I've been pretty sceptical - and critical - of the BBC's TV over IP efforts, including Project Canvas:
Project Canvas is a proposed partnership between Arqiva, the BBC, BT, C4, Channel Five, ITV and Talk Talk to build an open internet-connected TV platform, subject to BBC Trust approval.
The partners intend to form a venture to promote the platform to consumers and the content, service and developer community.
Like the UK's current free-to-air brands Freeview and Freesat - a consumer brand (not canvas) will be created, and licensed to device manufacturers, and internet service providers owners who meet the specifications.
‘Canvas compliant’ devices (eg set-top boxes), built to a common technical standard, would provide seamless access to a range of third-party services through a common, simple, user experience.
That's despite - or maybe even *because* - it proclaims itself as "open":
A technology project to build an open, internet-connected TV platform
As well as a lack of standards in the internet-connected TV market, there is no open platform. This creates two main problems:
* The UK's current free to air TV platforms Freeview and Freesat have been unable to evolve and keep pace with technical innovation in the consumer electronics industry. While some internet services are emerging on some commercially-owned/ pay-TV platforms - these platforms are working to their own (proprietary) closed standards. A fragmented market is emerging, which could put internet-connected TV out of the reach of consumers who don't want to subscribe to pay-TV.
* The internet services need to have a commercial relationship with the TV platform to obtain a route to the shared screen. This, combined with a fragmented market of varying standards, is slowing the development of internet-connected TV services.
Project Canvas intends to build, run and promote a platform that solves both problems: providing an upgrade for free-to-air TV, and an open platform of scale that will bring a wide range of internet services to the shared screen.
We all know how debased the term "open" has become, so frankly I expected the worst when the technical details were released. Looks like I was wrong [.pdf]:
Linux has been selected as the Operating System for the Device.
Linux has been ported to run on a large number of silicon products, and is currently supported by the vast majority of hardware and software vendors in the connected television ecosystem. Porting to new hardware is a relatively simple due to the architecture of the kernel and the features that it supports. The Linux environment provides the following functionality as a basis for the development and operation of the Device software:
• Multi-processing.
• Real-time constraints and priority-based scheduling.
• Dynamic memory management.
• A robust security model.
• A mature and full-featured IP stack.
Linux is deployed on millions of PCs and consumer electronics devices, and the skills to develop and optimise for it are common in the industry. In addition, a wide range of open source products have been developed for, or ported to Linux.
It's pretty amazing to read this panegyric to Linux: it shows just how far Linux has come, and how it is taking over the embedded world.
Even though content will be "protected" - from you, the user, that is - which means the platform can't really be regarded as totally open, the Project Canvas designers and managers still deserve kudos for opting for Linux, and for publicly extolling its virtues in this way.
Update: I haven't really made clear why that's a good thing, so here are some thoughts.
Obviously, this is not a pure free software project: it's a walled garden with DRM. But there are still advantages for open source.
For example, assuming this project doesn't crash and burn, I expect it will influence similar moves elsewhere in the world, which may be encouraged to use Linux too. Even if that doesn't happen, its use by Project Canvas will increase the profile of Linux, and also the demand for people who are skilled in this area (thus probably helping to drive up salaries of Linux coders.) More generally, the Linux ecosystem will grow as a result of this choice, even if there are non-free elements higher up the stack. Correspondingly, non-free solutions will lose market share and developer mind-share.
And finally, having Linux at the heart of the Project Canvas project will surely make it easier to root...
Follow me @glynmoody on Twitter or identi.ca.