IRFS Weeknotes #110
For a team looking at such ethereal things as synchronisation, metadata or privacy, our life seems to revolve a lot around a most tangible of materials: paper. Post-its and flip charts are amongst our favourite tools of course, but this week our use of paper takes us in a rather unusual direction...
Indeed, as part of our "10% time", Andrew and Chris have been building a small, internet-connected thermal printer. This was been made very easy by the fantastic instructions and architecture provided by .
Once their components had arrived they very quickly put together the controller and uploaded the sketch (the programme that drives the printer and allows data to be sent to it over HTTP). They then spent considerably more time trying to get it to work on the 91Èȱ¬'s network (an infuriating maze of DHCP servers, network proxies and static IP addresses).
After printing printed their first they started work on their first semi-useful application. This application allows you to print out tweets from a particular user, or from a list of favourite tweets. They've been talking to our colleagues over in about an interesting studio application for this, and we'll share more about that (here or on ) when it's ready.
Printing the best of @bbcirfs tweets
Meanwhile, our two main projects have been making great progress, on paper and otherwise.
The Future-Internet (FI) team is still looking into authentication and device pairing. Chris N and Theo have been working out the 'happy path' framework for the next FI phase which focusses on authentication of services, users and devices. From this, Theo tells me: "I'm confident that we'll have a basic design to paper prototype and test with role play:Who wants to play the bookmark? I'll be the telly."
Our everyday tools
Meanwhile, the ABC-IP team has been testing its tagging experiment prototype for usability issues and technical robustness before we release it to the Global Minds panel. The rest of the IRFS team played "test subjects" on the experiment, uncovering a few bugs but generally liking it a lot.
Still, the ABC team decided to simplify the interaction flow to remove some of the usability issues and improve the instructions. Andrew also spent a good chunk of the week fixing bugs. There's a really exotic one in Chrome that only occurs when you're not trying to diagnose it with the debugger. It's a bit of a scenario. Web development, Andrew notes, isn't often like quantum physics. Which I assume means he's enjoying it.
Andrew had actually started his week by seeing Mark Boulton talk at "" about what he called ''. It was a bit about how things age but more about honesty of materials. In product design, that's making sure plastic is like plastic and wood is like wood. (OK, but what about paper?) What does this mean on the web? It's about being honest with the materials that we work with. Mark spoke about his work for redesigning the . Instead of creating swish 3D visualisations of the experiments, they're going to display live data visualisations that the physicists themselves use, with clear explanations for the layperson about what they show.
Speaking of conferences, Yves is in Crete at where he just presented the paper . He also contributed to a panel on Big Data led by BT, and chaired an In-Use track session.
In spite of a somewhat sketchy internet connection, Yves got quite a lot of work done with the team back in London: he released the evaluation dataset we used within our paper on . And he has been spending the beginning of his week working on polishing our speaker segmentation and clustering algorithms and testing them on all archived episodes of From Our Own Correspondent.
Meanwhile, Chris Newell has been investigating whether the LiveStats log processing software could be adapted to generate some real-time, aggregate statistics.
We don't usually mention meetings in our weeknotes, but I received some enthusiastic reports from quite a lot of people in the team: Libby and Chris Newell came out of a workshop on our recommendations work with lots to think about; Sean joined a very interesting meeting to discuss MPEG-DASH with colleagues from around the 91Èȱ¬. Among all the competing proprietary protocols for streaming A/V over HTTP, such as Apple's , Adobe's and Microsoft's , looks like a very promising standard way to provide adaptive bit-rate streaming, multiple channels and synchronised metadata.
And finally, in preparation for the upcoming , Theo and I ran a one hour creative session to gather our ideas and generate some new ones using a technique called 'Breaking the Rules'. It was bedlam... but we managed to generate a lot of energy and enthusiasm!
One service, many devices...
Comments Post your comment