On 06/11/2007, David McBride <[EMAIL PROTECTED]> wrote: > ... and then later realised that:
What do you mean "realised later"? These are the kind of things that would be in the feasibility study. 10 minutes with even a half way competent engineer would have shown you that proprietary technology was never going to be portable enough. Short of creating both OS emulators and Processor emulators (Binaries are tied to CPU architecture remember) that run on every target platform then there is no way platform neutrality could be achieved with a proprietary solution. Even then you rule out platforms that need high performance. I was speaking with an engineer from a very large and well respected company and he was saying they implemented a lot of things in Assembler, and utilised specialist hardware just to get performance up and power usage down (they made set top boxes btw.). The BBC has completely locked companies like them out of the market as the only person who knows how Windows DRM works is Microsoft so the BBC has effectively given them an exclusive license to produce iPlayer compatible set top boxes (no one else can achieve the same performance as they have to install WinXP which on a set top box cripples performance.) > * That the arguments for DRM that you'd previously accepted do not make > sound > technological sense, They never made sense, not now and not then. Long before iPlayer was made the concept of reading and writing in the binary machine level was known. Ever heard of punch cards? DRM was as insecure then as it is now. > * The regulatory agency that you report to is indicating that the current > platform support is insufficient Is it a new thing that the BBC can't show commercial bias? Surely that was the case back when iPlayer started. > * That the proprietary technology choices that you'd made for the > distribution > and DRM components of your infrastructure are not portable to > all of the necessary and ideal target platforms (Mac, Linux, smartphones, > iPods, etc.), That is highly obvious now, and should have been obvious at the time (apart from the iPods and smartphones parts). > * You're being forced to publically defend the decisions that you'd > previously > made using the rationale you used at the time, and finding that the > arguments you're making are unconvincing (at least to this audience), Must have been unconvincing at the time. Technology may have advanced but the basics have remained pretty constant. > So far as I see it, Ashley has only a couple of options: > > 1. Try to continue down the present course - procuring or developing DRM > and/or distribution technology as necessary in order to satisfy both the > BBC regulators' and the rights-holders' requirements. > (See also: the recent Adobe Air streaming announcement.) Adobe Air is still not platform neutral though is it? It's heavily dependant on the Adobe platform. As I mentioned above there are companies who need to use hardware and assembler to squeeze out extra performance and cut power consumption (good for stopping gobal warming). The BBC is actively preventing these type of companies from developing products. Without actually releasing a specification they have to use a general purpose CPU that supports instruction sets defined by the BBC (in general x86) which although popular in desktop boxes it is not heavily used in embedded systems. The engineer I was speaking to was more than happy to show off their new set top box. Running Linux on MIPS (it was a prototype, I'm not sure if they intended to keep the OS and chipset). Does Adobe Air run on MIPS? Does Knotiki? Does WMP? What about ARM and Sparc? Again the BBC is locking out very good people purely because they won't build an inefficient system running on a specific CISC processors. > 2. Develop and advocate a major shift in strategy, that involves: > - Dropping the design requirement for DRM on all distributed content, Not really possible, or needed. There are many open standards for DRM. And none are less secure than Microsoft's offering. It was broken before release, it couldn't even provide protection for a few seconds, it was dead on arrival. I find it hard to even design something less broken. > - Retooling the existing production infrastructure as necessary to support > open distribution and content standards, Shouldn't take too long, if the BBC needs a hand I am almost certain I can find some BitTorrent server software. Bit Torrent seems to scale well, and it works in a platform neutral manner. > - Either convincing the BBC legal team that they have the rights to > distribute the programme-makers content sans-DRM under their existing > broadcast / streaming agreements, -or- I don't think you can convince the legal team they have something unless they actually have it. Lawyers are a bit picky about this kind of thing. But as I said in point 1, DRM can be done in an interoperable manner. > - Re-opening negotiations with the programme makers to secure internet > distribution rights sans-DRM, -or- That could be useful, it would help if the BBC knew how DRM and computers actually worked and could explain how trivally easy it is to break them. Google has some examples of broken DRM schemes, here are a few: DVD-CSS (completely dead) AACS (many keys have been recovered, not sure if licensing authority can revoke them quick enough either) Window DRM > - Restricting the programmes that may be downloaded by the iPlayer service > to in-house content that they clearly can offer for access by UK > residents. It's relatively simple to restrict it to a country at the server side using IP to country lookups (might want to see: http://www.maxmind.com/app/country $50 one off payment, $12 a month after first month. 6 quid a month not bad, or there's a free version: http://www.maxmind.com/app/geolitecountry ) (not entirely sure what would happen if you tried to use above data to build IP Tables rules, might break performance, but you could always distribute encrypted files to the world via P2P and send the keys to UK users, would reduce the traffic that needs to be checked for IP. Not hard to do in an open way either, just pick a cypher (e.g. AES) and a way of encoding it, and send via HTTP, job done.) Or a third option: 3. Resign. Then maybe someone who knows what they are doing can take over. Andy -- Computers are like air conditioners. Both stop working, if you open windows. -- Adam Heath - Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial list archive: http://www.mail-archive.com/[email protected]/
