Re: Geospatial XG telecon - 21 June

Dan,

Fair enough, I need to post the draft Note this week which will  
explain the neogeo concepts a bit more clearly. The serialization  
should be the same as GeoRSS, so the examples there should be useful  
(which we also need to expand).

Chris, what do you think about adding in geo:lat and geo:long to  
neogeo for "backwards compatibility", at least for now? There will  
presumably be a mixed reaction, but compatibility would be useful as  
long as we clearly indicated how the properties are being defined in  
the context of the newer ones.

The immediate question is what to do past this week once the  
incubator expires. Should we continue with experimentation through  
georss.org or SWIG? Others have supported establishing a GWIG  
(geosemantic web interest group), but it clearly needs canvassing of  
support to go forward. In particular, we need a vibrant forum if we  
are going to formulate any Recommendation level work such as the  
geospatial relationship markup language (GRML?) that Mike and I have  
discussed.

Thoughts?

--Josh

On Jun 19, 2007, at 9:31 AM, Dan Brickley wrote:

>
> Joshua Lieberman wrote:
>> Chris,
>> Glad you'll be able to join on Thursday. The incubator's charter  
>> expires next Saturday, so we have to get neogeo ensconced  
>> somewhere and figure out if / how to propose a geosemantics  
>> interest group as a follow on.
>> There has been plenty of discussion about geo:lat and geo:long.  
>> With due respect to Dan for being a pioneer here, the consensus  
>> has generally been that a different vocabulary would more clearly  
>> mark a transition to feature-based tags.
>
> I'm plenty fine with a new vocab being created. I was more  
> concerned with potential for non-backwards-friendly changes, than  
> with standing in the way of progress. The point of doing the SWIG  
> vocab was to do the tinyest thing we could possibly have done in  
> RDF geo space while still being useful. It's time to go a bit  
> deeper :)
>
>
> 		 We could certainly, however, add them as feature
>> properties to neogeo and express the assertion that taken together  
>> they also should be considered to map to the general feature model  
>> with point geometry in the same way as georss:point. Is this  
>> valuable? It would be useful feedback, of which we have not gotten  
>> very much to date.
>
> Sounds plausible to me, but I don't know neogeo well. Perhaps we  
> could get some test instance data together? And experiment with  
> transformations (using SPARQL, XSLT, whatever, ...).
>
> Dan
>
>> Cheers,
>> Josh
>
>
>

Received on Tuesday, 19 June 2007 13:59:04 UTC