Blue-Sky thinking
This week, BSkyB’s COO, Mark Darcy, penned an opinion piece in the “Organ Grinder” section of the MediaGuardian, on the subject of Project Canvas.
This piece was particularly interesting, because as far as I know it’s the first whisper on the subject we’ve heard from Sky since the last consultation closed (although I don’t think they’ve disclosed their response to it as they have previously).
It was doubly-interesting, because Sky has finally made their ulterior motives clear, where previously all we could be sure of was the fact that they had some.
Regular readers will know that my position on Canvas has been aligned with Sky’s to an extent in the past, based upon the representations that they’ve made public. This has nothing to do with any great love for Sky, and more to do with the fact that some of their arguments were fairly sound. The danger has always been, and the comments on the Organ Grinder post bear this out, that the name “Sky” diminishes the value of the arguments that they’re making, especially when accusations of hypocrisy are trivially levelled. In all honesty, I wish Sky had kept quiet, as they’ve done no favours at all to those of us who have genuine concerns with respect to Canvas.
Let’s start with things we agree about, as they’re fairly straightforward. Top of the list is the structure of the whole affair. My position is that, as is the case with over-the-air broadcasts, if Canvas wants to deliver something worthwhile for the future, it should be limited to a set of specifications which cover the technology and required features (with compliance required for legitimate use of the branding, as is the case with other suites of standards). Broadcasters and manufacturers then commit to adhering to the standards and things work more or less swimmingly. As the principal standards-collating body based in the UK with membership made up of the broadcasters, consumer electronics manufacturers and other interested parties, the DTG is best placed to do the legwork here. In fact, much of Canvas is supposed to be based upon the DTG’s work anyway.
Most of the rest boils down to metadata discovery and management. The Canvas approach is collective control amongst its members, but in a somewhat cack-handed manner enforced by way of its unified interface. My suggestion would be that Canvas would be well served by taking a good long hard look at RadioDNS’s approach (I outlined discovery mechanisms in my response to the third consultation).
Now, it’s certainly true that Sky is guilty of hypocrisy when it bitterly complains about the mandated UX given that Sky forces all Digibox manufacturers to adhere to its mandated user interface. That rebuttal alone doesn’t mean the mandated UX is actually a good idea, though. If you pretend that the arguments about it come from somebody who isn’t in Sky’s position, they do make sense: there’s no clear benefit (and no justification has been given) for mandating a particular user interface across Canvas devices.
Some have suggested that a mandated user interface is necessary to ensure that features which might otherwise be neglected (especially in the realm of accessibility) are present, but this is a dubious argument at best: if the branding scheme (backed by the Canvas partners) requires these features to be present and useable, then manufacturers won’t have a whole lot of choice but to implement them, just as is the case with Freeview receivers today.
It’s also been suggested that tight control over the EPG is necessary in order to maintain coherence and consistency and avoid horrific user confusion. However, there are two key factors at play. First is that the documents submitted to the BBC Trust state that Canvas’s EPG numbering will align itself with DMOL’s numbering scheme (which is the scheme used by Freeview); the second is that utilising Canvas to deliver linear services (i.e., those actually requiring an EPG entry) exclusively over IP doesn’t really seem to be on the agenda — though it’s difficult to tell from the representations to the Trust, and this would be a grave error if this is the case, as it would severely limit the potential of Canvas.
However, if either of these is the case, the need for the Canvas Joint Venture to have any involvement in the EPG is removed: it’s either DMOL, or it’s whatever-platform-linear-services-use (be that Freeview, Freesat, or whatever).
All of the mock-ups and demonstrations of Canvas that I’ve seen show off IP-delivered content augmenting over-the-air linear broadcasts, whether it’s on-demand content, interactive features, or additional linear streams (e.g., ‘multi-screen’ features). None of these things requires Canvas control over an EPG. In fact, none of these things requires anything more than a set of common standards, metadata and service-discovery, and a connection between my Canvas device and the broadcaster’s servers.
Next, there’s the rather strange and ill-defined partnerships with ISPs. It’s not remotely clear what the plan is here, especially given that Ofcom is very big on people being able to switch ISPs with as little hassle as possible. The only possible scenario that I can see based on the available information is that ISPs sell Canvas “services” as part of their existing packages. Thus, alongside your phone-line and broadband, you also get Canvas “stuff” delivered to you, which may or may not include some kind of value-add on-demand services which you can only get from them. This raises a few questions, though. Devices on my network can talk to the BBC’s servers already, as can yours, so is there any value-add if I don’t care about the ISP’s extra services? If there is, what does this mean for ’net neutrality? Certainly, BT Wholesale’s Content Connect, or some variant of it, is part of BT’s plans for Canvas, and that definitely doesn’t bode well for ’net neutrality.
Similarly, what happens when I switch ISPs? Is the Canvas box mine in the same way that my Freeview box is and do I just lose the extra services, or is it rendered useless when my contract with my ISP is terminated? Will the ISP I switch to insist on sending me another Canvas box (which could be better or worse than the one I’ve got, but will of course have the same mandated UX), leading to old ones getting dumped on a regular basis? What are the WEEE implications for all of this?
There are other things which have been left deliberately unspecified at this point: conditional access and DRM. That’s fine, though, because from a technical perspective none of them are particularly difficult to do. First-draft specifications for both over IP can be turned around in a week if you know what you’re talking about (and the first step there is knowing the difference between CA — which encompasses geographic restrictions — and DRM). The fact that DRM is by definition flawed in most contexts, especially those involving communication over the Internet, doesn’t change the fact that it’s conceptually pretty simple stuff (even more-so if you’re designing, rather than reverse-engineering it).
That’s what Sky and I agree on. Not just Sky and me, though — Virgin Media, the DTG, and a collection of others seem to share common ground on much of the above.
As I said at the start of this piece, though, Sky’s ulterior motives have now become clear. As thephazer points out in his comments on the piece, Sky wants to have its cake and eat it, too.
For the avoidance of doubt, my envisaged future for TV is one where broadcasters — whether they’re exclusively over-the-air, exclusively IP, or a mixture of both — can make their content available safe in the knowledge that people will be able to receive it without any more difficulty than they can with Freeview today, and with a superset of the features available. People like Sky don’t actually figure in this equation except as yet-another-broadcaster. Sure, they might apply a conditional access scheme such that their content is accessible to Sky subscribers, or they might mandate conformance with the specifications as part of the agreements they have with Digibox manufacturers, such that IP-delivered services can augment the linear broadcasts being carried today as part of Sky’s packages. None of this magically alters — nor should it — the relationship Sky has with the BBC, ITV, Channel 4 or anybody else. The same applies to Virgin, BT Vision, Tiscali TV, Freewire TV, and anybody else carrying linear broadcasts. All should be about is making sure that off-the-shelf device x can pick up content y associated with channel z. This is not conceptually different from buying a TV which had Teletext support back in the day: some TVs had it, others didn’t (yet), and broadcasters didn’t need to have special relationships with anybody in order to deliver it.
What Sky actually wants, aside from the legitimate concerns about Canvas, is for it to enable it to carry the BBC’s content piecemeal. That is, rather than carrying BBC One and allowing Canvas-compliant services to augment those broadcasts (and supporting those services in its Digiboxes), Sky would pick and choose which parts of the BBC’s output it wanted to syndicate and just carry those.
This is completely bonkers. It’s essentially the Chewbacca defence applied to toys-out-of-pram demands. It’s not remotely within the Canvas remit in the first place. Honestly, I have no idea what Sky thinks it’s playing at. I only hope that Sky included this argument in its submission to the BBC Trust so that they could have a good laugh about it.
Of course, the big worry is that an incoming Government will think Sky’s idea is an awfully good wheeze and strong-arm the BBC into permitting it in the face of all reason.
I still think Canvas is a really poor implementation of a great set of ideas, though.
