We’re back from FOSS4G and have been reflecting on our jam-packed week in Denver. Our only regret is that we were so busy we missed some of the plenaries, workshops and talks that we wanted to attend. In case you’ve found yourself in a similar situation or would like to review an OpenGeo presentation we’ve compiled the materials below.

First two videos that we’re lucky that the organizers captured:

Paul Ramsey	An exploration of open source business models
Mike Pumphrey	Why your product sucks

And next (almost) all of the materials presented by the OpenGeo team:

Paul Ramsey	An exploration of open source business models
Paul Ramsey	State of PostGIS
Sebastian Benthall & Rolando Peñate	Designing a better SDI
David Winslow	GeoServer Scripting with Python and RESTConfig
Bart van den Eijnden	Introducing GXP: Webmapping made easy
David Winslow & Alyssa Wright	Mapnik2GeoTools
Paul Ramsey	Tips for the PostGIS Power User
Mike Pumphrey	Why your product sucks

We had a great week in Denver and have always felt that FOSS4G is the most interesting and informative conference of the year. FOSS4G certainly proved to be no exception. If you had a conversation with someone at our booth, or at a presentation, and would like to follow up please feel free to Contact us. See you in Bejing!

 

Talks, talks, so many talks!

Opening plenary Thursday was all open data, all the time. It started with Peter ter Haar from the UK Ordnance Survey, not an organization renowned for its openness, but one that has been wrestling with existential questions and moving towards open over the last few years: how to meet government open data mandates while still performing cost recovery? Peter says OS has moved from an everything-closed licensing model to a freemium model, where some data and services are free and other premium ones are not. It will be interesting to watch Ordnance Survey continue to evolve over coming years.

Then we got a taste of Open Street Map, which made me wish I’d been able to come a few days earlier. The OSM community continues to be a real hotbed of great news ideas for engaging folks in geography. My favourite from the talk: a mobile address verification application, but cast as a video game, with a medieval theme and rankings of the top players.

Finally, Michael Byrne presented on the open data aspects of the FCC Broadband Map and Developers API initiatives. While the open source aspects of Michael’s story is compelling (turning around a high visibility project with incredible load in only a few months using an open source geostack) I find the open data aspects even more compelling. The whole site is based on APIs. So the architecture completely separates the data feeds from the user interface, which makes the UI completely modular and ensures that everything is available to third parties for re-mixing. This architectural practice — building APIs, then building sites using only those APIs — is the best possible long-term approach to open data, because it embeds openness at the core of the system.

Friday is PostGIS day in the Windows room (yes, the Windows room), power users, 2.0 features, raster, replication, query, oh my!

Let the games begin!

At the opening plenary, Peter Batty again took the conference to the “neutral zone” and talked about evaluating all software, open source and proprietary, on a level playing field. Does it do what you need, is it supported, does it have a development future, it is a reasonable price? Good things to think about while evaluating options!

Arnulf Cristl covered the OSGeo point of view, how the organization brings together the diverse communities of open source GIS into an cooperative organization to promote the general growth and health of open source geospatial.

And finally, I got to spend 20 minutes on stage talking about open source business models. As usual, it was a harrowing but enjoyable experience to talk to so many people at once! Video? Of course there’s a video, many thanks to the conference organizers for capturing it.

After the plenary, the technical sessions began and I started off with a recap of the state of the PostGIS project, then sat down to take in a case study from the Norwegian Forests agency. They are migrating from Oracle to PostGIS and have built a clever tool to allow them to replicate from their Oracle to their PostGIS instance during the long period in which they have to run both system simultaneously. Change takes time!

Unfortunately, for the rest of the day I had meetings on the floor and elsewhere, so I took in no other talks to report on! The energy in the conference, on the floor and in the plenaries so far has been amazing. I am tempted to do my best Juan Antonio Samaranch impression and declare Denver to be the “best FOSS4G ever!” Here are some pictures from the conference.

This year, the second workshop day included a new FOSS4G feature: the “FOSS4G for New-comers” session. The all-day track was basically a non-technical introduction to the philosophy of open source, case studies from organizations deploying open source, and a guide to some of the terminology folks could expect.

The session kicked off with Peter Batty explaining his history with FOSS4G, which (gratifying for me) started at FOSS4G in Victoria, exploring PostGIS. Peter covered his rationale for moving increasingly into open source, including his recent shepherding of the Ubisense infrastructure from a largely Google-bases system into an open source system. It all comes down to control: Google’s technology was fine, but the terms of use and lack of control over the future of system was too big a business risk to take over the long term.

A recurring theme throughout the session, pointed out by moderator Brian Timoney, was that of “hybrid approaches”. Using open source or proprietary software is not an either/or proposition, it is an and/also proposition. Every case study cited involved some kind of bridge or connection between open source and proprietary software.

My favorite presentation was from Michael Byrne of the FCC on why they chose open source for the National Broadband Map. And not just because he said the support he got from OpenGeo was “fantastic”. They benchmarked their options, and build prototypes to make sure they were making the right decision, and chose open source on the merits: it was faster, it was more flexible, it fit their needs for handing the expected huge traffic (500K visitors on day one) the map would generate.

Unfortunately the audience had very few questions, so I wasn’t able to meet my personal goal for the session: getting a gauge of what issues are top of mind for managers taking their first foray into the world of open source.

Tomorrow the plenaries and technical sessions begin, and there are many many talks to see. OpenGeo has lots of talks (too many, sadly, for me to go to them all and also see the other ones I want to) this year, and we’ll be having office hours at the booth again this year for folks who need their PostGIS/GeoServer/OpenLayers/GeoExt/GeoNode questions answered.