Jon Udell

Last week Kevin Curry dug into some data about school violence in his district. In this case the data was made available as HTML, which means it was sort-of-but-not-really published on the web. Kevin writes:

Whenever I come across data like this the first thing I want to know is whether or not it can actually be used as data. In order to be used/usable as data the contents of this HTML table need to be, at minimum, copy-and-paste-able into a spreadsheet.

Or, alternatively, the HTML table needs to be parseable as data. In this case, I was surprised to find that a couple of tools I normally use to do that parsing — Dabble DB and Excel — didn’t work. That’s because Kevin’s target page doesn’t include a static HTML table. It’s dynamic instead: First you select a district, then the table appears. This mechanism defeats tools that try to parse data from HTML tables, so it’s a bad way to publish data that you want to be available as data.

Lacking the option to parse the HTML table, Kevin’s only choice was to copy and paste. That’s clumsy, and you have to be really motivated to do it, but it can be done. Here’s the Google spreadsheet Kevin made from the data he copied and pasted. And here’s the same stuff as an Excel Web App.

If you haven’t tried out the new Excel Web App, by the way, it’s interesting to compare the two. One key difference, at least from my point of view, is — not surprisingly — the Excel Web App’s ability to roundtrip with Excel. A Google spreadsheet is, at this point, more functional in standalone mode. While you can edit both a Google spreadsheet and an Excel Web App in the browser, for example, the Google spreadsheet can insert and modify charts, whereas the Excel Web App only edits data.

Of course if you have Excel you’d rather use it to insert and modify charts. It’s a lot more capable than any browser app is likely to be anytime soon. So it’s pretty sweet to be able to open the cloud-based Excel spreadsheet, edit locally, and then save to the web. A related limitation of the Google spreadsheet is that you lose charts when you download to, or upload from, Excel.

Another key difference: The Excel Web App currently lacks an API like the one Google provides. I really hope that the Excel Web App will grow an OData interface. In this comment at social.answers.microsoft.com, Christopher Webb cogently explains why that matters:

The big advantage of doing this [OData] would be that, when you published data to the Excel Web App, you’d be creating a resource that was simultaneously human-readable and machine-readable. Consider something like the Guardian Data Store (http://www.guardian.co.uk/data-store): their first priority is to publish data in an easily browsable form for the vast majority of people who are casual readers and just want to look at the data on their browsers, but they also need to publish it in a format from which the data can be retrieved and manipulated by data analysts. Publishing data as html tables serves the first community but not the second; publishing data in something like SQL Azure would serve the second community and not the first, and would be too technically difficult for many people who wanted to publish data in the first place.

The Guardian are using Google docs at the moment, but simply exporting the entire spreadsheet to Excel is only a first step to getting the data into a useful format for data analysts and writing code that goes against the Google docs API is a hassle. That’s why I like the idea of exposing tables/ranges through OData so much: it gives you access to the data in a standard, machine-readable form with minimal coding required, even while it remains in the spreadsheet (which is essentially a human-readable format). You’d open your browser, navigate to your spreadsheet, click on your table and you’d very quickly have the data downloaded into PowerPivot or any other OData-friendly tool.

Some newspapers may be capable of managing all of their data in SQL databases, and publishing from there to the web. For them, an OData interface to the database would be all that’s needed to make the same data uniformly machine-readable. But for most newspapers — including even the well funded and technically adept Guardian — the path of least resistance runs through spreadsheets. In those cases, it’ll be crucial to have online spreadsheets that are easy for both humans and machines to read.

Last week Scott Hanselman summed up the principle of keystroke conservation like so:

There are a finite number of keystrokes left in your hands before you die. Next time someone emails you, ask yourself “Is emailing this person back the best use of my remaining keystrokes?”

Several of the comments on Scott’s post focused on the notion that keyboards will one day be obsolete, and that speech recognition will break the typing bottleneck. But that’s not the real bottleneck. The keystroke conservation principle is just one way of getting at the notion of scalable communication powered by network effects.

One of my favorite stories comes from Larry Moore, who was a Lotus executive. To illustrate why people didn’t “get” Lotus Notes, he used to talk about the early days of the telephone business, when there were roadshows to introduce people to the concept of telephony. Demonstrators would set up two phones on either end of a stage, with a wire strung between, and talk to each other. But it made no sense to the audiences. Obviously those people could already hear each other! Who needed the wire?

It’s the same thing with the principle of keystroke conservation. If I talk to one person, or a few people, faster than I can type messages to one or a few, I can communicate more, but not orders of magnitude more, and not in ways that fully exploit the power of the network.

Forget keystrokes for a moment and look at how Sal Khan is rewiring math and science education. He started out doing one-on-one tutoring with his cousin Nadia. It’s clearly ridiculous to say that his ability to scale that effort is constrained by the rate at which he can talk. On his instructional videos he talks no faster than normal. But he has strategically placed those videos in a pub/sub network where they can be discovered, subscribed to, shared, and reused. There are nearly 60,000 subscribers to his YouTube channel. That’s scalable communication.

The problem with examples like this one, of course, is that most of us aren’t rock-star performers like Sal Khan. If we push all the communication that we can into open networks, we’re not going to boost our reach by five orders of magnitude. Maybe only two. Maybe even just one. But that’s significant! You’ll never type a message 10x faster, or speak it 10x faster. But you can easily reach 10x more people by adopting communication habits that make it more likely that your message will be discovered, shared, and reused.

Face-to-face discussion, phone calls, email, and text messages are narrowcasting modes that don’t scale in this way. Blogs, Twitter, Facebook, wikis, and audio or video podcasts are broadcasting modes that do. How do we use both together in the right ways for given situations? It’s subtle. One commenter on Scott’s post writes:

My emails very rarely contain anything to blog about or update a wiki with.

What amount of email do you think is actually appropriate to becoming a blog entry in your life or in a less technical person’s life?

For what it’s worth, I think in terms of an inventory of reusable parts and the DRY (don’t repeat yourself) principle. For example, I’m often asked about how to publish iCalendar feeds from popular calendar apps. So I’ve written up a series of how-to blog posts. And I’ve encapsulated that series into a query: http://delicious.com/judell/icalpub+howto. None of those posts would have been email messages. But there are many email messages in my outbox that contain links to the series. Because the link is a query, it yields fresh results for anyone who has ever received the link in email as well as for anyone who ever will. The same posts are also quite often found directly by way of search.

Counting keystrokes is just one way to think about the underlying pattern. It’s not about typing versus talking. It’s about choosing the mix of modes that will best repay the effort you invest in communication.

Wakened this morning, about three o’clock, by Mr. Griffin with a letter from Sir W. Coventry to W. Pen…

So begins today’s installment of The Diary of Samuel Pepys, as rendered by Phil Gyford. It’s a remarkable project that maps January 1, 1660 (the start of Pepys’ famous diary) to January 1, 2003 (the start of Phil’s Moveable Type recreation of the diary) and has continued faithfully ever since.

The Pepys blog is enhanced in all sorts of useful ways. People, places, and topics are cross-linked with indexes, places are mapped, all references are viewable on a timeline — it’s a brilliant example of advanced blog customization.

Back in 2003 I mused about what kind of content management system would enable somebody to do a project like this without a lot of inspired hacking. The question came up again recently when my sister Ruth decided to recreate an archive of letters that my parents wrote home from our 15-month stay in New Delhi during 1961 and 1962.

I’ve long held that blog publishing systems are really lightweight content management systems that can be used for almost any purpose. So I pointed her to WordPress.com, explained that you can use pages instead of posts to arrange items however you like, and waited to see what would happen.

Well, it didn’t work. It’s true that you can build an arbitrary collection of pages, but there’s no way Ruth would be able to manage that collection without automation. I could write code to help her, but I don’t want to. That’s partly laziness, and partly curiosity about how to use the standard kit to achieve the desired effects.

One of the biggest limitations of pages, in WordPress, is something I’d never noticed until now: No tags! So ended my plan to have Ruth use tags on pages to achieve a lightweight version of Phil Gyford’s indexes.

Why not just use posts? Originally I thought it would be cool to mimic the Pepys diary: start with a date in 1961, and continue in “real” time. But Ruth doesn’t want to do it that way. She wants to be able to process the archive in any order that’s convenient. And she wants it to read forward, like a book of letters, not backward like a blog. These perfectly reasonable requirements turn out to be harder to satisfy than you’d think.

It turns out that you can make the letters run forward on the Posts page by manipulating the publication dates. So here was the scheme I tried first:

July 2 1961 -> Jan 01 1961 15:01
July 4 1961 -> Jan 01 1961 15:00
...
Oct 19 1961  -> Jan 01 1961 04:01
Oct 22 1961  -> Jan 01 1961 04:00

In this scheme, every letter maps to the same day, chosen arbitrarily as Jan 1, 1961. Every month maps to an hour of that day, each letter maps to a minute within that hour, and the times run backward. Since WordPress reverses the sequence again when displaying items on the Posts page, that makes time run forward in that view.

The benefits are huge. Now Ruth can use tags to organize sets of letters, imposing as much or as little structure as she wants. Views by tag are neatly presented as sets of blurbs with “Continue reading” links. Each item automatically links to its predecessor and successor.

But there’s irreducible weirdness too. For example, the Jan 01 1961 date — which has now become an abstract database key used only for sorting — is part of every post URL. You wind up with patterns like this:

/1961/01/01/june-30-1961-from-anita/

This gets even weirder because dates prior to the start of Unix time — Jan 1, 1970 — don’t display in the management UI. However that turns out to be both a feature and a bug. It’s a feature because WordPress reverts to the current date for display, so you see “Posted on June 28, 2010 by Ruth” instead of “Posted on January 1, 1961 by Ruth.” And it’s a bug because you can’t easily scan and adjust the dates that control sorting.

More weirdness arises from the deeply hardwired assumption — in WordPress, but also in all blogs, really — that entries post in reverse chronological order. Although the backwards time mapping seemed at first glance to work, it turned out to be broken in two ways. On the Posts page, after the break, the link pointed to “Older entries” which were really, in our scheme, “Newer entries.” And within posts, the next and previous logic was also reversed.

So for now I’ve gone back to a forward mapping of hours and minutes within Jan 1, 1961. I’ve ditched the default Posts page in favor of a hand-crafted page that presents items in ascending order. Once you’re in an item, the next and previous links work as expected because, when you move from item to item, WordPress uses a forward arrow of time.

I’m not complaining. It’s astonishing that WordPress provides a free service that Ruth can use publish this archive of letters, and I’m hugely grateful. I think we’ll be able to come up with a technique that will satisfy her requirements — without demanding heroic effort from her or custom software from me. But it sure is interesting to see what happens when you mess with a blog’s notion of the direction of time.

In Defensive surveillance for cyclists I made a LazyWeb request for a helmet-mounted camera that can strongly identify passing vehicles. John Faughnan isn’t in a position to satisfy that request, so he did the next best thing: he refined the specification.

That makes at least two customers. How many more, I wonder? If only there were a way to make the demand visible.

Yesterday the Flickr blog announced that the Keene Public Library has joined the Flickr commons. I’ve been watching the library’s photo stream for a few years now, as its archive of historical photos and postcards and has been steadily and carefully uploaded, described, cataloged, and tagged.

Last weekend I went rock climbing with some friends at the Stone Arch Bridge and wondered what it looked like when the trains ran. Here’s a postcard that answers the question.

Postcard of the Cheshire Railroad Bridge (Stone Arch Bridge) in Keene New Hampshire, also called the Keystone Arch Bridge. The bridge had a 90 foot span and was 60 feet high at the center of the arch.

“An enduring example of the excellence which characterized the construction of the Cheshire Railroad. It was designed by Lucian Tilton. The stone came from a quarry on the Thompson farm, within a half-mile of the bridge. The keystone was set Dec.9, 1846. The removal of the original parapet and the substitution of an iron railing has detracted somewhat from its beauty.”

This postcard says “Largest Stone Arch in New England.”

Nice!

I’m just wrapped up a screencast about the elmcity project. It’ll stand in for me at an upcoming event I can’t attend, and serve as an explanation I can point others too. This is the first screencast I’ve worked on in ages, and also the first in which I appear as a picture-in-picture talking head. The process has been challenging, and I want to write about it while the details are fresh.

Software teleprompters

After writing the script, I realized I’d need a teleprompter in order to read it effectively into the camera. You’d expect to find lots of software prompters floating around on the web, including some free ones, and you’d be right about that. But I had to work through a bunch of them before I found one that worked well for me. I tried CuePrompter, TeleKast, and many others. All failed in some dimension of control: margins, speed, transport. Finally I settled on PromptDog, which is free to try but is the one I’ll buy when I go this route again. It does everything well, but what really put it over the top for me was the way it wires scroll speed to the mouse wheel.

If you’ve read from a software-based teleprompter before, you’ll already know this, but it was new to me. No matter what scroll speed you choose, you need to vary it as you go along. That’s because words and sentences take varying amounts of time to speak, but you need to keep your eyes focused near the top of the screen where the camera sits. With most of the programs I tried, you manage this focal zone by stopping and starting the scroll. But for me, at least, the stops and starts were distracting. PromptDog’s mouse-wheel-driven variable speed control made it much easier to stay in the focal zone. Reading from a software teleprompter is hard, at least for me. I was happy for all the help I could get.

Picture-in-picture video

For this screencast, I upgraded from Camtasia 5 to Camtasia 7. It can record directly from a camcorder, but my second-hand Panasonic PV-GS400 doesn’t seem to work well in that mode. So I recorded to tape, imported the results to a file, and imported that file into Camtasia as a PIP (picture-in-picture) video. On import you tell Camtasia how big your PIP window will be, and where it will show up in the larger video window. I made the PIP window a quarter the size of, vertically centered in, and flush right with the larger 1024×768 video window.

I’d sssumed you could move the PIP window around, and grow it or shrink it, to accomodate different kinds of underlying screencast action. But that assumption was wrong. For a given segment of PIP video, the window stays where you put it. This leads to my first feature request for Camtasia 8: a PIP preview rectangle when recording the screen.

Often it’s OK to let the PIP video just overlay the screen action. But sometimes you don’t want it to hide an essential part of the screen. To avoid that, you have to compose the screen around the PIP window. Lacking a visual cue for the PIP window’s borders, I had to guess. Often I guessed wrong, and had to recompose and reshoot a piece of screen action.

Note that you can vary the size and location of the PIP window by splitting the PIP video into segments and assigning different sizes and locations to each segment. That’s a lot of work, though. And you don’t really want to split the PIP video into segments because then you can’t manipulate the whole track.

Editing audio, motion video, and screen video all together

I made things hard on myself because I’d forgotten that Camtasia invites you to do more integrated editing than you should. In principle you can, for example, run a noise-reduction pass on your audio in Camtasia. In practice, I would prefer to do that in Adobe Audition, which does the job faster and better. What I should have done is grab the sound track out of the captured motion video, run Audition’s noise reducer, recombine the audio and video, and then import into Camtasia for editing.

Instead I edited everything down in Camtasia, then tried to do an export/process/import pass on the audio. When you export, Camtasia renders the audio based on your edits. Unfortunately it came out a few seconds longer than expected. I think that’s because the differing frame rates for screen video on the one hand, and motion video plus audio on the other, make it hard to keep things in synch. Next time around I’m going to try matching the frame rates to see if that helps.

(In the end I decided it was worth redoing the edit anyway, so I split the AVI file I’d recorded from the camcorder, fixed the audio, imported it back into Camtasia, and redid the relatively few edits I’d made to the PIP video.)

In the past, I’ve done some carefully edited screencasts where things that I say are tightly synched to things happening on screen. (“…when I click on this link, we see that …) It’s easy to pull that off when you can’t see the speaker, because you can mess with the screen video, or the audio, or both. When you can see the speaker, it’s much harder. Motion video isn’t nearly so forgiving as audio, so you have to do almost all the synch adjustment in the screen video. Or else re-record some or all of the motion video.

To PIP or not to PIP?

Is all this effort worth the trouble? When Scott Hanselman surveyed his readers about screencasts, he asked, among other things, “PIP or no PIP?” More than half agreed with the statement: “Too much PIP (Picture in Picture) video of the presenter is distracting.” And I think that’s true for screencasts that show how to do stuff with software.

When a screencast shows why to do stuff with software, though, I think the talking head may make more sense. Now, my instinct is to be a voice only, as I am on my podcast. But if the screencast is going to represent me at an event, it seems like I should try to project myself there.

More broadly, the topic is something I care about and have struggled to communicate effectively. If this method of presentation works better than others I’ve tried, even if only for some people, then it’s worth doing. My communication kit needs as many tools as I can pack into it.

Now that I’ve knocked the rust off my screencasting skills, I’m looking forward to redoing this video based on feedback. And since it was made for a ten-minute conference slot, I should probably also do some shorter versions that will work in different contexts.

One thing that’s becoming terribly clear: If I want these to make sense to broad audiences, I need to speak plainly and illustrate with simple everyday examples. I’ve been embarrassingly slow to figure that out, but I am learning. In the screencast I just wrapped, which is all about syndication, I never use that word. It’s a start!

This could have been me:

A bicyclist riding along Old Homestead Highway was hit by a vehicle Friday evening.

At about 6:43 p.m. Swanzey Police and Fire Department responded to a reported hit-and-run accident on Route 32.

The vehicle was described as a white SUV, possibly a Chevy Blazer, with a black roof rack. It’s missing its passenger-side mirror as a result of the accident, according to Cpl. Robert Eccleston of the Swanzey police.

The cyclist suffered serious injuries and was transported to Cheshire Medical Center/Dartmouth-Hitchcock Keene.

A couple of years ago it was me. I got sideswiped on a bike ride in another part of the county. In that case too, the impact also broke off the passenger-side mirror. Luckily I only suffered a bruised leg. According to a follow-up report, this cyclist suffered “skull fractures on the left side of his head, where his helmet hit the pavement, a broken shoulder and severe road rash.”

When it happened to me, I was furious for weeks. Every time I saw a sedan similar to the one that knocked me off my bike I looked for the telltale missing passenger-side mirror. And I formed a clear idea of a product that might have prevented the hit-and-run, or failing that, nabbed the perpetrator. It’s a pair of bicycle-mounted cameras, front and rear, that trigger on approaching traffic and take sequences of shots that can identify approaching vehicles.

Here’s why I imagine this could work. I don’t know about yesterday’s hit-and-run, but in my case it didn’t feel like an accident. We were the only two vehicles on the road. There was plenty of room for the car to give me wide berth. But some motorists like to hassle cyclists verbally, and once in a while that escalates to a cat-and-mouse game. That’s a game people these people play because they think they can get away with it. There’s no expectation that the sideswiped cyclist will be able to prove that it happened, or capture the identity of the car. In my case, when I jumped to my feet after tumbling along the roadside, only to see the car speeding over the top of the next hill, I remember thinking: “You bastard, if I only had your license plate number you would regret this.”

Defensive surveillance isn’t just a capability that cyclists need, of course. It makes sense for motorists to identify and record oncoming traffic too. But car-on-car violence is a game played on a level field. Car-on-bike violence is so unequal that I’ll jump at any advantage I can get.

Does the product I imagine already exist? Maybe, but I don’t think so. There are obviously scads of cheap helmet- or bike-mountable cameras. What I’m looking for, though, is one that’s optimized for defensive surveillance. I think that means a gadget that senses oncoming traffic, and then shoots sequences of high-resolution stills. Ideally it’d come with two pairs of mounts. One pair would be fitted to my bike’s handlebar and seat. The other pair would be fitted to my car’s dashboard and rear deck. For extra credit, the car would keep the cameras charged so they’re always ready to defend the bike.

PS: Meanwhile, my low-tech solution is a helmet-mounted rear view mirror. I have always used one, and can now scarcely imagine what it used to be like to have to crane my head around — and wobble my bike — in order to see what’s behind me. With a helmet mirror, situational awareness only requires rapid eye flicks that become an automatic habit. Obviously the habit wasn’t fully automatic, but after the incident a couple of years ago I’m even more vigilant. I watch every car that approaches from the rear, and am always mentally preparing a dive into the ditch.

Noting that Windows 7 has been shipping with multi-touch support since October 2009, Charles Fitzgerald recently asked: Where are the Windows 7 tablets? Well, I’ve got one. It’s the Acer Aspire 1420P, which is same the machine that Microsoft PDC attendees got last fall. The moniker is “convertible tablet PC” but for me, it’s really a “do-everything PC” because I use it in three modes: as a desktop, laptop, and tablet.

In tablet mode it’s no iPad, I’ll be the first to admit. But as a general-purpose machine that morphs into a tablet, it has exceeded my expectations. Conventional wisdom holds that Windows 7 running standard apps can’t make effective use of a multi-touch screen. But while standard apps clearly aren’t optimized for multi-touch, basic gestures work and are very useful. Tapping and scrolling are my staple gestures, but I was delighted to find that pinching and spreading map to font size adjustment in browser windows. I do this all the time now.

My primary use for tablet mode is reading — mostly reading web pages. Before I got this machine, I had already developed the habit of loading up a bunch of pages into browser tabs, using Readability to discard the cruft, and then kicking back on a sofa, or in an airplane seat, to cycle through the tabs. Now I can do this with the Acer in tablet mode. At 1.7kg (3.8 pounds) it’s not something I can conveniently hold for a long time without propping up with my legs or with a pillow. But to put things in perspective, Wolfram Alpha reports that 1.7 kg is 0.68 x the mass of the book A New Kind of Science.

Full-on tablet mode is just one option for reading, though. The other day, sitting in an airport bar reading, I used the machine in laptop mode but with the screen spun around so that the keyboard was safely away from my drink. Later, in a meeting with colleagues, I spun the screen to a variety of angles to show things to them, and to enable them to show things to one another. Now that I’ve had a taste of this kind of flexibility, I’ll never want another laptop that doesn’t have a screen you can spin around and fold back.

As a pure laptop it’s a bit of a compromise, as you’d expect. The keyboard is solid, but the screen outweighs the body of the machine which can make it tippy. The screen is also wider and skinnier than I’d like. That said, multi-touch makes it a different kind of laptop than I’ve ever used before. Now, when using other machines, I find myself reaching for the screen to scroll or adjust fonts. It’s true that general-purpose computers aren’t optimized for touch. But it’s an incredibly useful adjunct. I won’t ever want another computer that doesn’t support touch.

With previous laptops I’ve always used docking stations. For the Acer, though, I just plug in a giant second monitor. And I use a USB keyboard/mouse adapter to command the machine from my Captain Kirk chair.

Now I’m really looking forward to a next-gen version of this do-everything computer. It would be a bit squarer. It would be a bit lighter — say, .4 x the mass of A New Kind of Science. The accelerometer and multi-touch display would be more responsive. Given all that, though, I’m not sure I’d ever need or want a slate-style tablet. This machine has raised my expectations for just how flexible an all-purpose computer can be.

In my town there’s one guy who does shoe and leather repair: Ed Hutchins. He resoles my Birkenstocks, fixes kids’ hockey skate boots, refurbishes leather jackets. If you search the web, you’ll find two reviews of Ed’s work:

1. They do a great job repairing shoes here and just recently he did a fantastic job repairing my son’s hockey skate boot. His rates are very reasonable the only downfall is he takes his time so don’t be in a hurry to get your item back soon.

2. Ed Hutchins is a very nice man who does very good work. The only complaint I can think of is that he is not the fastest at getting work done. He does a fabulous job though and his repairs last!

Those comments appear on a number of sites: Kuzdu, InsiderPages, LocalTom. Beyond address and phone number, that’s all the web knows about Ed’s business. The comments are accurate. I’ve been waiting months for my Birkenstocks. It’s clear to me that there’s unmet demand here for shoe and leather repair. But you’re somebody who could help Ed meet that demand, how could you know about the opportunity?

In principle, the demand can be made visible. I think it was Esther Dyson who coined the phrase visible demand. In 2006 she published an issue of Release 1.0 entitled Visible Demand: The New Air-Taxi Market. The idea, which I discussed at length with DayJet’s Ed Iacobucci, is that when we use the web to aggregate demand — in this case, for direct flights among regional airports — we can optimize the delivery of services.

The same idea shows up in Eventful Demand:

1. Demand that your favorite performers come to your town.

2. Spread the word to get your friends and family to join the Demand.

3. Eventful will alert you when your events are scheduled.

DayJet was up and running until the 2008 credit crunch killed it. Eventful Demand is alive although not really kicking. (It’s unlikely that you’ve ever Demanded a performer. And I suspect it’s also unlikely that you’ve ever attended a performance at a venue chosen to satisfy a Demand.) But the idea of visible demand seems so powerful, and so right, that I hope it will play out on a broader stage.

How? That’s a $64 billion question which I hope people smarter than me are trying to answer. Meanwhile, I’ll just put this fact out onto the web. If you’re a great shoe and leather repairer, and you’d like to ply your trade in a picturesque New England town, the folks in Keene will welcome you with open arms.

Last week Peter Wayner wrote in the NY Times about the Canon Hack Development Kit, which makes it possible to write scripts to control Canon PowerShot cameras. The article describes how hobbyists fly this kit on weather balloons to perform high-altitude surveillance.

It can also be used for low-altitude surveillance. Last week I moderated a panel at Gov 2.0. One of the panelists, John Crowley, showed ultra-high-resolution (9cm/pixel!) imagery of the Gulf oil spill that was taken from a kite carrying one of these hacked Canon cameras. This isn’t just way faster and way cheaper imagery than we’ve seen from official sources. It’s also way better.

John Crowley doesn’t regard this as a hobby. Working for the Harvard Humanitarian Initiative and STAR-TIDES, he researches and develops emergency infrastructure for stressed populations. That means shelter, water, power, and sanitation, but also information and communication technologies (ICT). Kite surveillance of the Gulf was one of his compelling ICT examples. Another was an OpenStreetMap project that collaboratively mapped out the affecteed areas of Haiti in the days following the quake. I don’t yet have links for these examples but I’m going to ask @jcrowley to provide them, and I hope he’ll join me for an Innovators podcast.