What I was envisioning for 4) (but didn't elaborate clearly) would be
that our data model act as if indefinitely recurring events more than,
say, 3 years in the future don't exist. We would preserve the fact that
they're indefinitely recurring, all the relevant occurrences would be
updated nightly and exist as fully fleshed out first class items. So in
my picture of 4), the query system wouldn't need to have any semantic
understanding of recurrence.
Wouldn't that provide the wrong query result for any view that doesn't show a
limited time range? For example, a summary/table view should only
show recurrence masters and possibly recurrence exceptions
If I understand correctly, when a user does a contains search on calendar
events, we would show a search results pane that only shows recurrence
masters, not recurrence instances. This is not what iCal does, however,
iCal shows regular recurrence instances in the search results pane. I
can imagine showing exceptions (particularly if the recurring meeting
exception contains the search term and the recurring meeting master
does not) but I think our plan is not to show unexceptional instances.
Lisa
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "Dev" mailing list http://lists.osafoundation.org/mailman/listinfo/dev
