[uf-discuss] Resoling the "exclusive end date" issue

Andy Mabbett andy at pigsonthewing.org.uk
Tue Aug 21 03:06:53 PDT 2007


There is an outstanding hCalendar "issue":

<http://microformats.org/wiki/hcalendar-issues>

around the use of exclusive dtend for whole-day dates:

	<abbr class="dtend" title="2007-01-01">31 December 2006</abbr>

which is also one of the concerns with the misuse of abbr. This behaviour
is clearly counter-intuitive and problematic for publishers:
	<http://microformats.org/wiki/hcalendar-examples-in-wild#Examples_with_some_problems>

and is not available in some CMSs, for example MediaWiki; it is notable
that even the microformats "wiki" cannot have end dates entered using this
pattern:

	<http://microformats.org/wiki?title=events&diff=19616&oldid=18921>


Though hCalendar did not go thorough the microformats process, it seems
unlikely that publishers are already publishing exclusive end dates on web
pages; and that this pattern would not have been used had the process been
followed (since a negative cannot be proved; evidence to the contrary is
invited).


One of the proposed remedies is to create a new class in hCalendar (and
other microformats requiring and end date), perhaps "dtendinc" (for
"inclusive date end") or "dtendwhole" (for "whole-day date end"), such
that, in :

	<abbr class="dtendinc" title="2006-12-31">31 December 2006</abbr>

parsers are instructed, when exporting iCalendars, to increment the date
in order to generate the exclusive date-end required by iCalendar.

A "dtendinc" with a time component, or simply comprising a year or
year-month, would be equivalent to the same value with "dtend".


As discussed previously:
	<http://microformats.org/discuss/mail/microformats-discuss/2007-August/010433.html

it would be sensible for this solution to be implemented, subject to the
usual caveats about process and testing, before the watershed formed by
the release of Firefox 3.

Other solutions to replace the abbr-design-pattern with something more
accessible should, of course, also allow for this change.

-- 
Andy Mabbett
** via webmail **



More information about the microformats-discuss mailing list