We’ve just completed our first internal hack day and made Twitter-powered post-its for the web. It’s not earth shattering, but it was quite fun.

The idea

Exploring what could be better than a list of comments at the bottom of an article

We came up with the idea at a Mozilla/Guardian ‘jam’ exploring how technology will affect journalism in the next few years. One of the challenges was entitled ‘beyond comment threads’ and was aimed at exploring what could be better than just having a list of comments at the bottom of an article.

One improvement, we thought, would be to allow inline commenting on an article so that people could share their thoughts and have a bit of a debate with a bit of context.

The Sketch

We came up with this artwork to convey the idea:

It’s not going to win any prizes, but enough to give the general idea.

We thought the best way to power such a tool would be through Twitter since it’s a) very popular and trendy, b) easy to use, c) would allow comments to be shared across sites d) could be made quickly without the world’s major sites sharing our vision.

It seemed such a simple idea that we thought that it probably existed already, but Mozilla’s @cyberdees liked the idea and thought it was original so we developed it a bit further.

Tam (the most dedicated Twitterer at Outlandish Ideas) was helpfully sceptical about the idea asking the much needed questions a) why is it better than any other link shortening service, b) how can it make money, c) how would potential users find it, d) when would you use it.

@rasmuswinter and @harryrobbins came up with variously credible ideas and a 30 minute static prototype that was enough to convince @tamlyn that it was worth spending a day on, if not exactly that it was a good idea.

The Proof of Concept

At Outlandish Ideas we’re very keen on making things quickly and making them iteratively. In other words, if an idea seems like it might be worthwhile spend 2 minutes making  a sketch. If the sketch goes down well spend 30 minutes sketching it with code (e.g. the simplest form of the idea). If that all still make sense then spend a bit more time on it, until the ideas is either saleable or exhausted.

In this case we took it a bit further than proof of concept and made a prototype that works.

The Prototype

We spent a day (two half days to be honest) hacking a working version together.

Rasmus created the database backend and API, Tam used javascript wizardry to create a neat little bookmarklet that allow you to add notes to page, and Harry created the scripts that allow you to view a page with its Twiki.

The prototype is still very much a quick and dirty hack, but it does work (mostly) and demonstrates the idea. You can try it out at http://twiki.co or see some examples here and here.

You can also see all the notes that have been attached to a page – for example the BBC Homepage.

Outcomes

We enjoyed ourselves and learned a lot.

We learned to check the Intellectual Property Registry for trademarks before you chose a name for a product (we received a cease-and-desist notice from the nice people at twiki.org within about 3 hours of ‘launching’).

We learned that we can make a quite well-polished prototype in a day.

We learned that Twitter’s search is very inconsistent and can stop indexing tweets or even entire users if it thinks what they have to say is pointless.

We learned that (again) Tam is almost always right.