[georss] Resimplification

[Allan Doyle]

On Mar 2, 2007, at 16:55, Pat Cappelaere wrote:
>>
>> Define Data Provider. USGS? Mass GIS? City of Boston? Local police
>> blotter blog? Birdwatcher blog? High-school weather station?
>>
>> I really care more about the end of that list than the beginning.
>
> Then you really have a problem because the specs as written may not be
> targeted to that high school level!  They should and I agree with you.

Are we reading the same spec? http://georss.org/simple.html

I think this is understandable by a high-school kid who already knows  
HTML. Or it's pretty close. More examples would surely help.

...snip...

>>> We are going to have a hard sell to say:
>>>
>>> Guys, you need to support:
>>>
>>> 1. GeoRSS/Atom 1.0 GML Simple
>>> 2. GeoRSS/Atom 1.0 GML Complex
>>> 3. GeoRSS/RSS 2.0 GML Simple
>>> 4. GeoRSS/RSS 2.0 GML Complex
>>> 5. GeoRSS/Atom 1.0 KML
>>> 6. GeoRSS/RSS 2.0 KML
>>
>> Stranded on a desert island, I would pick these, in this order
>> 1. GeoRSS/Atom 1.0 Simple
>>
>> And I would not call it GeoRSS/Atom 1.0 GML Simple. I would prefer to
>> call it GeoRSS.
>>
>> I caution against a sentiment of "our list of choices is too complex,
>> we have to throw a bunch of stuff out, so let's keep the most general
>> thing."
>>
>> Allan
>
> But this is not our current message to data providers.
> We are saying:
> GeoRSS/Atom 1.0 GML Simple
> GeoRSS/RSS 2.0 GML Simple

I don't understand where the "GML" has crept into our naming scheme  
for Simple.

>
> And then
> GeoRSS/Atom 1.0 GML Complex
> GeoRSS/RSS 2.0 GML Complex
>
> And coming soon:
> GeoRSS/Atom 1.0 KML
> GeoRSS/RSS 2.0 KML

This makes me wonder. If OGC is endorsing KML as a best practice and  
if KML is going to have limited harmonization with ISO and GML, then  
why did we originally have to jump through so many hoops to make  
GeoRSS Simple be a subset/profile of GML? Particularly if there's now  
a move afoot to have a KML version of GeoRSS.

>
> This is the current message to the data providers.  You are not  
> stranded on
> an island.

I didn't mean that I was stranded on an island of non- 
interoperability. I meant if I were literally shipwrecked and I was  
only able to save one spec from the list, I would save Simple.

> GeoRSS encoding is being used for WFS-Basic (or Simple).  Some  
> clients will
> ask for RSS2 or Atom1.0 output, right?

But that's a mere XSLT script away (I think). I wonder if  
feedburner.com's automagic feed translator would just handle this.

> So now encoding might be simple or complex (depending on the  
> feature set
> being served)
> And we need to support KML for the other users!
>
> I guess I am talking as the GeoBliki Data provider.
>
> This is my reality.
> Pat.
>

-- 
Allan Doyle
+1.781.433.2695
adoyle at eogeo.org