I'll
> > post if I find anything further.
> 
> I've tracked down what causes this behaviour -- it's actually a repeating
> timestamp which is from a year ore more ago (contrary to what I posted above).
> 
> This:
> 
> ** <2011-04-17 +1y> Test     :holiday:
> 
> or this:
> 
> ** <2010-04-17 +1y> Test     :holiday:
> 
> appears in the agenda *with* the <> timestamp included. This:
> 
> ** <2012-04-17 +1y> Test     :holiday:
> 
> does *not* appear with the <> timestamp included. The difference is the
> *starting* year.
> 

Indeed - I can reproduce that. It happens in org-agenda-get-timestamps,
in the call to org-agenda-format-item: this function takes a regexp
argument, remove-re, and removes any matches from the string it
produces. The regexp is constructed from the *current* date though:

          (concat
           (regexp-quote
            (format-time-string
             "<%Y-%m-%d"
             (encode-time 0 0 0 (nth 1 date) (nth 0 date) (nth 2 date))))
           ".*?>")

so it becomes "<2012-04-17.*?>". Hence it removes the date in the third
example above, but not in the other two.

The question is whether this is intended or not: personally, I don't see
any reason for the difference in behavior, so it might be a good idea to
generalize the regexp to match *any* year.

Nick