[georss] GeoRSS Location References

[Sean Gillies]

Raj,

1) I don't want to blow off RSS 2.0 users, but it seems like the link to
geometry idea is coming entirely from the Atom quarter. My own use case
for it is connected to use of other features absent from RSS 2.0
("related" links, for example), so I don't have a need for a
Atom-independent way.

2) Hinting the geometry type isn't important, but I think that hinting
whether you are linking to a geometry entity or a "feature" entity (Atom
entry, KML placemark) will be important.

Sean

Raj Singh wrote:
> I'd love to get the location references issue decided soon. We've been  
> sitting on this stuff too long. So let me summarize what I think is  
> out there right now.
> 
> Andrew proposed a couple options -- http://georss.org/proposals/external_geometry 
>   -- that preserved some of the GeoRSS language in the link (either  
> <where> or the geometry element).
> 
> Josh and Sean support a strategy that's a more generic, universal way  
> of handling external references (see quoted text below) by working  
> with <atom:link>.
> 
> This leaves me with 2 questions:
> 1) Sean, does your proposal only work with Atom? Do we care?  
> Personally, I'm for dropping support of all the other "legacy" RSS  
> formats.
> 
> 2) Is there any important advantage for software to know a little  
> about what the referenced link will contain prior to dereferencing it?  
> If so, then Andrew's ideas where you still have a <where>, <point>,  
> <line> or <polygon> is useful.
> 
> If the answer to #2 is not really, which I think it is, then I'm a big  
> +1 on going with something like what's shown at http://atlantides.org/trac/concordia/wiki/AtomLocationByReference 
> .
> 
> ---
> Raj
> 
> 
> On Mar 28, 2008, at 8:03 PM, Sean Gillies wrote:
>> More specifics here:
>> http://atlantides.org/trac/concordia/wiki/AtomLocationByReference
>>
>> Sean Gillies wrote:
>>> I'm -1 on that one (origin:
>>> http://lists.eogeo.org/pipermail/georss/2007-November/001752.html,
>>> IIRC). -0 on the first, which is more of an interesting idea than
>>> something we should start implementing now.
>>>
>>> I think some standard link relations (ala
>>> http://lists.eogeo.org/pipermail/georss/2007-November/001755.html)  
>>> might
>>> do the trick.
>>>
>>> Sean
> 
> _______________________________________________
> georss mailing list
> georss at lists.eogeo.org
> http://lists.eogeo.org/mailman/listinfo/georss
>