emacs-orgmode
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Inheritance of LOCATION when exporting ICal (Was: Re: [Orgmode] Bug


From: Carsten Dominik
Subject: Re: Inheritance of LOCATION when exporting ICal (Was: Re: [Orgmode] Bug in org-export-icalendar)
Date: Fri, 14 Nov 2008 08:02:00 +0100

Hi Harri,

On Nov 13, 2008, at 3:41 PM, Harri Kiiskinen wrote:

From: Carsten Dominik <address@hidden>
Subject: Re: [Orgmode] Bug in org-export-icalendar
Date: Tue, 11 Nov 2008 17:34:20 +0100

yes, only times in time stamps will be used in iCalendar export. Times in the headline are too unpredictable. The agenda uses them as a service because nothing really bad can happen. But if you want consistency across applications, you should define times in a consistent way, i.e. in the stamp.

Ok, explanation quite satisfactory. Second wish:

When exporting to icalendar, would it be possible to make the property
LOCATION to respect the setting in org-use-property-inheritance? I
like to group seminars under one heading, so I could just set the
LOCATION property there, and still get it exported to the .ics file
for each of the sub-headings.

If I understand correctly, the place to change is in org-exp.el, lines
4336 and 4458: (org-entry-get nil "LOCATION") from nil to 'selective'
or something.

almost correct,

  (org-entry-get nil "LOCATION" 'selective)

would be the right thing to do.  Why don't you try locally
and let us know how it works?  One problem is that if you have
large file with N entries, using property inheritance will
eventually make the time required for export scale as N^2,
because for each matching entry, Org will have to scan back
to top-level to find if it should inherit.  On could use a
different implementation to make sure that is scales as N, but
the current algorithm does not.

Perhaps the selective inheritance could be allowed for DEADLINE
and SCHEDULED, too? I could imagine using this, too.

Same problem as above, and I am no sure if it would break anything.
Maybe maybe not.

- Carsten



Best,

Harri K





reply via email to

[Prev in Thread] Current Thread [Next in Thread]