So I just finished attending my first Reboot – and as many people tried to explain before hand, “it’s weird”. Not just European Conference weird with completely different crowd, most attendees speaking in a foreign language (their own, or in English which is foreign to them), varying minor variations on “what matters”, and of course dealing with wall plug adapters.

No, Reboot is weird in a different way. In attempting to learn if it was more traditional conference, or unconference, the answer was never clear and often just “yes” to the multiple choice question. Now I know why they said that, it actually makes sense. But let me hopefully be slightly more lucid.

What is Reboot?

Reboot asks for anyone to submit topics ahead of time, and these are then chosen both by a committee as well as public voting – so it’s a very open system, BarCamp-like, but these talks are then chosen with a speaker and assigned times. So then Reboot becomes more traditional conference with many “stage talks” in a face-forward audience setting. There are some side rooms that will have sessions scheduled that are discussions, but that definitely isn’t something to plan on. So the presentations are more traditional.

But then where Reboot was really surprising is the amount of the conference that happens outside of the sessions. This isn’t just a “hallway track” discussions – they are full-fledged, conference long sessions working on projects. People are dedicated to building things, gathering together information, creating, ideating, collaborating, advising – all during the conference. I’m quite sure a sizable number of Reboot attendees never go to a ’session’ but merely use the venue as a mechanism to gather together many like-minded people who are driven to do something, and leverage the brain-power and thoughts that are coming out of the sessions to act on something bigger.

I love BarCamps – they’re discussion-centric, synergistic, and connect people in networks to carry forward and achieve great things in the future. What Reboot does, by comparison, is not wait. Have an idea? Get started on it now. Or at least be very good in capturing the idea, disseminating it widely and getting it moving as quickly as possible.

Now perhaps with a little better, or at least verbose, explanation of what Reboot was like from my perspective – it also makes the name itself more meaningful. The conference is Reboot, not “Reinstall”, or “Start-over”, or “Redo”, just “Reboot”. When you Reboot something you retain some measure of the longer-term state it was in; applications are installed, configurations are set, and so forth. But what Rebooting does is to go back to a fresh state, with the long-term memory and skills and infrastructure, and get up and running and back into business.

So the conference is about considering what’s around you, what you’ve built and have, and through the sessions, collaborations, projects, whatever, to take a fresh perspective and jumpstart on moving forward.

Action

The theme, or topic, of Reboot 11 was “Action”. Simple, single word: Action. Take action, make something happen. There is a parellel in Tim O’Reilly’s “Work on things that matter”.

The proposed reasoning behind having Reboot focus on “Action” was that we are in a global economic crisis, there are short-term issues such as disasters, corrupt governments, and long-term problems of environmental quality, health, and education. It was a push for us to work on these issues and figure out how we can help enact change.

However I found through talking with many people that had attended several Reboots that there was “nothing new”, or “revolutionary” in many of the topics and that the individual felt they already had a good grasp of what was going on in the space. It was this malais of “amazing things” that made me realize why it’s really time to take action – and Bruce Sterling to cast just enough “Gothic Hi-Tech” to make it solidify”"

We have the tools, we have the power

The tools we’re all using and building with have been in active use for several years now. We have our wikis, blogs, social networks, mobile devices, media devices, connectivity, realtime communications, hardware interfaces, API’s, and more. We have all created an amazing toolset that has been used to create many varied, and some quite crazy, applications, worlds, communities, or systems.

And if we have these mature tools, with many choices and the ability to quickly pull them together to accomplish nearly everything – we have to grow up and realize that these are not just toys or hacks nymore. These are the very tools that can, and are already, making the world of the future.

And we, the technologists, designers, thinkers, citizens of the next generation that are now in control. We have grown up with these tools, and in many ways we’re already using them to change governments, raise communities, run businesses, and live in our world.

Act now

For me, what I took away is to take responsibility and consideration for what I choose to do. Hacks and toys are fun – but as Bruce said, “if it’s not beautiful enough to show your friends, and doesn’t have a narrative attached to it, throw it away”.

Work on things that matter, and make them work by focusing on them like they matter. Your actions will make a difference, and if they don’t – you’re doing it wrong. Put it down, and move onto something else. Collaborate and work together to achieve great, actionable outcomes. You’re an amazing person with many talents, and we can all use amazing people like you. What action are you going to evoke?

It’s been a long while since I’ve talked here in-depth about a project I’ve been helping with for a few years now. If you’re not familiar with it, Mapstraction is a JavaScript library that provides a common interface to more than 11 other major mapping providers such as Google, Microsoft, MapQuest, OpenLayers, and FreeEarth, just to name a few.

The project originally started with just three providers: Google, Yahoo, and Microsoft; and was developed by three developers: Tom Carden, Mikel Maron, and Steve Coast. At the time, it was succint and easy – using constructs such as switch statements in a single mapstraction.js file.

Over time, as more providers were added – this became quite unwieldy. 11 case statements in any method, various callback hooks necessary, and in general quite slow to load and run for the client.

Fortunately, some very bright JavaScript experts: Rob Moran, Derek Fowler, and Adam DuVander, all congregated to help architect a complete new version that takes incorporates the current evolution of Mapstraction to provide a smaller file download and computation overhead. In addition, because each provider is now split out into an individual file such as mxn.google.core.js, it makes it easier for developers to support single providers without having to worry about any other provider or their impact on other code.

New Code Neighborhood

In order to start promoting the new API and encouraging developers to come and help out, we put the source code and tickets into a Google Code project. Although, being Subversion, you still have to submit patches to get changes accepted for now. So I personally suggest working from the Github version, which will be kept up to speed with git-svn, and then you can submit patches from here to push into the ‘official’ subversion repository.

The demos on the Mapstraction homepage do well at showing the capabilities of the library, but are difficult to maintain and users and potential contributors can’t really play with them at all. So as part of our Where2.0 workshop, I put together a Mapstraction API Sandbox, built with Google’s AJAX API Playground, and running on AppEngine.

In this sandbox, you can see more demonstrations of the API, associated code, and even modify JavaScript or HTML and create your own personal copies. For example, you can see complex marker creation, the initial version 2 demo, or the marker filtering demo.

There is definitely a lot of recent excitement around the API. We decided to use Mapstraction as the basis for our own GeoCommons Maker map integration – making it easy for developers to work within a common framework (demo). So get started contributing!

For more recent news on Mapstraction checkout the recent posts onO’Reilly Radar and ProgrammableWeb.

In the last few years the increased access to streaming data, user-generated content and mobile media capturing devices has opened up new capabilities in responding to emerging, dynamic situations. I’ve been fortunate to be invovled in a number of projects dealing with immediate crisis such as hurricane evacuation or wildfires, to slow rebuilding of cities through community participation, and even more general technology development through work with Ushahidi and now Swift. I’m still new to the domain and have a lot to learn about on-the-ground action and effectiveness, and integration with first responders and officials.

It’s for just this reason – the mix of expert crisis responders, with technologists, sociologists, policy makers, and many others that there has been a desire for a venue to bring them together to share experiences, thoughts, and collaborate on innovating together. Mikel and Jesse have talked about Disaster Tech and their lessons learned – but there still needed to be a broader discussion of international, local, regional, fast, slow, prediction, analysis, response, and rebuilding around crisis. So it was very exciting that at Government 2.0 Camp, a meeting of interested people metup and we finally pulled together to make this conference a reality.

CrisisCamp will be a 2-day unconference – June 13 + 14, 2009, held this year jointly in Washington, DC and London, UK – to bring together this wide array of people, organizations, and ideas to connect and build solutions together. In DC, the event will be hosted at the George Washington University School of Policy and the Internet – a wonderfully open and well-situated venue that also hosted TransparencyCamp. Equally as exciting is that there will be a Friday night Ignite session on Crisis experience and tools hosted at the World Bank Headquarters auditorium.

Help make it happen

Besides coming to the camp as an active participant, we’re also looking for sponsors. The costs are quite low, thanks to the generosity of GWU, but there are still fixed costs to cover security, A/V equipment, rentals, and potentially food for attendees in order to keep them fueled and together to collaborate. Please contact us via email (crisiscommons (at) gmail.com) or through the sponsorship page directly. We truly appreciate your support and will make sure everyone knows about it – they will appreciate it as well.

In addition – we’re looking for Ignite speakers. While the unconference itself will be open with no preset agenda, the Ignite talks are fixed, 5-minute lightning talks meant to highlight particular stories and lessons that you would like to share with the attendees. Again, email us – crisiscommons (at) gmail.com – to let us know that you would like talk.

I realize this may come as a late notice – things have been very hectic with Where2.0 and other conferences, but we truly hope you can come to this amazing unconference to share your thoughts, tools, experiences, and help build the next generation of crisis techniques.

Signup here for CrisisCamp

Erik Wilde was pondering:

thinking of geofeeds where feed paging does not take you back in time, but increases the search radius. but how to specify paging semantics?

My first feeling is that ‘zooming out’ is not really a link. Pagination is just a crutch to deal with returning the full set of a single query in meaningful window sizes due to server response, bandwidth, client parsing, and maybe human interface. Zooming out implies actually performing a different search and would be a function of a client interface.

This mechanism is provided by OpenSearch-Geo, which communicates how a client would use a bounding box or polygon search. Therefore a client interface could choose how to zoom out or in and has the capability to query the system this way.

Feed Clusters

However, I still could imagine more specific uses for such a concept. Erik’s original idea is perhaps thinking more about using geography as a way to indicate limiting search set results. This might be done using clustering mechanism, such as k-means, similar to how one might view dense data on a map in clusters, but in search result feeds. The link elements would provide looking into any of these clusters, or zooming in.

<feed>
    <title>Search for 'coffee'</title>
    <georss:box>38.87,-77.2,38.89,-77.0</georss:box>
    <entry>
        <title>8 results</title>
        <georss:box>38.87,-77.2,38.91,-77.1</georss:box>
        <link rel="self"
          href="http://server/search.atom?q=coffee&amp;bbox=38.87,-77.2,38.91,-77.1<"/>
    </entry>
    <entry>
        <title>15 results</title>
        <georss:box>38.87,-77.1,38.9,-77.05</georss:box>
        <link rel="self"
          href="http://server/search.atom?q=coffee&amp;bbox=38.87,-77.1,38.9,-77.05"/>
    </entry>
</feed>

Structured clustering

Another way this concept could work more literally would be to utilize a hierarchy, or several hierarchies, that a client may be able to easily snap to in order to query larger or smaller geographic areas. These hierarchies are not apparent in a simple search template and can provide more semantics to indicate the larger area rather than just simply “zoomed out”.

For example, using the GeoNames GeoTree or GeoPlanet woeid’s, a search result could provide links from the specific bounding box query up to regional or districts that contain this query – as well as perhaps subsets contained within the bounding box.

<feed>
    <title>Search for 'coffee'</title>
    <georss:box>38.87,-77.2,38.89,-77.0</georss:box>
    <link rel="up"
        href="http://server/search.atom?q=coffee&amp;woeid=2347605"
        title="Search Virginia for 'coffee'" />
    <link rel="down"
        href="http://server/search.atom?q=coffee&amp;woeid=12590311"
        title="Search Arlington County for 'coffee'" />
    <link rel="down"
        href="http://server/search.atom?q=coffee&amp;woeid=12590343"
        title="Search Fairfax County for 'coffee'" />
    <entry>...</entry>
</feed>

Written on no wifi, sitting in the Detroit airport after a red-eye flight from WhereCamp.

This weekend, Tim Waters (chippy) noticed that the WhiteHouse is using OpenLayers mapping library and OpenStreetMap basemap tiles in their new Delivering on Change page.

Whether you were already serving your country, or are responding to the President’s call, share how you are delivering on change in your community.Whether it is an hour per month helping those struggling in the current economy, tutoring kids in your neighborhood every day, or anything else, we want to highlight what Americans are doing to strengthen our country.

This is very interesting on several levels. Foremost is the use of government provided (TIGER/Line) and crowd-sourced data (OpenStreetMap) in an official US Government Site. This is definitely an indicator that what were cutting edge tools have reached a critical mass to provide broad usability and appeal. Open Source? check

Looking underneath the hood, the data is provided via a KML feed (), so you can pull the data out and upload or map it however you want. Open Data? check

The site itself, Delivering on Change, is asking citizens to contribute stories and media about their personal engagement with change. This is an incredibly exciting step to ask for people to contribute to national storytelling and character. Citizen-sourced data? check

The new US administration is continually doing amazing, and open, initiatives. There is incredible excitement around Recovery.gov as a testbed for the next generation of transparency and embrace of technology and open data feeds.

Small next steps

My thoughts on interesting applications wouldn’t be complete without pointing out a couple of suggestions. While many defend the default OpenLayers controls – I personally think that implementors should take that next step and apply minor customization to better integrate the look and feel of the map controls into their site. I’ve talked before about how easy it is to change some CSS to replace the controls. Perhaps even just a darker blue background to match the White House blue in the logo. Customized?

Another, less highlighted but very important for Government sites is the integration of accessibility controls. OpenLayers supports map navigation using keyboard inputs – which provides for alternative interfaces to navigate the map. It’s not clear if this is official “508 compliant”, but at least demonstrates the potential. Accessible?

How you can help

So do you want to help make Change, especially with mapping data and technology? Come join us at the Washington, DC mapping party – currently planned for June 20 + 21, 2009 somewhere in DC (details coming soon). Or join a mapping party near you.