Re: property searches for #+CATEGORY

[Carsten Dominik <carsten.dominik@gmail.com>]

On  8Nov2007, at 5:55 AM, Bastien wrote:

> Carsten Dominik <carsten.dominik@gmail.com> writes:
>
>> However, I do think that Adam's initial request to make the
>> category available as a special property for queries in not
>> unreasonble.  Or does anyone disagree?
>
> I'm convinced it's not unreasonable :)
>
>> I am not sure, though, if the #+CATEGORY category should be
>> available with `org-entry-get', because it would then be very
>> hard for the property API to make a difference between a value
>> that is intimately associated with the current entry, and a
>> value that might be derived by some other mechanism.  So here I
>> differ somewhat from Adam's feeling that category is just like
>> TODO or a tag.  It is different.
>
> Then a search like CATEGORY="cat" would also return entries which
> CATEGORY property is not "cat"... ok, maybe this doesn't hurt that
> much for search purposes.  But I expect someone will come in three
> month complaining that `org-entry-get' didn't return the category,
> even though he set it up through #+CATEGORY.

OK, here is what we will do.  We will make `org-entry-get' return
a value from #+CATEGORY if the INHERIT flag is set in the call
to `org-entry-get'.  In this case the value is inherited not from
a higher level entry, but for the "file environment", and even
top-level outline entries can ihnerit it.
We actually already have this mechanism:

    #+PROPERTY: Name Value

So we could then see

    #+CATEGORY: work

as a short-hand for

    #+PROPERTY: CATEGORY work

... which makes all of this suddenly look as if it was designed
like this from the beginning.  I like it.

Thanks again to everyone who contributed to this discussion.

- Carsten