I spent this weekend at the Guardian offices for the Guardian/Rewired State SXSW Hack Day.
The theme was to look at 21st century tools for journalism, and used covering the SXSW conference as a starting point.
Sorely lacking in any useful coding skills, I think my best input came at the beginning of the weekend when a few developers sought me out to talk through some of the barriers and frustrations journalists face when trying to cover events.
The most interesting of these conversations was with Sym Roe, who came to the event wanting to create something that would have a wider use for journalists beyond reporting SXSW.
We talked over a number of issues journalists face.
One notable one that didn’t make it past the discussion stage was how to filter out noise on Twitter to get to the really interesting stories. This is really needed when you report on an event that has hundreds of social media-savvy people present and (as with SXSW) many sycophants willing to furiously retweet certain folk in the hope they might mention their startup, app or idea.
We decided there might be a very simple tool that could count retweeted links, but then allow you to inspect the results in quartiles. That way you could check for stories below some of the more frenetic retweeting. A hack for another time, perhaps.
However, the idea that got us both really interested was what became known as Fluffbox.
Fluffbox, which was developed by Sym and Premasagar Rose, is designed to let journalists curate from a variety of social sites and file all the interesting stuff they find into one, searchable “box”.
This box then lets you drag and drop the pictures, tweets, links, audio, etc, into a document that renders them all in lovely, clean html.
This is fantastic in two ways: one, it allows us journalists to have one place to store all the little bits and bobs we might want to use for a story. And, two, the clean html also makes the finished document something that can be copied and pasted into pretty much any editorial content management system.
Fluffbox was highly commended, but I personally think Sym and Prem’s work is probably the most practical journalist tool that came out of the Hack Day… but then again, I may be biased!
As if he wasn’t busy enough, Sym also helped me realise a second hack that I’d been wanting to do for ages. This one, however, had absolutely no practical use.
The Romp-o-meter pulls in all the stories that contain the word “romp” that are published in the UK tabloid press. They are aggregated together, the sport-related “romps” are removed, and then the UK is given a “romp” score, based on the amount of naughty nookie appearing that day. In my unwittingly double entendre-filled presentation, I noted that “romps can go up and down” and that this might indicate then general moral (or morale!) levels of our nation.
Sym had less than 40 minutes to pull this hack together, so I was impressed he even had something to demo! There are plans, however, to make the fully-fledged version. I’m really hoping it will involve a romp-o-meter swing-o-meter.
You can also browse some of the hacks on Rewired State’s project page.
Hi Joanna
Re the retweeted link counter, something else that may be handy is to chat the connections between folk who have tweeted a particular link?
For example, here’s a chart of folk who, according to BackType, tweeted a link to this post: http://www.flickr.com/photos/psychemedia/5444891248/
And connections around another of the #gsxsw posts: http://www.flickr.com/photos/psychemedia/5438208045/
Hi Joanna,
Nice to see your write-up.
There were a few aspects of the Fluffbox editor that I neglected to mention in our demo, which may be of interest:
* The article editor lets you write either plain text, HTML or Markdown, which is a simple-to-use format for writing text that gets converted into HTML – for example, you can easily create lists, links and images, without having to use the verbose HTML notation.
* The app automatically saves your work offline, so that you can resume draft articles at a later stage, and write articles on a train or without Internet access.
* Each type of media asset (“fluff”), such as a Flickr image or a tweet, or a generic image, link, video or Audioboo produces its own custom HTML when you drag the asset into the editor.
* The HTML structure for these different assets can be easily modified to be in the preferred format for the newspaper’s CMS – e.g. The Times will use a different HTML structure for quoted tweets than The Guardian, you may want a Flickr image to include the photographer’s name, or not, and so. It’s pretty flexible.
[BTW – your link to the GitHub repository is a bit broken].
Cheers,
Prem
Pingback: links for 2011-02-14 « Sarah Booker
@Prem Thanks for the github link spot! Should be corrected now.
It sounds like Fluffbox is even more customisable than I realised! I really hope you and Sym take it further. I already know _exactly_ what I’d like to use it for. 🙂
@Tony That’s great – thanks! We talked a bit about Truthy on the day too, so that we could visualise how a trend had been disseminated. Sounds like BackType is the way to do that.