BBC iPlayer in the Mesh
As you may have seen in David Treadwell’s piece on stage today at PDC, Anthony Rose, Head of Digital at BBC Future Media and Technology, presented a project we’ve been working on with the BBC over the past couple of months and that has kept me heads down for quite a bit of time.
Hopefully the demo represented some useful ideas for what you might want to construct using the Live Mesh , but I’ll recap on some of the thinking behind this project.
The key benefits from Mesh in the case of iPlayer (which will no doubt be applicable in many cases) could probably be described as:
- Offline capabilities (with Silverlight for instance)
- Automation of transport of media, and personal data across devices (“me sync”)
- Application sharing amongst groups of friends (“we sync”)
The first two points above solve some of the engineering challenges associated with the requirements of a platform such as the iPlayer: they’re important and could be expensive to construct and maintain so getting these things “out of the box” is clearly useful.
The third point – sharing amongst groups of friends – is really about the move to social media rather than lists, or general recommendation engines. Enabling groups to recommend within the group, and see other group members’ activities opens up a number of possibilities around identifying key influencers (and perhaps rewarding them) and also enabling behaviours related to (for instance) peer pressure: “Did you know all of your friends have watched programme X?” For the BBC, this can drive increased reach for their content through the social and viral nature of these calls to action, and for the user it provides a much more personalised and relevant set of information.
So, how did we construct the application?
We used the Microsoft Technology Center here in Reading and worked with the BBC and my own crack team of Microsoft Consulting Services consultants (Paul (who you may have seen on stage), Carl and Shen) to develop the application as a prototype in around a 2-month time frame.
As with most prototyping exercises, there’s a lot of ground that wasn’t covered as the key driver is to ensure the unknowns become knowns. In this case, that meant ensuring the Mesh was capable of supporting the following kinds of scenarios:
- Silverlight applications working with no issues.
- Connecting to 3rd party web services (BBC Syndication feeds, Live Messenger Contacts API, etc.)
- Syncing personal data around chosen devices
- Consistent experience across chosen devices
- Online/Offline experience (which means constructing caching services for images (for example))
- Media transport (moving media into the Mesh, and then to devices for offline viewing)
- Syncing group data across application instances (and capturing that in relevant ways for overall Mesh experience and iPlayer aggregate information)
- Some kind of social feedback and recommendation capability (in this case we constructed a social timeline for moods and a mood barometer).
So this was quite a lot to execute on in the constrained time frame, particularly when you consider that we also needed to build a user experience that reflected the existing iPlayer experience.
On top of this the Live Mesh APIs were being actively developed at the same time?
Carl carried out most of the Mesh-based heavy lifting and is looking to follow up some of the work he did with the Mesh APIs specifically as an article for MSDN, so I’ll point to that once it’s complete.
So, checking off the items on this list proved the application viability. Live Mesh really has the potential to change the way we think about experiences across devices, but also how we think about interacting with each other through application experiences.
As Anthony said, “Tomorrow, your friends will tell you what to watch”.
