AGI North’s Where2Now Conference

Yesterday we had a great one day packed geo-conference in Harrogate, North Yorkshire. Cheers to all the sponsors, the AGI, Rollo for organising it and GeoPlan for hosting it. Folks from Yahoo, Google and Microsoft were present.

Presented about Open Historical Maps – with the example of the collaborative georectification and digitization application being built for the New York Public Library. You can see the slides here:

Some things of note:

  • Theres nothing geoSpecial about geoSpatial.
  • GI is just bits on computers.
  • Google does testing of their cartography and maps live on users and have metrics to see how these groups interact with them.
  • Any successful geo presentation must either have a Vermeer’s Geographer painting or Snows infected pump.
  • Microsoft Bing maps are looking bling with the OS stuff.
  • Ordnance Survey looking to release toolkits in a box using OpenLayers, Geoserver etc. No vendor locks ins there!

All the presentations were filmed,
and should be able to be viewed here:
also  keep an eye on for some more slides as and when.


Digital Geography in a Web 2.0 World

Attended “Digital Geography in a Web 2.0 World” in Manchester on Monday, hosted by the National Centre for e-Social Science (NCeSS) which support from ERSC. Primarily it was showcasing relevant projects of the Centre’s “Nodes”, from Leeds, Nottingham, Bristol, Leicester, although most of the talks were from University College London (UCL). NCeSS expressed an interest with working with OpenStreetMap and other data sources, and seems to be a good initiative – somewhat designed to make palaeographers adopt some new “emerging” practices .

Talks included “The Names project: profiling for the public and by the public”, “GMapsCreator and MapTube”, “sim city for real”, “3D visualisations”, “Intelligent Agents”, “Web 2.0, Neogeography and the Virtual World”. Interesting stuff overall, even if it showed the time lag between academic research and the current state of the art.

Now for some observations:

Andrew Hudson Smith, in a talk entitled “Web 2.0, Neogeography and the Virtual World” missed the point about OpenStreetMap somewhat – talking about his iphone, and how it was cool, as it had a GPS on it – he talked about OSM, as it “tracks you around automatically” and “GPS traces are automatically uploaded, where it makes maps”. He did showcase some of the work of UCL in Second Life, particularly in visualisation of some GIS concepts, such as Schelling’s segregation model, the game of life, which was interesting.

Richard Milton talked about GMapsCreator, a nice bit of kit that “takes a shapefile containing geographic areas linked with attributes and automatically generates a working Google Maps website from the data.”  – written with GeoTools, too. He also introduced MapTube is “MapTube is a free resource for viewing, sharing, mixing and mashing maps online”. It has a nice interface, you can layer and reorder KML and google tile layers on top of each other – although in practice it seems to exist as a showcase for outputting UCL maps. For example the map they did for Radio 4 Mapping the Credit Crunch. You can only link to a map, no embedding, no GeoRSS, no sharing. Would be nice to see some kind of discovery of datasources, publishing the catalogue to make it findable by search engines etc .The issue of copyright of some layers seems to have been skirted somewhat.

Martin Clarke. SimCity for Real. Urban and regional modelling. There was a bit of discussion at the end about the intention to increase the user base of these models. The response was that “oh, models are so complex, we cant have non-experts using this”. This is for non-expert academic and policy makers, let alone members of the public! “If people made dodgy analysis about where to place a new shop, then it would disastrous!”. Same arguments about opening up spatial analysis, until Google changed things and about public data (people would misinterpret it!). Shame. If Web 2.0 is about democratising data, web 3.0 would be about democratising the analysis of data.

Asked whether these models are used for public policy – the answer was that no, councils etc. whilst getting models and bespoke modelling software – they do not use it for making decisions. Then explained that their private customers (they have a consultancy in addition to academic use) used it, as it could be proven that money could be saved. So there is a monetary incentive not to release this analysis as service. Interesting that this public research body is indirectly helping private consultancies. Perhaps in the future grants should be awarded on condition that research and tools be open to citizens?

Most interesting presentation was from Nottingham University.

The Locata game is a game to test spatial awareness – you see a map, or relief model, and given an image of a view shed, you have to work out where that “photo” could have been taken. It requires Shockwave to play, so I didn’t test it. Another application they developed which I didn’t try, but is worth a look is “geoCode” – “Simulating QR Code Location-based Services”

Second Life got a few mentions in a few of the presentations – we are now seeing the first results of the wave of research projects using the software started 1-2 years ago, when the hype was big. Furries were mentioned about being a hindrance in publishing and talking about research! But things were interesting especially regarding the potential to use it as a virtual learning environment – and for demonstrating 3D agent based models.

overusing the word “open” at where2.0

Where 2.0 has started, alas, I’m not attending in person this year, but Nick Black spotted that my image is there instead, on the conference programme!

Big big thanks to who are streaming most of the sessions, plus are archiving a lot of them too if you miss them. Also, the irc backchannel is #where2008 on freenode. The current over-used and misused word is “open”. (last year it was “GeoRSS”).

For live blogging, John Mckerrell is doing exhaustive transcribes too.

Things of note so far:

google includes geosearch in it’s API, and tries to kill Mapufacture. People freak out about augmented reality enhanced police states. Dash opens their API, ESRI and google have a “partnership”. This partnership, basically uses google earth and Arc Server may give a push into good open source / Free Web Processing Servers, as more people see the value of real geographical and spatial analysis, and not wanting to pay tens of thousands for a paleo-program and a new server with ridiculous specs for the privilege.

BarCamp Leeds Map

Here’s a map of the registrants for BarCamp Leeds 2007, from the registration data given to me by Imran. Nice to see a wide spread of people from around the place, and, as you can see, they are mostly up North! It uses  the very good Openlayers JavaScript mapping api, and the default underlying mapping is of OpenStreetMap  but also Yahoo, and Google Hybrid map can be seen (thanks Chris for the OSM tiles!).

Looking forward to barcamping, going to do something geo, whether opensource geostack, openstreetmap, or desktop GIS stuff, who knows! Any suggestions?

(Got this up on on the day I’ve registered it)