Tuesday, June 27, 2006

OGC Technical Committee Meeting - Edinburgh

I'm sitting here in Edinburgh at the OGC TC meeting. Its day two and I've caught up on my email :-) Day 1 went quite well, we started off by proposing to create a security WG to separate securing your data on the web from digital rights management. With any luck this will lead to fewer flame wars on GeoWanking when people produce open source products which allow you to secure your WMS. I might have upset Graham Vowles from Ordnance Survey by alluding to OS being seen as the spawn of the devil.

There was a brief discussion about the future of SLD with respect to what OWS 4 was up to. This seemed to boil down to SLD is dead long live Symbology Encoding. The only change is to remove UserLayers from SLD and move the binding of a style to a data layer to an annex to WMS, WFS, WCS etc.

On day 2 I'm now sitting in GeoDRM (we're discussing the future of GeoDRM) the debate seems to still be how should we do this, are proxies the way to go or do we need to build directly into the services, which language should we use and how do we integrate geo in to DRM.

I'll post later with some thoughts on the crisis management session and tomorrow's grid workshop.

Friday, June 02, 2006

What am I doing?

I seem to have sent out a lot of email this week with links into my online bibliography so I thought I'd share my author and tag cloud here. This should give people a better idea of what I'm currently researching for my day job. You'll see that I'm doing a lot of reading on collaboration, geocoding, geographic name disambiguation and visualisation.

Most of this week has been taken up with writing a paper for the ACM GIS meeting in November, we've been experimenting with writely as a collaborative tool, but most of my co-authors prefer to use word :-(

Next week looks like a coding week :-) with the need to reintegrate StyledMapPane in to the GeoTools trunk/2.2.rc1 code base so that we can use it in GeoVISTA STUDIO and some problems with STUDIO to be fixed too.

Wednesday, April 26, 2006

Random thoughts about maps etc

Well we finally made it through demo season here in GeoVISTA. I managed to build a collaborative map exploring/sharing application (which sadly is on a locked web site). But basically it works in the way I laid out here a few posts back. As users request data from GeoServer I intercept the request and store the bounding box and other details of the request and then pass the request to the server. Users can then request a geoRSS feed of these transactions and see them displayed on a map using mapbuilder. If desired users can opt to follow another user so that their map view is updated to match the other users as they move about the map. All in all we were quite pleased with the results. We've also started to implement using a WFS/T to allow user annotation on the map. This works, but we want to extend it so users can add to each others comments and older comments become more translucent with age.

I've recently become fascinated by folksonomies such as del.icio.us and CiteULike. I've been mulling over what you would need to do to make a similar system for maps. Ideally I'd like a user to be able to look at any map on the web and tag it to a central database. But to start with I was thinking of just building a custom map front end on top of a catalogue and allowing users to browse and tag through that. Then the next time you wanted to find a map of sea surface temperature for the Mexican Gulf quickly it would be at your fingertips along with other relevant hurricane maps.

This then leads back to work as I'm also investigating ontologies, semantic web type things and how they might integrate with mapping and analysis tools. So I might actually get around to building a test system.

Wednesday, March 29, 2006

Top 10 WMS layers

Matt at PerryGeo lists his top 10 wms layers - it's an interesting list but I feel a little amerocentric but I haven't actually checked how many of them extend worldwide as my network is flaky today.

He makes a good point though that it is very hard to find good wms layers.

Friday, March 17, 2006

More on collaborative mapping

Previously I talked about how I would like to build a collaborative web portal system. Well during spring break I got some coding done and now things are starting to take shape. Here's a diagram of how it works (or should work).

Basically someone using the web portal (in yellow) has a mapbuilder client in one of the portlet windows which makes a wms request (1) to the collaboration facade, a simple servlet which passes the request on unaltered to GeoServer (2) (or which ever WMS you are using) while doing this it also extracts the bounding box of the request and the layers etc requested. This information it puts in a georss feed using a modified version of Informa that knows about geography (4). Meanwhile the WMS has returned a map to the client (3).

Another user of the portal polls the georss feed (5) and displays the bounding box in a small reference map so they can "see" where their colleges are looking at. If they are collaborating directly they second user can set their client to follow the leader (6,7) so that they are seeing exactly the map that she sees.

So far I've built the collaborative facade and I'm working on 1) making the boxes in the georss feed show up in mapbuilder and 2) thinking about how to do follow the leader in mapbuilder.

The first step has moved forward today as I've upgraded to firefox 1.5.1 and the mapbuilder head. I can see the demos of georss working now. So I think the next step is to work out how the shipping demo draws the tracks and look at drawing boxes somewhere similar and reading more complex georss documents.

Thursday, March 09, 2006

Guardian Bashes Ordinance Survey

There is a big debate going on over the freedom (as in beer) of public(ish) geodata in Great Britain started by the Guardian (a left wing newspaper). Obviously Ed Parsons has something to say in his completely not speaking for OS blog. GeoCarta: British Newspaper Bashes Ordinance Survey also has some discussion.

My take on the issue is that Crown copyright data should be freely available as generally speaking the public taxes have paid for it so they should see the benefits. Ed of course argues that taxes don't fund OS any more, which is true if off the point. What he conveniently forgets that until recently the OS was wholly funded by tax payers. So the majority of features on OS maps were mapped by a tax funded body - has your house changed much since 2000? OK so there are new areas of building and new roads and such like to be added but that's really not that much in the overall scheme of things. It should also be noted that for the wild and inaccessible parts of the country the government already pays OS to map these areas, sort of a mapper of last resort since no one else would map them but they are needed for emergency planning etc.

So the question is would it kill the OS to free its data? I guess that depends on how they do it. Obviously if they have to continue to make a "profit" for the treasury then they would be in trouble, though I'd guess that people would still buy the printed OS maps. However if we got smart and reckoned in the cost to other government (national and local) and public-sector agencies (universities, hospitals etc) as a cost against OS we could afford to fund the civil servants like Ed directly from taxes and still show a profit. Then money OS could make from consulting, selling paper maps etc would be profit too.

Disclaimer: I like Ed - he bought me a drink once, my wife used to work for the government (so I know about the idiocy that goes on) and I have moved to the US so it won't be my taxes that pay for this!

Thursday, March 02, 2006

Our new House

Our House
Originally uploaded by ianturton.
A picture of our new house on the edge of State College. Geotagged to within a picodegree of it's life in case you ever want to visit.