Content From The Edge

I attended the JXTA User Group meeting last week, and got a chance to hear about a really cool project called Paper Airplane. And to view some truly spectacular UI mockups while I was at it.

The project, headed by Brad Neuberg, is developing a user-friendly tool to allow people to publish content from the edge of the Internet. In its ideal form, Paper Airplane would incorporate distributed storage, relieving users from the need to run and maintain a web server or pay for bandwidth. It’s a truly revolutionary idea – if most of the knowledge is contained at the edge of the network, what better way to release that information and encourage innovation than to lower the technological barrier to sharing information? Paper Airplane was conceived with this purpose in mind: making it easy for people to create and share information.

That said, the ideal solution and the project’s current incarnation are quite different. Although the software will still achieve its primary purpose of allowing easy publishing, the more difficult elements of the implementation have been pared down. The lack of one of the most useful features of the original design, distributed network storage, means that end users will still need an “always on” connection to the net to allow their peer to serve content to other users.

In an ideal world, Paper Airplane would implement all of its original designs, plus more. For example, I’d really like to see this project try to provide a solution that co-exists more closely with the traditional web infrastructure. I envision a dynamic DNS-P2P bridge which would allow a user to enter a URL in a web browser and have the URL resolve to the IP address of a peer that could handle the request. This not only would allow individuals to publish content without running their own webserver, but also would allow the load for popular web sites to be distributed across their readership. For example, readers of a popular site like Slashdot could mirror the latest content on their local peer, reducing the load on the main website and solving what Neuberg affectionately terms the “tragedy of the dot-commons”.

I also got a chance to present an updated version of an idea I’d previously presented here. I’m hoping to put an updated paper together on the topic in the next couple of weeks, and join Neuberg in his quest to push the boundaries of information distribution to the edge of the Internet.

Paper-Thin Security

Foolish paper-based security solutions have been seriously annoying me over the past couple of days. Two recent examples come to mind: my recent trip to the DMV and my ongoing application for Irish citizenship.

The DMV lived up to its reputation this weekend – a purgatory whose screaming children and close quarters seem to be specifically designed to concentrate psychological distress. Welcome to the DMV, we’re here to serve you. After a mere three hours, we arrived at the counter only to have the DMV employee point out that my middle name did not appear on my newly-issued US Social Security Number card. This was unacceptable, given that my middle name appeared on my other identity document, my passport. “Ever since 9/11…” the DMV employee offered in half-hearted explanation.

Only when the employee’s supervisor came over and signed off on the condition that the information be matched against the computer did my application get the go-ahead. Funny, and here I thought the Orwellian concentration of personal data in government computers was going to be a bad thing. Apparently I shouldn’t have worried, as it doesn’t appear the government is actually using any of the data it’s collecting in any significant way during the normal course of business. Had I not even provided my passport (it wasn’t required, I brought it just in case), the integrity of the DMV process would have hinged entirely on a 3″ by 2.5″ piece of cardboard.

Has anyone considered how ludicrous it is to still be relying on paper for our security? With high-resolution printers and sophisticated imaging technology in the hands of every web designer on the planet, one would hope for something more sophisticated than heavy paper and pretty patterns. But what’s worse than pointless paper-based security? How about badly-designed security solution that are supposed to address the deficiencies of pointless paper-based security systems. Like those of the Irish government.

To be clear, the Irish government and Irish institutions have a weird love-hate relationship when it comes to paper. For Irish institutions, a utility bill is often enough to provide proof of identity. But for Irish immigration, this apparently isn’t enough – no, no, no, they want notarized documents. Sounds appropriate, right? But they also want the original documents to be sent to them as part of the application – if they need the original documents, what’s the point of sending notarized copies of these documents as well?

The devil is in the details – it’s not enough that the documents be notarized by a notary public. No, the notary public has to be personally known to you – meaning that you must have known the notary for a minimum of six months prior. Oh, and you can’t identify yourself to the notary using either a driver’s license or passport – no, they just have to know you, by telepathy or some other unspecified means. According to the Irish government, if you were to introduce yourself to a notary, wait six months, and then present documents for notarization, this would provide a much better proof of identity than just asking a notary (or the consulate) to do their job and verify the authenticity of the documents using one of the many computer systems available.

In the end, it probably doesn’t matter – I doubt they even bother to check the identity of the notary public! After all, they probably just rely on another paper-based security mechanism: the notary’s seal. No one could possibly duplicate embossed paper technology!