³ÉÈËÂÛ̳

We're coming to the end of the current phase of development of The Programme List. This has been an exploration of how people remember programmes and tools to help them do that. We launched the first version in September as a simple way of keeping a list of programmes that you might want to watch or listen to. Since then, based on feedback from our users, we've developed the concept to focus more on delivering reminders and alerts for your programmes. The last prototype update in December added Twitter-delivered alerts - sent 15 minutes before a programme is due to start, with the restriction that these could only be set on individual broadcasts.

This week Andrew's been implementing the GEL re-style of the prototype, including a new settings page for alerts and emails. Dan's been modifying the API underneath to support these changes and finishing off what's necessary for the emails. Theo's been on hand for any UI tweaks and Duncan's been adding some custom analytics so we can better analyse the usage of it. We'll be releasing the next version shortly and it will have recurring reminders that can be set on episodes, series or brands (that's what we call our long-running programmes) and will be delivered by Twitter or by email. We're also starting to put together a video that explains more about the concepts behind this prototype and how we're imagining a common platform with which these lists and alerts might be shared amongst your TVs, radios, computers and mobiles. You can try .

This week on our other projects:

is wrapping up after three years, Libby's tidying off loose ends and Vicky B is storyboarding a 1-minute video to summarise the whole project and created .

ABC-IP is a little quiet this week but Chris L has been evaluating the automatic segmentation code by comparing it to manually added chapter points in the World Service programme The Strand.

In Sean has been writing up the recent work on adding the HTML5 startOffsetTime attribute to Firefox. More about that soon.

In the programme recommendations work it's Becky's last week with us and she has been finishing off her analysis of the field trial. This has shown very good results for a collaborative filtering recommendation algorithm and is able to produce instantaneous recommendations based on a small amount of information provided by the user.

In Vicky and Joanne are planning the diary study that will be run to capture people's viewing activity. This will end with lab sessions where participants are shown their viewing history and asked about their attitudes to privacy and personal data. Pete has been wireframing the prototype dashboard for these sessions while Barbara and Chris N have been dealing with the collaborative side of the project.

And in other work, Sean has been building a a simple HTML5 RadioVIS demonstrator for use at an and Michael has been exploring using as a documentation format. And Marc has been looking at how to use web-based timelines for news stories; from to via !