I’d like to thank Caleb Clark for recording and posting a video of the talk I gave last month at the Marlboro College Graduate School. I watched it the other night and I think it’s my best explanation of a cluster of things I’ve been thinking about and working toward for a long time. The list includes:
- the local web
- LibraryLookup
- webscale identifiers
- REST
- public data
- loosely-coupled cloud services
- lightweight service composition
- structure and transformation of data
- the elmcity project
- the pub/sub pattern
- feed syndication
- personally authoritative data publishing
- social and decentralized information management
When I look at that list, and realize that I’m always trying to do (and describe) all of these things at the same time — because they’re all deeply intertwingled — I can see why it’s been so hard to tell the story. Apparently, given an hour, I can now tell it reasonably well. But I’ll rarely get that hour. So I also need to condense it into a five-minute Common Craft-style summary. A hard challenge, but a good one!
Nice talk, Jon; a great way to introduce the how and why of mashups using a locally-relevant context!
What you need is a wiki and a pattern language!
Jon, I spent a whole hour and watched this talk without doing any multi-tasking. It was worth the time.
I was struck (not surprised) by the difficulty in dispensing lucidity around pub/sub-syndication systems. I think some of it stems from the fact that the polling or pushing of data is not intrinsically understood by humans. Calling methods is an easier concept to grasp, or just boiling it down to events (ie a feed was updated) likewise.
I’d love to discuss further.
@John: Thanks!
@Bill: A pattern language for sure. That’s exactly what I feel I’m working toward, with the provisional title: Fourth R Principles.
@Dave: There’s a ton going on here that is not understood by (most) humans, because it’s not part of our cognitive development in the real physical world. The list includes:
– indirection and loose coupling
– the properties of globally-unique names
– the nature of structured data
Around the same time as that Marlboro College talk I was in Newport, RI, for the Traction Users Group conference. In a session on Observable Work we talked a lot about what things belong on that list, how we who intuit them make use of them, and what to do about the great majority who don’t intuit them.
http://tractiontug.blip.tv/file/4284588/
http://tractiontug.blip.tv/file/4284439/
I’ve never really had the chance to hang with a group of people who have internalized all these patterns, and to collectively reflect on them. It was a wonderful experience and I need to have more of those!
Is there any chance of an audio mp3 version?
Yep: http://jonudell.net/audio/marlboro-college-oct-2010-the-local-internet.mp3
Viewed your talk. Like you, I find it is hard to describe the circumstances and value of some things in an elevator ride as the investment community likes to require.
I am always intrigued by those who think they have discovered a new thing and they give it a name not previously used and persist in thinking that it is something new and unique. I have been involved in the data processing of publish and subscribe for 50 years. Push technology is just a broadcast feed on a request/response system that has no on going context.
I can help you and Microsoft help your beloved libraries and their suppliers, the copyright holders, not only for books but all the other media also.
Will you engage other than in this form?
Specifically the financial securities industry has been publishing prices for a long time and individuals have been subscribing to them. All of the functionality of the current internet financial securities presentation I was doing in the 1970’s. Much of it I invented.
I have applied and adapted this technology to the issue of digital media piracy in the Internet environment. Your passion for cooperative information dovetails well into this discussion. I urge you to expand your horizons.
Note the new wordpress account to better communicate!
Further I was the first location to publish the output of Businesswire and PR Newswire along side the Dow Jones news ticker (what we called publishing). And eventually many other sources. The newspapers today take this information (subscribe) and that of the Associated Press (AP) and repackage/edit it sometimes by a computer program and publish it as their own.
Hopefully my avatar will appear this time ;-)
Jon, I enjoyed the talk. It was unclear to me what the issue around 59 minutes (Drupal) was. Can you elaborate?
Drupal exports iCalendar. So the Hannah Grimes site, built on Drupal, was able to syndicate its calendar to the Keene hub. Then Hannah Grimes migrated to Constant Contact, not realizing it doesn’t export iCalendar. More on this at http://radar.oreilly.com/2010/11/the-icalendar-chicken-and-egg.html
Don’t be afraid to use simplistic diagrams just because you’re presenting to adults.
Silly fuzzy monster-as-computer image.
Monster can only eat peas.
iCalendar data = plate of peas in front of monster with big smile on his face
HTML/PDF with random date formats = plate of 9 vegetables, one of which is peas, sitting in front of monster who is angrily picking out the peas and sometimes corn by accident.
…
There are a lot of metaphors and analogies to use (for the message goal) that I think you are missing perhaps from being too close to the forest.
Cooking : You don’t hand someone a finished and iced cake when they ask you what is in it. You hand them the data: the *ingredient list* on the recipe.
Etc…
Don’t be afraid to use simplistic diagrams just because you’re presenting to adults.
Oh, absolutely. It’s very clear to me that I need a version of this talk that’s condensed into a snappy five-minute Common Craft-style screencast.