« Previous | Main | Next »

Research on BBC content for GNU/Linux

Post categories:

George Wright George Wright | 16:44 UK time, Thursday, 30 October 2008

This week, we're really happy to be sharing some work we've commissioned to deliver BBC content (mainly radio shows from the BBC Audio & Music team) on demand for users of GNU/Linux.

In RAD, we've wanted to experiment with non-DRMed content for a while - and managed to assemble some content which was available to access in a number of ways.

One of the complex things about putting content on desktops or OSes (apart from the two closed operating systems) is the number of options available. Free software's all about choice - but for a large corporation like the BBC, sometimes choices take time. So, in my team, RAD, we have made initial discussion about our work with a number of vendors, and selected Canonical (who produce the Ubuntu flavour of GNU/Linux) for this piece of work - other work in a similar vein is currently under discussion with other vendors/distros. Of course, this work is also available for people/vendors to port to the two closed operating systems, should they wish to do so.

We then worked with colleagues around the BBC and selected a list of available, updated, current content which we could make available both in and outside the UK, a feed of this content using the URIplay metadata framework, with our partner MetaBroadcast.

We then worked with Canonical and their software partners Collabra to deliver this feed to the Totem media player, and Gstreamer multimedia platform, to display available content in a simple, browsable window using a free software plugin, where the feed is updated using the Internet, when the user starts the player.

totem_parse430.png
Click for full size

Lots of this work involved changes to the underlying infrastructure of Gstreamer, as well as developing the plugin for Totem. We then moved on the adding content to the feed, and helping to optimise the playback experience - eg to download suitable codecs if they were available to the end user. As the service develops, we will start supplying content in several different formats - some of these are totally free and open, some aren't - we've reflected the wishes of content owners here, obviously.

We're happy with this first iteration, and are already seeing patches to the upstream software sources, to allow other OSes and distros to use these improvements. Longer term, we'll be looking to improve the content sources, as well as optimise the UI - to show channel or programme icons, for example. Because this is free software, we can make changes to this - and so can you/other users! The whole stack is free software - from URIplay through to Totem, the media player. Some codecs will involve a download, and in some territories (mainly outside the UK) may be restricted, but the underlying framework is free and open.

metadata430.png
Click for full size

This, quite clearly, is not a competitor to BBC iPlayer. It surfaces a lot of BBC content that is already available, but you won't see Doctor Who or Dragons' Den on there anytime soon. What you will see is an increasing list of content, in both audio and video formats, that we can share with you in an experimental way, allowing us to explore and test new ways of viewing and listening to some of our TV and radio shows. Some of the blog posts already online have mentioned that this involves all BBC content - that's incorrect - it's a limited subset designed for us to explore and evaluate how new platforms might need new distribution systems.

I'll be posting more technical information explaining how URIplay fits into the system, and our other plans for it, later this week, as well as sharing some of our plans for work with other free software partners.

George Wright is Portfolio Manager, Rapid Application Development, Research & Innovation, BBC Future Media & Technology.

Comments

  • Comment number 1.

    That's very positive progress. Good work :)

  • Comment number 2.

    I appreciate the effort, but you are making this too difficult. Simply provide an HTTP link to an MPEG4 or h.264 encoded AVI. This will work on ALL platforms. Not just Linux, but Windows, OS X, *BSD, etc.

    Why on earth are you making this so hard? There's absolutely no reason to require people to install gstreamer just to watch a video. Look at what Archive.org is doing. They do it right.

  • Comment number 3.

    BBC is going about this all wrong. Apparently there is still the "we have to have an application" mindset running this.

    Think about how web browsers work. You don't create your own web browser for people to access BBC. Instead, you produce documents using web standards like HTML and CSS and deliver them with internet standards like HTTP and TCP/IP. You have a wide choice of software to carry out these tasks to produce and deliver these documents, just like users have a choice of web browsers to view them.

    The approach to be taken for non-DRM audio and video content is to use an existing standard. You could use MPEG4 or H.264. If you want to be sure your listeners and viewers are not encumbered by patents, you could use OGG with Vorbis for audio and Theora for video. As long as your produced output conforms to the standards, multiple player programs should work (or their developers will have the incentive to make them work).

    If you want to be sure your product plays well in free open source players, then I would also suggest participating by financial sponsorship of some of these projects.

  • Comment number 4.

    Hatta0, you might read a little bit more a bit GNU/Linux distros and it's multimedia-history. Ignoring the aspects of user experience and content rights, the main technical trouble is about patents on audio and video codecs on the client side. Which doesn't mean that there are no code implementations, more the lack of rights to use those codecs.

    Have a look at http://lwn.net/Articles/178285/.

    I really appreciate the effort of the BBC, in this project, improving the user experience, by simplifying the access to content and on the other side, by developing dirac, a patent free codec, which is solving licence issues. And in general using free software.

    You might have a look at http://www.diracvideo.org/

  • Comment number 5.

    You are wasting time and the license fee payers' money here.

    You should just focus on getting the content available through only open-standards, non-patent restricted formats, and clearly document how developers can access the data.

    You won't need to waste your time developing plugins, as a standard and well thought out content publishing system will mean that developers will just write suitable software themselves. The Canonicals of this world can then just select what they feel is the best software for their distribution.

  • Comment number 6.

    The BBC has always made things too difficult. That's why I've been listening to NPR etc. streams for years now and why my BBC radiophile days are long gone.

    Here's what I wrote in *2005* in response to some Guardian blog post when I could still be bothered to be angry about the whole sorry business:

    “When will the BBC stop "assessing public demand" and "experimenting" (as they call it every time they put up an mp3), and begin to make good use of internet technology? I don't know what they intend to do with their TV output but their record as far as audio is concerned is abysmal: Where are the Radio 3 ogg streams to match Vltava and NRK? Where are the mp3 (or preferably ogg) streams to match what's available from WNYC, KPFA, KCRW and a host of other NPR stations? How is it that I'm able to listen to two year old editions of Michio Kaku's "Explorations" show from the KPFA archives, and Brian Lehrer's podcasts etc. but when the BBC temporarily puts mp3s of the "Reith Lectures" or Beethoven up on its site, or provides a handful of podcasts, they're called "experiments"?

    More than two years ago, the wholly commercial Virgin radio company saw the benefits of (and public demand for) ogg and duly provided some streams - shortly after the supposedly public service oriented BBC had permanently shut down "testing" of its own ogg streams. The BBC could and should have been the world leaders in the use of the internet and its associated technologies. They should have invented 'podcasting' and been using the best and most accessible of modern stream formats all along. In 2005, their experiments should be with things like speex, bittorrent, theora and their own "Dirac" format - not with doing things that others have done as a matter of course for years. ”

  • Comment number 7.

    I agree with most of the others here. I'm really happy an effort is being made for cross compatibility, but the BBC really is going about it the wrong way. If you start off with open standards (really open, not microsoft's idea of open), and then you'll save yourself more hassle in the end.

  • Comment number 8.

    The comments concerning patent encumbered formats are well taken, but I don't think that changes anything important.

    Encode your video in unencumbered dirac/vorbis, wrap it in an AVI, serve an HTTP link to it. User clicks on link, video opens in their player of choice. It doesn't get any easier than that.

    All this is is a serious case of Not Invented Here syndrome.

  • Comment number 9.

    Hatta0, you might read a little bit more a bit GNU/Linux distros and it's multimedia-history. Ignoring the aspects of user experience and content rights, the main technical trouble is about patents on audio and video codecs on the client side.

    This is hardly a problem in the real world. Software patents are a grey area legally, especially this side of the Atlantic, but whether these codecs are actually patented people just use them anyway. The process of adding a 3rd party repository and getting the right packages is trivially easy and there are plenty of good instructions on the web (just google "distroname" multimedia codecs). Anyway the whole discussion is moot as there are already good, unpatented audio and video codecs such as the Ogg and Theora ones that anyone can use.

    But to the more general point - why are you making it so difficult to get access to your content BBC? As others have pointed out setting up audio and video streaming using open technologies is relatively simple, and will work on all OSes on the client side without special configuration. And you guys really need to drop this DRM nonsense - I can go to any torrent website and download a video file of any Dr Who or Dragons Den episode quickly, with and in a format that will play on virutally any OS or device with no hassles - from my Linux computer to my DVD player.

    I'd be happy to get these shows from the legitimate source but that just isn't an option. The reasoning that you need DRM to stop it being pirated is laughable when the only practical way to access the DRM'd content for a lot of people is to use one of the easily available pirated copies.

  • Comment number 10.

    "Hatta0, you might read a little bit more a bit GNU/Linux distros and it's multimedia-history. Ignoring the aspects of user experience and content rights, the main technical trouble is about patents on audio and video codecs on the client side."

    This is hardly a problem in the real world. Software patents are a grey area legally, especially this side of the Atlantic, but whether these codecs are actually patented people just use them anyway. The process of adding a 3rd party repository and getting the right packages is trivially easy and there are plenty of good instructions on the web (just google "distroname" multimedia codecs). Anyway the whole discussion is moot as there are already good, unpatented audio and video codecs such as the Ogg and Theora ones that anyone can use.

    But to the more general point - why are you making it so difficult to get access to your content BBC? As others have pointed out setting up audio and video streaming using open technologies is relatively simple, and will work on all OSes on the client side without special configuration. And you guys really need to drop this DRM nonsense - I can go to any torrent website and download a video file of any Dr Who or Dragons Den episode quickly, with and in a format that will play on virutally any OS or device with no hassles - from my Linux computer to my DVD player.

    I'd be happy to get these shows from the legitimate source but that just isn't an option. The reasoning that you need DRM to stop it being pirated is laughable when the only practical way to access the DRM'd content for a lot of people is to use one of the easily available pirated copies.

  • Comment number 11.

    "Do one thing - do it well"

    codec, wrapper, plugin, application UI and distribution + partners tie in ?

    You are definetly not doing one thing.

  • Comment number 12.

    "george,

    What you will see is an increasing list of content, in both audio and video formats, that we can share with you in an experimental way, allowing us to explore and test new ways of viewing and listening to some of our TV and radio shows. "

    thats fine and dandy george, but you could have setup simple Muticast tunnels on your servers years ago.

    wrote a wrapper for the VLC and Mplayer apps that are cross platform compiled for a very long time now so as to easly pass from wastful unicast streaming weblinks to Multicast streaming to many people with a single content feed.

    saving you lots of bandwidth and forcing the worlds ISPs to reconsider turning Multicasting back on in all their ISP routers and related kit.

    YOU DIDNT do this last step for your BBC Multicast trials , instead asking the likes of Virgin media and BT to peer with you and turn thier multicasting protocol back on to the end users.

    they dint do this so we were excluded from taking part.

    ring any bells, your providing some of the work, yet over complicating things that dont nedto be.

    all we need is a simple point to multipoint multicast tunnel on YOUR servers and a local multicast matching tunnel on your localhost ports that we can connect to, and use as other have said.

    useing direct links to near realtime content with a webpage we can que it up and cache it when we want to see it, simple and YOU would be providing some real (old MBONE network)INNOVATION for todays bandwidth straped wastful unicast streaming users and services.

    the ISPs dont seem interested in turning Multicasting back , its up to you adn other providers of video streams to put these tunnels in place and give us the ability to connect to your content, simple easy and all wraped up in a quick code revision of VLC/Mplayer ,a Mtunel java , + Bamboo java DHT codebase to use it on the worlds open markets....

    the multicast code (java and others)exists today, your alredy running multicast trials and have been for a long time, you miss outthe one part required, that multicast tunnel+wrapper to connect it all together, please just write it,install it on your web content servers and let us use it for years to come

  • Comment number 13.

    Hello George,

    I must concur with the previous posters.

    Please let go of the concept of ``immersive end-user experience'' and just provide URLs to Vorbis and Theora encodings of the unrestricted material. A simple HTML index will suffice; no need for media player plug-ins.

    Alternatively, take the budget for this apparent make-work scheme and use it to help the Gnash developers work towards supporting iPlayer.

  • Comment number 14.

    Concur with much of the above, I'm really happy that BBC technology (which previously was very open source minded, lots of Perl, for example) has not totally forgotten the open source world.

    There's a saying, KISS (Keep It Simple Stupid) which could be (Keep IT Simple Stupid) that the BBC need to repeat as a mantra. The iplayer is complex and unreliable and this is mildly misdirected overkill.

    It would be very useful for -someone- to step back and do some strategy (so that 'we' don't end up with, and pay for, five different players and content sites) and (why not) ask a few licence payers and users (and potential contributors, if we're in an open source 'world') before heading off down a particular development avenue.

  • Comment number 15.

    There's an interview with George and some more information/discussion at Backstage here.

    Nick Reynolds (editor, BBC Internet blog)

  • Comment number 16.

    Good news for us Amarok users.

  • Comment number 17.

    This comment was removed because the moderators found it broke the house rules. Explain.

  • Comment number 18.

    Thank you.. http://www.gelsesli.com/ sesli sohbet

 

More from this blog...

BBC iD

Sign in

BBC navigation

BBC © 2014 The BBC is not responsible for the content of external sites. Read more.

This page is best viewed in an up-to-date web browser with style sheets (CSS) enabled. While you will be able to view the content of this page in your current browser, you will not be able to get the full visual experience. Please consider upgrading your browser software or enabling style sheets (CSS) if you are able to do so.