On Mon, Dec 8, 2008 at 6:30 AM, Aristotle Pagaltzis <[EMAIL PROTECTED]>wrote:
>
> * Peter Keane <[EMAIL PROTECTED]> [2008-12-08 05:30]:
> > But the spec says it "assigns no meaning to the content" and
> > the RNC seems to allow it.
>
> If the validator went strictly by the spec (the RNC is informal
> anyway) then it couldn't flag many obvious errors because of
> Atom's mustIgnore nature. Typos like `<caetgory>` would have to
> be considered valid, f.ex., even though nothing will process such
> a feed correctly.
>
> > I am getting an error from the feedvalidator for text in the
> > atom:category element:
> >
> > <category term='breed'>siberian husky</category>
> >
> > I'd like to use atom:category here instead of:
> >
> > <x:breed>siberian husky</breed>
> >
> > which DOES validate.
>
> I am not sure what it is supposed to mean. Isn't the text in the
> element semantically significant? It seems to me you are looking
> for this:
<category
>
scheme="http://example.org/breed";
term="siberian-husky" />
>
Yes, that's the idea -- I need to apply arbitrary key-value pairs to an
entry. I'd like to use inner text, though, since the value could be of any
length (maybe even an entire essay). Right now, I approach it like Google
Spreadsheets does:
<entry xmlns:gsx="http://schemas.google.com/spreadsheets/2006/extended";>
....
<gsx:type>Task/Job Request</gsx:type>
<gsx:eid>pppp</gsx:eid>
<gsx:name>hgkjhgk</gsx:name>
<gsx:text>some arbitrary textj</gsx:text>
</entry>
But was thinking I might use atom:category instead. I realize that
atom:content would be a logical place for this stuff (perhaps as RDF or
XOXO), but many of these entries *do* in fact, have textual content that's
not as much like "metadata" as these key-value pairs.
--peter
> Maybe with a `label="siberian husky"` thrown in for humans.
>
> Regards,
> --
> Aristotle Pagaltzis // <http://plasmasturm.org/>
>
>
