So last year the ³ÉÈËÂÛ̳ and Yahoo! ran the stunningly cool Hack Day London. It was a brilliant success, and frankly I’ve been biting at the bit to organise another one. Last week I got the nod from Ashley Highfield that the budget had been cleared and we could run another one!
It’s very early days yet, but so far we know that it’s going to be late June, venue to be decided and we’re looking for other people to work with. Obviously we worked with Yahoo! last year (who’s idea Hack Day is) - and this year we want to work with as many people as possible.
I think the only major change will be that we plan to run some kind of conference on the Friday before the weekend event and that we wont be having a massive band on the Sunday night - rather we’ll be doing something (and this is soooo cool) on the Saturday night.
Oh… one last thing… we’re changing the name this year to avoid any confusion… henceforth it’ll be called Mashed.
More news as the planning progresses.
From the Mailing list,
Personally I'm waiting for the time when we can pause a program and scroll over the items on-screen and it'll tell us what they are and where we can buy them, like when Ed Norton describes his apartment at the beginning of Fight Club.
A new prototype based on Radio 1's now playing data but this time from a ³ÉÈËÂÛ̳ member of staff working in their 10% project. Simon goes into details.
We're working on a new 10% time project over here at FM&T Audio and Music - and we thought we'd give you guys a super sneak preview. There's a few of us involved here, including Yasser Rashid, Cathy Bartlet and Ramon Dodd.Its around visualizing now playing information by pulling in data from across the web.
The plan for this is to eventually build a flash version which is full-screenable to provide a visual companion while listening in the office, or on the web etc. Future data sources we hope to build on include Musicbrainz, Wikipedia, YouTube, song lyrics,Yahoo Music and loads more. At the moment, we've just got as far as last.fm, flickr and the webcam, but its a start!
One of last year's Innovation Lab commissions was a small R&D project
to make a visual filter for news stories. The result so far is a pair
of sliders which filter four news feeds into a single list.
The focus is on the ³ÉÈËÂÛ̳ website's 'Where I Live' section, and the aim
is to allow users to choose a balance of News and Local Features, as
well as varying the geographical mix. The slider may have wider
applicability depending on how useful and intuitive it proves to be.
We'd like your feedback.
Two sliders are used, the first to mix two pairs of feeds into two
lists, and the second to mix the two lists together. Stories are
given a value based on their order within the feed, and as the slider
moves across, low value stories from one list drop off, to be
replaced by high value stories from the new list.
The application is in two parts:
* a Perl script which consumes RSS feeds (using XML::FeedPP) and
generates an embedded JSON array containing the stories from all four
sources and their rank.
* a Javascript function which combines and generates the list(s)
visible on the page.
Since the timescales of 'News' and 'Local Features' differ so widely
(news changing by the minute, features dating back weeks) it was not
possible to elegantly rank the lists strictly on their publication
date -- instead we chose to simply score them according to their rank
position in the original feed.
In this prototype the source feeds are hard-coded into the script,
but this could easily be changed. Unfortunately the ³ÉÈËÂÛ̳ does not have
a very RESTful interface to their 'Local Features' XML, and with no
consistency in the URL format there is no simple programatic way to
access all the ³ÉÈËÂÛ̳ regions.
The slider is very basically styled, but is it intuitive? Does the
data shift as you might expect? Are two sliders too complex? Is a
slider appropriate here, or should something else be used? Is the
sorting algorithm right? What should we do about duplicate entries?
We're interested to know what you think.
