Weeknotes #3 (26/02/10)
Glen was up in BBC North seeing R&D colleagues up there, showing our work and seeing some of theirs. Tris ran a retrospective on the Signatures project this morning - we don't usually do these but we should. This project went really well so this felt like a good one to start with, not too many recriminations hopefully!
A hefty 67-page draft report arrives from the indie working on Resolver - Chris G and Tristan sit down to read it. They finish the day on a phone conference with the indie and Matthew P to discuss some of the policy and strategy elements. Akua was doing a lot of research into Olympics services to inform our ongoing work, whilst Theo and Chris N are working on some hardware after Theo went on an Arduino workshop last week
We had another team-wide chat about space, not much progress. We have a new room available and need to figure out the best way of using it. Every possible option seems to have too many cons.
Tris etc made some headway architecting Coventry and deciding on what bits we can prioritise. Looks like there's a short-term prototype and some longer-term research that could come out of it. After lunch we take a look at our project planning board, it's looking a little empty one month ahead but that's ok cos we have a planning and strategy session next.
Duncan's started to fill up his microblogging database and is seeing some interesting patterns. Sean discovered that his plan to relay the microblogging stream over the proxy would not be able to handle the volume of data. Duncan read about and installed RabittMQ on a local VM along with Ruby AMQP client implementation, and hooked up with a streaming library, to pop incoming updates onto a Queue. He also added script to strip incoming updates and final script to import into the DB. Glen's microblogging ingest API is done and now running as a service at Watford. I'm a bit worried that we have an Enterprise version and a lightweight version of this work, and they're not quite meeting in the middle.
Sean was happy, as he got his first ejabberd module building (for our work around a prescence system.) Sam had some great progress with P2P-Next: feed search, distributed search and LIMO/NextShare integration. Tris and I met with Jigna, who's working on green tech for BBC R&D. We think there are a number of things we could prototype in this area. Adding them to our list.
We have the final review of the Resolver deliverables, we've sent the indie off with a bit of document restructuring to do. Tris shows Mythology to some colleagues who are working on similar concepts for music and radio. They're doing it for production but they're broadly similar in concept and our work could help them push for some more advanced features. I meet a colleague from Children's R&D to discuss what we might work on with them, then rush over to W12 to prepare for a demo on Thursday. The shuttle bus is late so I sit on the Westway for far too long. Boo.
Vicky and I went to W12 to present recent work to BBC FM&T's COO and her team. We got a really positive reaction. Some of her team will make a return visit to our office to see more work and to follow up Multitouch. Annoyingly, our (custom made by BBC R&D) multitouch table got mangled in transit, so we are having to get that fixed. Hmmm. Duncan and Glen were hard at work on the microblogging service - we've now deployed both app and ingest scripts to our Watford boxes. We are presenting this work on Friday - so Duncan wrote SQL to generate stats for the presentation.
We all go (back in my case) to W12 at the end of the day, for the launch of BBC R&D's new South Lab. Lots of people in the new space. Gilbert and George (the R&D mascots) are safely outside. We all admire the library and the building, have a drink or two, chat to colleagues and join in the celebration for what's clearly been a difficult task.
Presented the microblogging work to Jem from A&M. He liked it and had some suggestions about what to do with it next. We're pleased that this went well. Glen and I have a catchup to discuss how things are going at the moment. Vicky, Duncan and Theo lock themselves in a room to go through next week's work on the microblogging work. No meetings apart from that so we're all just slogging through our backlog and ignoring emails.