call for presentations at the Data Model sessions in Cambridge , September 2007

[Alasdair Allan]

Doug Tody wrote:
> Alasdair Allan wrote:
>> Lets not make the same mistake the professional GIS community made  
>> with Google Earth...
>
> Which people - the scientific GIS community, or the public which
> uses Google Earth instead of something like the open source World
> Wind visualization tool,

World Wind is seriously impressive, but up until very recently it was  
not cross platform, and many of the content creators weren't using  
Windows. So there has been a much smaller take up. Possibly the Java  
version will solve this, but I'm not holding my breath. But then  
World Wind reads KML too, doesn't it?

> or the many other GIS visualization tools out there?

The open source vs closed source debate isn't really as relevant here  
as it is elsewhere, the file formats, APIs and other such stuff are  
all published standards. If you don't want to use Google's tools  
that's fine, write your own. But I'd advise you to support KML as an  
interchange format if you do, because that's what people are going to  
use, and that's what most of the content is going to be marked up in...

>  I suspect that products like Google Earth or Microsoft
> Virtual Earth will rule where public outreach is concerned, but
> open source tools based on open standards will rule where serious
> scientific visualization is concerned.  The public-oriented tools
> are unlikely to ever address all the issues important to either GIS
> or astronomical science applications.

Depends what you mean by public-orientated tools. I think you're just  
not seeing the opportunities that the new tools are presenting us  
with. Just because they're not not addressing issues you find  
important, it doesn't mean they aren't addressing issues I find  
important, or other astronomers find important.

 From a Google Earth end of things I know several professional GIS  
people who are ditching ESRI ArcGSI and, at least for visualisation,  
are now using Google Earth. Because it's "good enough". Sometimes  
that's all that's needed...

> Regarding KML, it appears to be pretty much a Google thing at the
> moment,

Yahoo have also adopted it inside Flickr as their export option for  
geo-tagged photographs. My bet is that it'll become the standard for  
such things, there really isn't any competition worth the name.

> however I see that it is being submitted by Google to OpenGIS
> for development of KML 3.0 as an open GIS standard.  It is oriented
> towards driving GIS-style visualization, and is very different from
> something like VOTable.

Err, yes? I don't see your point. I wasn't suggesting replacing  
VOTable with KML, that's obviously nonsense. I was suggesting  
adopting KML for KML type scenarios inside the VO, and not going out  
and reinventing the wheel. For instance, for visualisation purposes  
KML is an excellent alternative return from something like Cone  
Search or Simple Time Access Protocol (STAP) or its eventual  
replacement Simple Event Access Protocol (SEAP).

> I doubt if KML itself would address something
> like coordinate systems, rather this would be something the client
> app and server would need to negotiate and support.

True, KML uses a <Position>x,y,z</Position> tag and really doesn't  
specify what those mean that heavily. It's up to the client as far as  
I can see.

When it comes down to it, I think KML is going to start getting  
heavily used. I think the IVOA should think seriously about putting  
it's official blessing on such use, because it's going to happen with  
or without it...

Al.
-- 
Dr. A. Allan, School of Physics, University of Exeter
eSTAR Project, http://www.estar.org.uk/