[georss] using categories for disambiguation

[Pat Cappelaere]

Andrew,

I do not see anything wrong to say:
<category>mountain</category>
No even a need for geo:type=
Pat.

> From: Andrew Turner <ajturner at highearthorbit.com>
> Date: Thu, 8 Mar 2007 15:52:03 -0500
> To: "Gregor J. Rothfuss" <gregor at apache.org>
> Cc: Pat Cappelaere <pat at cappelaere.com>, <georss at lists.eogeo.org>, Raj Singh
> <raj at rajsingh.org>
> Subject: Re: [georss] using categories for disambiguation
> 
> On 3/8/07, Gregor J. Rothfuss <gregor at apache.org> wrote:
>>> A category is a category is a category!
>>> Having multiple categories tags make the decoding harder from an
>>> interoperability standpoint.  Everybody will be doing it differently.
>>> You can have more than one category tag anyway!
>>> 
> 
> I don't think of the location type or relationship as a "category",
> they are actual geographic meaning an relate the location
> name/geometry to the RSS item. And because they have different meaning
> (not just another category) you would end up putting in
> <category>geo:type=mountain</category> or something anyways.
> 
>> i suspect those two snuck in because they have a meaning in GIS. they
>> sure don't in the larger RSS world.
>> 
>> the 'just in case' school of standards leads to 600 page monsters that
>> no one reads unless forced to.
>> 
> 
> Well, as one who has already admitted to not sitting down in a comfy
> chair to read 600 page specs, I agree with simple & not "just in
> case". But you're advocating changing the standard just because it
> hasn't been widely used yet.
> 
> It doesn't really matter to me *how* we relate geometry to the story,
> I just want to be able to publish & consume location information and
> know how that relates to the RSS item. If that's machine tags in the
> category I guess that works, but there is a solution already somewhat
> in place in {relationship,featuretype}tag so I think it would just be
> better to promote using it.
> 
> Andrew