ratpie

Category: technology

  • GRD? More Like GOELRO, Comrade!

    Music industry watchers will have noted the recent demise of the Global Repertoire Database (GRD), a project so misconceived that its collapse should more properly be greeted with relief than hand wringing and woe.

    Some background. In 2008 EU Commissioner Neelie Kroes posed a question to a small and odd group of interested parties at two roundtables. The meeting notes are online. What, she asked, could be done to move digital music towards a single internal market in Europe? The territorial nature of copyright, and the fragmented administrative structure, were identified as barriers. Sir Mick Jagger was present at the first meeting, as was Steve Jobs. Sir Mick seems not to have had an opinion about anything. Jobs suggested investigating an EU copyright to subsume national rights as a long term aim, and articulated the idea that became the GRD:

    Apple therefore advocates the creation of a central electronic repository for ownership information into which all collecting societies and other right holders feed information in a common format, and which could be accessed by licensees, so that the problems caused by the information vacuum can be eased and timely and accurate remuneration can be fostered.

    SACEM rehearsed a line that needed to be retired at the point Ebay was growing out of Beanie Babies – everyone in the music industry should be forced to repeat three times every morning ‘there’s really not that much data’, but instead you get this:

    SACEM points to the difficulty that collecting societies face when they deal with millions of works: for some works, they perfectly know who owns the rights; but for others, they need to check carefully the information.

    That’s right – mere millions. A third year undergraduate project.

    The rest, as we have recently learned, has just become history. Committees were formed, politics was played, Deloitte was hired, the slush fund was spent, nobody could really agree either scope or funding, incumbents were nominated to do what quickly stopped looking like innovation. Most importantly, and it is a lesson that the music industry desperately needs to learn, no ‘outsiders’ were given the chance to bring a new perspective or new capabilities. In other words, those who had failed to keep updated an infrastructure and set of business practices that were acknowledged by all involved to be obsolete, were asked to oversee and execute the regeneration.

    When the Soviet Union had no power infrastructure it was entirely appropriate to build power plants and a grid, and Lenin (more here) famously recognised the importance of the project, but also its massive scale and complexity and need for experts:

    Communism is Soviet power plus the electrification of the whole country, since industry cannot be developed without electrification. This is a long-term task which will take at least ten years to accomplish, provided a great number of technical experts are drawn into the work.

    Surely by now it is obvious that the music industry is already electrified! We need no grand Electrification project, especially not one which looks more like an effort to preserve a set of organisations and infrastructure that I sometimes refer to as the Union of Soviet Socialist Republics of Music. And it is a form of Soviet malaise, that faced with an exponential growth in complexity – which brings opportunity as well as threats – many music insiders look to the already failed system to protect them.

    But it can’t. In the new electrified music industry, data processing which is significantly weaker than music’s partner platforms (Google, Amazon, Apple, Pandora, Spotify…) is a handicap (remember ‘there’s not really that much data’). There’s no justification for lengthy discussion around standard terms on standard uses; new models require a thorough examination, which is something best handled by a team of experts with lots of industry input, so that could be considered a core competence for the collectives. Beyond well controlled new deal terms approval and CRM however, every function currently undertaken by collecting societies could be better either outsourced or federated.

    And the horizontals are where the exciting things can happen, particularly where large groups of individual rights owners are affected. Instead of pushing songwriters and performers into national monopoly providers, communities of interest could bundle royalty collection and distribution with a range of supportive services. Private business could innovate in technology to drive down the cost of administration.

    Music already has too many databases and not enough competition to kill off the weakest. GRD was an old school move, to reach for yet another database, by a group of organisations understandably unable to imagine a future in which there is no need for them to exist. Our newly electrified music industry needs copyright utilities for sure, but, comrades, it’s time to leave the Soviet power behind.

  • To Get Paid We Should Set Music Free

    Digital music, having started out as a tied product with iTunes and Windows Media locking the files to devices and software players, achieved a certain freedom as DRM came off, but is now migrating fast back to proprietary encrypted formats. These locks benefit only their owners and operators. Switching from Spotify to Rdio, or Rdio to Deezer, or Deezer to Zvooq, means abandoning your library and playlists, and starting from scratch.

    It is unarguable that this approach is providing a significant and growing flow of money into the music industry. IFPI reported that subscription and streaming services produced $1b in 2013, and grew by over 50% year on year. In their bid to sign up subscribers these services are innovating rapidly in user experience, and racing to achieve global scale. There’s a long way to go, with around 30m active subscribers across all services. If music follows broadband into the home, there are already about 700m households in the market.

    On a standard Rogers model this means that we are out of the ‘innovator’ market and into early adopters. But only just. And the early adopters are very important, as they have the highest degree of opinion leadership. This segment is where a new paradigm flies or flops. And this is where a kind of inverse evolution is at its most dangerous.

    Evolution, in the popular phrase, is bad for the individual but good for the species. It is a binary test for specific unfitness. Early adopter markets on the other hand are good for the individual, but terrible for the species. No player has a strong incentive to sacrifice present growth for future sustainability; each prefers to act as a highly competitive individual rather than a category.

    And the truth is that the qualities of a successful category are very different to the qualities of a single offering. Choice, variety, many price points, and often some technical compatibility with competitors are what drives adoption into the mass market. In recorded music we have had many iterations of technical delivery, so we know this well. A £29 portable CD player plays the same CD as a top end audiophile model costing £29,000 (yes, they do exist).

    This is what we need for digital too if we are to push out of the early adopter market into the early majority. We need the influential early adopters to be saying not that one or other service is great, but that subscription music is great, for all the category benefits outlined above. Music otherwise risks sacrificing its paid-for niche to the irresistible rise of unpaid platforms, and for a far less certain environment for investment in new artist development and high quality music production.

    There is a smart way to push this process forward, and that is through library and playlist portability, enabled through a simple metadata file format (the audio itself is becoming ubiquitous anyway so would not need to be transferred unless missing in the receiving service), and a voluntary trademarked scheme for the services themselves. A ‘non-portable’ flag in the original supply chain metadata could allow music owners to offer enforceable exclusives to services if that was seen as valuable. All but the biggest service would have an incentive to participate as there would be more to gain than to lose; a service removing portability in response to losing customers would suffer a big loss in reputation and trust.

    Throughout the history of recorded music, innovation that adds value has on balance expanded the market. Cassette tapes made vinyl records more valuable by making the music portable. ‘Rip, mix, and burn’ added value to CDs in a similar way. Millions of tracks available with simple search and click, and sophisticated playlist and library management tools are doing the same in digital music. Giving consumers full ownership over their investment in the time and trouble to learn and use these tools will create more value for them, and more value for music owners and services as subscription heads into the mainstream.

    Portability will release a wave of investment in music service innovation to serve all kinds of consumers better, until, as with CD players, it would be inconceivable to imagine a well equipped home without one. It is not particularly challenging, either technically or legally. We should, once again, give music libraries and playlists the right kind of freedom, from restrictions, or risk being overwhelmed by the wrong kind, freedom from payment.

  • Downloads versus Access? Follow the Science AND the Money!

    One of the big debates in music is whether the download is dead, killed by ubiquitous connectivity and streaming services. Certainly the expensive (>$0.5) download seems to have plateaued, and streaming revenue is growing strongly tempting new entrants who bring with them investment in better consumer experiences. The end game, goes the argument, is that the device you plug the earbuds into becomes a terminal, connected to choice and service such as the music fan has never before experienced.

    Two things make me pause. The first is a nagging doubt that the mobile and wireless networks needed to deliver always on access really will live up to expectation in the near future. Telecoms seems to be in turmoil around the developed world, with companies facing huge investment decisions at a time when the traditional sources of profitable revenue are crumbling, cannibalised by feature substitutes carried over undifferentiated data. Regulation quite rightly favours innovation over protection, but has not yet addressed the funding of infrastructure when networks have been turned into a utility rather than a service. As for seamlessly transferring sessions between protocols and providers, well I am not holding my breath.

    The second pause point is the astounding innovation in storage, seemingly in all configurations and target uses. The standard server hard drive now carries 4TB of data, with 6TB on the way and densities in labs promising as much as 60TB. The advanced science suggests that we are in the foothills; IBM posit that a mere 12 atoms are needed to store 1 bit of data, suggesting a 100 times increase in storage density. That may be some way off, but each month brings announcements about technical breakthroughs promising twice to ten times today’s densities.

    In music only a very few tracks have enough genuine demand to justify providing a live connection either for a download or a stream, or the resulting torrent of reporting to share out the minute royalties generated. It’s easy to understand popularity in music as a pyramid of 3 multiplied by powers of 10. 3 tracks are screamingly hot, 30 in demand, 300 popular, and so on. Today the entire repertoire of drivetime radio would fit on a memory chip the size of a thumbnail, and as we know that selection changes so slowly it makes a real thumbnail look volatile. A square inch will soon be able to store 15,000 tracks – that is about 750 hours of music. We are less eclectic than we think; many of us turn off the radio if the DJ starts playing music we have not heard before.

    What are the business drivers here? A catalogue owner would do much better to license a selection of back catalogue on a once only fee to be sold as a chip embedded in a device, enjoyed for a year or two, and then recycled or destroyed. New and popular music would naturally remain on a per use or per download tariff as it justifies the cost of multiple deliveries and all that royalty reporting, and could be managed in a more dynamic cache on the device. Music services could get adept at using network quiet time to refresh and pre-cache, giving the network owners a real reason to subsidise the bandwidth music uses off-peak in preference to expensive and contended peaktime data. The embedded library could have something akin to a firmware update from time to time, if necessary. Importantly it could be entirely locked to the hardware, meaning a effortful loopback to a recording device if anyone wanted to set it free.

    The sell to the consumer would be relatively straight forward too. Buy a library of evergreen music for $50, and add all the latest chart hits for $25 per year if you want to, or buy a la carte if you prefer. At a wholesale rate of $40 music could be used as a sales incentive for just about any big ticket or recurring item. Remember this is back catalogue; I predict that the music rights owners would compete to get on the chip.

    Having strong musical preferences and making playlists has always been a minority sport. The technology and the business drivers both say that in a few years we will be buying downloads in vast quantities, but not choosing them track by track. We might not be paying for them directly. The result will be a much more efficient market, and most important for us all, a certain amount of freedom from over-dependence on privacy invading clouds and undependable mobile networks.

  • What the Future Used to Look Like, in 2006

    A new feature in iOS 7 reminded me that the sky, which was supposed to be falling eight years ago, has remained inexplicably buoyant. The feature is called the ‘Multipeer Connectivity Framework’ and it provides a simple way for devices to connect to each other over bluetooth in order to exchange files or messages.

    Eight years ago I was thinking about WiFi mesh networks, and wondering why what I thought of as the ultimate playground sharing device had not been brought to market. The idea was simple; a portable device which let you create sharing groups of friends, securely connect to each other, and pool collections of files, chat, and send photos to each other. If any one of a connected group of devices had an internet connection that too could be shared, giving everyone access to the global filesharing and chat networks. Private, ad hoc, and intermittent sharing could at the time be thought of as a copyright owner’s nemesis. Instant tape trees, updated and improved. Even back then the bill of materials for such a device, and the networking protocols, were available and cheap. It could have been delivered for under $100.

    For better or worse it never happened, like so many alternative futures we have dreamed of over the first two decades of ubiquitous digits. I still think it would be exciting, and terrifying, if such a device caught on in the playgrounds of the world. I wonder, for instance, if regulators would feel compelled to come up with new crimes alongside copyright infringement, such as ‘incitement to upskirt’, or ‘assault by clandestine chat mobbing’.

    I was at the time working with a music industry group on collective responses to such developments, the membership and discussions of which have to remain confidential. I can however share a paper I prepared, in August 2006, full of excitement about what seemed to me to be amazing advances in technology with huge potential for good as well as a bit of copyright mischief. Many of the references and links have disappeared, but for the sake of perspective and posterity, here it is:

    New and Exciting Ways to Experience Music, Part I

    Let’s have a key! How about:

    (n) for “now” – meaning implemented and in use

    (l) for “likely” – meaning the technical means are out there, and perhaps already being used by a few intrepid pioneers

    (p) for “possible” – meaning that existing techniques could be adapted, or new techniques developed using similar technologies to those either already deployed or in the pipeline

    It will also be helpful and avoid repetition if the NEWTEMS are grouped into families, so that the variations on a theme can be understood and their impact assessed. The status code should apply independently to the variations as well as to the family. Also, where possible, examples, with web links.

    Newness is a matter of viewpoint, so the criteria should probably be whether they are sufficiently both digital and musical not whether they emerged in the last n years.

    Internet filesharing, ‘classic P2P’ and its successors

    The components we need to consider are the communication protocol, the application, and the transport. To illustrate, in the case of Limewire the protocol is Gnutella, the application is Limewire, and the transport is TCP and UDP. The Gnutella protocol is released under an open source licence and is widely adopted. The transport uses fundamental internet protocols. The application is released in two versions, a paid for application and an open source project available under the GNU GPL. The Limewire source code has been used by other project teams to create new and modified applications.

    Gnutella is what’s known as a ‘decentralised’ protocol, where both searching and file transfer are managed by the application running on the user’s host computer, rather than by one or more central servers. It is however relatively easy to discover which host is advertising a particular file, and some implementations allow the shared directory of each host to be listed.

    There are many sources of information about filesharing protocols and applications on the web. I shall here just highlight a few developments that will have some bearing on the VRS agenda. Some applications will be hybrids of the following categories.

    Anonymous P2P (n)

    Anonymity is achieved by constructing routing messages so that it is difficult to determine the source or final destination of any given packet of data. Messages may also be encrypted.

    Example: GNUnet – http://gnunet.org/

    Encrypted P2P (n)

    Messages are encrypted so that it is difficult to impossible to discover what content is contained in the data that is flowing around the network.

    Example: ANts P2P – http://antsp2p.sourceforge.net/

    Filesharing over encrypted connections (n)

    An encrypted and private network is created over which filesharing applications are run. Increasingly internet chat sessions are being encrypted, and the chat applications are including easier ways to share files.

    Example: stunnel – http://www.stunnel.org/

    Filesharing over anonymous networks (n)

    An anonymous network is created using a series of temporary routes and proxies which masks the source and destination of some network traffic. Filesharing applications then use this anonymous network.

    Example: Tor – http://tor.eff.org/
    also: I2P – http://www.i2p.net/

    Trusted F2F networks (n)

    Generally what’s meant by the ‘darknet’, F2F (friend to friend) networks are formed through permission mechanisms, either where each member has to accept a security token from every other member, or where a member has to allow a new member to join and then the new member is accepted by all existing members.

    Example: W.A.S.T.E. – http://waste.sourceforge.net/

    Wireless Internet, WiFi, WiMAX, Mesh Networks

    Anything that works over the wired internet also works over wireless, so all the above apply equally to this section. These technologies also have the capability of forming local point-to-point connections, and local networks, without any internet connectivity. While some WiFi internet will be provided through commercial hotspots, and require contracts and accounts in order to gain access, some will be provided through free and open networks provided by municipalities, volunteer groups, charities, and through neighbour sharing of home wireless LANs.

    Shared WiFi (n)

    Rather than opening a standard WiFi router up to open public access some operators are controlling who can connect to their networks, which ensures a degree of privacy or trust. Terms and conditions are sometimes subject to a voluntary peering agreement which is designed to prevent others from using the network without contributing their own bandwidth to it. Open source tools are available to manage shared WiFi networks.

    WiFi is also capable of creating small scale point to point or group networks between mobile devices. These could be used to transfer files from one portable music player to another, or allow a group of friends to share each others’ content. WiFi could also enable such devices to connect to the internet.

    Sony Mylo
    http://www.sony.com/mylo

    Wifidog
    http://dev.wifidog.org/

    WiMAX (l)

    WiMAX is the marketing name for a technical standard that is best suited to providing high bandwidth network access over a distance of less than 5km. It can carry mobile telephony, TV and video signals, and internet access. WiMAX is currently either absorbing or concluding interoperability agreements with other rival standards, and is being deployed on a large scale in several countries.

    WiMAX Forum
    http://www.wimaxforum.org/home/

    Mesh Networks (n)

    Mesh technologies enable much larger networks to be created, with each node acting as a router as well as a point of connection. A mesh can also use the wired internet to hop between otherwise autonomous mesh networks. In practice, mesh networks are likely to be used for sharing internet access and files around buildings, small areas of a city, or at events such as festivals, where their ad hoc characteristics make them most suitable. They also offer a very quick and low cost way to bring internet access to disaster or conflict areas. Generically such capabilities would enable what could be termed a ‘playground sharing device’, with SMS and internet chat applications too.

    Locustworld
    http://www.locustworld.com/

    Radio Babylon (a state51 project)
    radio babylon

    SONbuddy
    http://www.sonbuddy.com/

    Short Range Radio, Cables, other Transmission Methods, Pluggable and Removable Media

    Much like the wireless technologies listed above, this set of communication technologies can be used for a range of purposes in order to provide point to point connections, networks, and access to other networks. The shorter range radio standards are generally considered as ‘wire replacements’, used instead of cables to connect devices and peripherals. Enabling technologies include USB, FireWire, Bluetooth, WiFi, UWB, Infrared, Memory Cards of all shapes and sizes, portable Hard Drives, and all other forms of removable media storage, including CDs.

    Sneakernet (n)

    Rather a broad term perhaps, but useful to collect a set of essentially person to person activities that are enabled by new digital technologies. Sneakernet refers to data transfer which involves at least one party carrying something containing the data. Here are a few subcategories…

    Bluetooth transfer: files are loaded onto a bluetooth capable device such as a mobile phone, and BT file transfer is used to copy the files onto another device. There’s an interesting case where Verizon Wireless, apparently in order to comply with contracts with content providers, crippled some bluetooth features on a phone including file transfer, and consequently fell foul of a class action.
    Verizon Wireless – ‘We never stop working for you.’
    http://www.verizonwireless.com/b2c/footer/legalNotices/v710.jsp

    USB Hard Drive (Thumb Drive): A small Hard Disk is loaded with files and can then be plugged into mobile phones, mp3 players, computers and games consoles in order to copy the files. Capacities have been growing and have now reached 16GB.
    CellDisk 16GB
    http://www.iocell.com/english/products/16gb_e.asp

    USB Data Transfer: Even without a host computer two USB enabled devices, including hard drives and memory cards, can be connected and their files copied.
    Macally SyncBox
    http://www.macally.com/spec/usb/input_device/syncbox.html

    On The Horizon

    Internet over Powerlines (l)

    While for the most part internet access is a fairly simple business, with households maintaining an account with an ISP, there are some interesting developments with wireless access, and with powerline networks, which make for a much more complicated picture. An entire street for instance could elect to share passwords on a single high-speed internet account, and use the electricity cables to provide access.

    Powerline adapter
    http://www.trendware.com/products/TPL-101U.htm

    The next stage would be to add encryption and mesh networking to provide each home with security and privacy, but to leave a segment of the network open for more community spirited activities, such as sharing the minutes of the street party committee meeting…

    Really Mobile Internet (p)

    Some of these networks might even be automotive, by which I mean that they move under their own steam, as in this unique and wonderful project which creates a flocking swarm of helicopter computers and a mesh network…

    The UltraSwarms
    http://gridswarms.essex.ac.uk/

    Store and Forward Internet (l)

    There is no need to create an end-to-end session to get the benefits of internet access, and therefore filesharing and other content distribution. So far, store and forward techniques are being used to bring the internet to remote and deprived communities, but they could equally work wherever people can get hold of compatible devices.

    Internet Village Motoman
    http://www.ratanakiri.com/

     

  • Preserving the Open Internet the Easy Way

    No regulation is easy of course, so to start, here’s an apology for the misleading title. Sorry. But it need not be as difficult as it might seem to make the few small adjustments needed to ensure that the next decade does not see the effective end of the open and public internet which promises to deliver so much to humanity.

    Consumer ISPs love to shout about how much bandwidth they are offering their customers. Just look at the words they use! “Unlimited BT Infinity 2 Superfast broadband” screams BT, the UK’s former nationalised network owner. In this case the consumer offering is 76Mb/s download speed, and 19Mb/s upload. But those are qualified – ‘up to’, which of course starts with zero Mb/s.

    As alternatives to the commoditised ADSL 2Mb/s heavily contended offerings appeared in the market, the UK’s Ofcom felt the need to help ISPs avoid misleading the public with false promises. Here’s their key message:

    There has been a noticeable trend for some ISPs to advertise their products based on faster and faster headline speeds . However, the evidence from Ofcom’s research indicates that these headline speeds are rarely achievable in practice by the majority of consumers that buy them.

    from the 2010 Code of Practice that you can read in full here:

    http://stakeholders.ofcom.org.uk/telecoms/codes-of-practice/broadband-speeds-cop-2010/code-of-practice/

    Observant regulators will of course have noticed that the speed being advertised applies only to the connection from the consumer to the ISP’s own network, and that thereafter it is anyone’s guess how much bandwidth will be available to a given content provider or service. And, that while to an extent the consumer ISP has no control over how much bandwidth a third party will have provided to get their message to the public, they do have some control over the way and with how much capacity their networks connect to the world.

    The movement of large and popular services off the public internet and into private connections is well understood and well documented by now. It would be an ignorant regulator who somehow conflated Google, Facebook, Amazon and the rest with an open internet agenda; surely such a regulator would be laughed out of town. For these large network owners, ‘open’ is simply an unfortunate stage you have to go through before cajoling consumer ISPs into closing again, and shutting public traffic out of the connections that carry their traffic.

    Faced with the prospect of competing on price per Mb/s, while making large investments in their core networks almost solely for the benefit of third parties who themselves are turning their backs on the open internet, it is not entirely surprising that the ISPs with the means to do so are trying to develop their own content services. In the UK, BT has bought sports video rights to try to tempt people into its TV service. Other ISPs around the world are buying film rights and other bundles of content, in the hope that their own bandwidth intensive services will enable them to compete internally for their customers’ attention and money in a zero sum game of their own making,

    So there, in a nutshell, is the intersection of several strong commercial incentives, which between them are conspiring to deprive the open and public internet of the investment it needs if it is to continue to enable the innovation that created the private peering behemoths in the first place. If trends continue, what’s left of the internet will become the ‘third service’ alongside ISPs’ own services and the major private networks (Google et al.). It would not be surprising, given that everything politicians hate – from illegal content to freedom of speech and protest – will be flowing down it, this third service will not be high on the list for care and protection.

    Neglecting the open and public internet would be a mistake however, even for Governments which are naturally concerned with preserving their currently pre-eminent internet business sectors. Here’s Neelie Kroes summarising for us at the IGF meeting in Kenya in 2011:

    The Internet is changing the world, and it’s not just trillion dollar marketplace. It is a forum where people connect, a platform for astounding innovation, and a powerful vehicle for human rights and fundamental freedoms.

    Kroes will surely recognise better than almost anyone else on the planet that if you have to define those terms, innovation, human rights, freedoms, and even the more narrowly focused network neutrality, you have failed before you start. As an aside, a proxy measure for the scale of this failure might be found in corporate spending on non-corporate activities – higher is worse – as companies spend to defend their franchises, not to open them to competition.

    The opportunity here, the ‘easy way’ in the mischievous title, harnesses the competitive nature of consumer ISPs to investment in the open and public internet. It pushes the proper questions of who should be allowed to run a computer network, and who should be allowed to connect their network to the internet, to where those questions are best answered, in national regulations and internet governance. It recognises the central role of IXs, internet exchange points, in enabling the internet to be what it is. Topically, it borrows the idea of capital adequacy from international banking reform efforts.

    It is simply this; ISPs should be obliged to provide a reasonable percentage of their total sold bandwidth through public peering at an internet exchange. Then as consumers we could be sure that we really are buying reasonable access to the open and public internet, and as participants in business and culture we could have confidence that our access to the world will not be starved away by neglect or by gatekeeping on the last few hops. And as an important side effect it would remove the incentives Governments have to help corporate interests grow unfairly at the expense of the conditions on which future innovation and freedom rest.

  • Netflix Explains Why The Internet Is Not Good Enough For Them

    The migration of new digital services off the public internet is an established pattern. Driven by economics, and with real money at stake, it is not likely to slow down any time soon.

    To recap, large content providers such as Google (with YouTube), Facebook, Amazon (mostly driven by its Web Services division), and Netflix, among others, are taking their traffic out of public exchanges and setting up private connections with consumer ISPs. They make a compelling offer. ISPs save costs by connecting directly and the exchanges that ISPs largely fund avoid the investment that would enable this traffic to remain on the public internet.

    Helpfully, Netflix has been sharing its technical architecture and its offer, called Open Connect, to ISPs. Here’s a presentation from a recent network operators’ meeting:

    [file no longer online – 5/11/2017]

    and you can see the original powerpoint (link updated 2017) here:

    https://dknog3.dknog.dk/agenda/

    and here’s the horse’s mouth:

    https://signup.netflix.com/openconnect

    Free is a very convincing price, and few network operators have the tools to measure the future impact of a lack of investment in common resources.

    Netflix even sweetens the deal, by offering a higher quality video stream to subscribers who accept the Open Connect equipment and connection. And this illustrates the policy issue very well indeed.

    If our policy makers believe that the open and shared internet is a driver of innovation and value for the public they should be concerned. It is being re-engineered to deliver a second rate level of service. We are still a long way from the point where ISPs don’t need the public internet at all, but arguably that is a destination now in sight.

  • The Internet is No Longer a Duck

    It waddles and quacks, but is the Internet still a duck?

    The Guardian published on March 28th 2013 an article by Cory Doctorow with the headline ‘Copyright wars are damaging the health of the internet’. His argument is simple and persuasive – that the Internet has become too important to our civil, cultural, and personal lives and well-being, and our ability to protect our freedom, to be compromised in the attempt to prevent copyright infringement.

    But many people, mostly without knowing it, have already deserted the Internet. Their traffic is now carried over private networks owned by large content and service providers. The Internet is by definition a network of networks, and your packets used to find their own way often across many other networks. The Internet still exists, but compared with these high capacity and well connected private networks, internet traffic is decidedly second class.

    If this seems surprising, consider a few of the more popular internet activities. Watching a cat video on YouTube, for instance. Google has been keeping up with YouTube’s huge traffic requirements by connecting to all the large consumer ISPs directly. GMail the link to a friend, G+ it, and you use the same private connection. Tweet it and even for such tiny data demand as tweets, quality of service issues could well have driven your message via Twitter’s private connections. Your DropBox is hosted by Amazon – yes really, the online shopping site -, another private network currently spinning its own web. If it is big or successful, the chances are it is getting off the Internet as fast as it can.

    This does not mean the Internet is not healthy. Far from it. The open Internet is enabling extraordinary innovation all over the place. Having the giants on their own expressway ensures that the smaller players don’t get flattened on the common roads.

    It does mean, however, that for ordinary users, including bloggers and tweeters of all persuasions, the health of the Internet is irrelevant. Regulators are now dealing with the behaviour of a relatively small number of very big companies, not with anything that can really be considered a ‘common’, or public space.

    A much deeper understanding of today’s reality and direction of travel is needed if those regulators are to protect open innovation and communication for us. The fact that internet protocols make seamless the connections between networks that allow public ‘rights of way’ – the old public Internet – and the new private networks, should not confuse us. It might still waddle and quack, but the Internet is no longer the duck it once was.

  • Yet Another Private Internet

    Lobbyists sometimes try to simplify issues, for policy makers and for grassroots and other supporters, and one of the most potent ideas surrounding Internet policy is that what you don’t like will ‘break the Internet’.

    Here’s John Naughton neatly encapsulating this approach for the UK’s Guardian newspaper in a comment on SOPA/PIPA early in 2012:

    …the most worrying aspect of these bills is that they would distort the architecture of the internet in ways that would cripple its capacity for enabling innovation

    Read the whole thing here under the headline ‘Sopa and Pipa: don’t let big business break the internet’: http://www.guardian.co.uk/technology/2012/jan/08/online-piracy-challenge-sopa-pipa

    The attraction of this is easy to understand. For any highly technical question with complete uncertainty about the outcome of any policy decision, all the attention-deprived Government official or politician has to do is ask, ‘will it break the Internet?’ If there is sufficient noise in the ‘ayes’, risk that they may be right will ensure inaction.

    While this is going on in public, privately companies are breaking the internet in all sorts of useful and interesting ways. For a small business I am a large buyer of datacentre services and connectivity, operating a highly connected music platform hosted at Interxion in London. For many reasons we own and manage our hardware – but I can now, should I so wish, buy private connectivity within the datacentre to Amazon’s Web Services. Here’s what they say about it:

    Using AWS Direct Connect, you can establish private connectivity between AWS and your datacenter, office, or colocation environment, which in many cases can reduce your network costs, increase bandwidth throughput, and provide a more consistent network experience than Internet-based connections.

    A quick survey shows that Amazon has been rolling out private connectivity via high quality datacentres around the world. You can bet also that the peering points which connect consumer access providers to the Internet also benefit from these direct connections, alongside connections to Google, to Akamai, Limelight, and other providers of media and application delivery services.

    The chances are increasingly slim these days, if you stick to fairly popular sites and services, that your bits will ever touch the Internet. In fact the real Internet is already significantly more expensive, more congested, less reliable and less consistent than all these private platforms.

    Why is this important? Well, perhaps monopolies built on private networks should be regulated as private businesses, rather than as users of a miraculously unlimited and unfettered Internet commons. So when Google in Germany launches a campaign saying ‘Defend Your Network’ (Verteidige Dein Netz), it is important to understand that Google no longer uses our public Netz, it has its own, private Netz.

    Google’s property rights in its network should of course be respected, and robustly defended, in all the territories it operates in. Help them if you wish, but not in the name of the extraordinary communal efforts that enabled them to achieve their scale, and which they have now left behind.

  • Leapfrog or Lag? Music Tech in Developing Markets

    We have been conditioned to think that developing nations are somehow privileged to have the opportunity to skip whole generations of technologies, moving straight to what we in developed economies are embracing because of its promise for the future.

    Here’s the Economist, writing in 2008:

    The mobile phone is also a wonderful example of a “leapfrog” technology: it has enabled developing countries to skip the fixed-line technology of the 20th century and move straight to the mobile technology of the 21st. Surely other technologies can do the same?

    The full article is here, and despite the common sense conclusion, that sometimes new technologies are dependent on a whole set of building blocks that might not be present, it reflects a prejudice against the past that only the rich can afford.

    Another example, from an arbitrary blog:

    The internet is growing in India, and most of it is on the mobile phone.  Many, perhaps most of the world, will access the internet only on their phones.  They are skipping the PC and not even blinking or thinking twice.

    So how important is the mobile OS market?  It will rule the digital world sooner than you think. (the rest here)

    On a visit to Vietnam in 2008 I too was struck by the way that the mobile phone had become the ubiquitous and essential tool for business, and for all the normal traffic of life. But there was something else too. Computer shops came in two types, one familiar from poorer shopping streets in developed countries, with components and a few low end desktops and laptops, second hand and repairs. The other type presented top end tower systems in perspex cases, objects of aspiration and desire.

    Market research site iSuppli sees this same pattern in China, with the PC being the aspirational home media centre for emerging middle class households.

    …the PC appears to be winning as the main source of evening entertainment in China’s households—for playing games, reading, watching videos, listening to music, communicating and emailing, as well as shopping.

    The fact that we are excited about our smartphones and tablets, and are devoting less attention to our two year old home PCs, should not lead us to assume that a broadband connection and a PC would not therefore seem like the apogee of well equipped home life in countries where it has so far been unattainable for the vast majority of families.

    So perhaps when you have had a fully connected and well equipped home for ten years the next frontier might be mobility. Ordinary families in developing economies might prefer to catch up with the rich, rather than go without the benefits of the last decade of connected home computing.

    For music this means preserving what we have learned about music on the PC, about downloads and home media servers, and about a more social and versatile approach to delivering music. It might well mean multiple modes of acquisition, including peer-to-peer, for much longer than seems likely in developed countries. Synching with cheap and portable MP3 players would be much more important than music through the cloud to smartphones and tablets.

    In fact, while the CD remains a mass market music product, we see in some developing markets all the conditions that fed the growth of Napster, but on a far larger scale. The disutility of a music subscription service, tethered inconveniently and expensively to an account and a limited number of devices, seems to be less exciting a vision of the future than the idea of ‘leapfrogging’ suggests. I suspect, given a choice, many of us would happily leapfrog right back to 1999. Maybe in China, Latin America, Malaysia, Vietnam, and many other countries there will be more of an incentive to embrace and develop with the market rather than fight it this time around.

  • Networks or Services – Who is Music’s Better Partner?

    When I formed the idea for a new business bundling broadband with music, in 2003, my thinking was guided by two simple principles. The first was that customers already saw music and broadband as a natural bundle. And the second was that making ISP networks able to manage and account for music services was a difficult but achievable stretch, and once completed would be a much stronger partnership and more effective competitor for unlicensed music than would over-the-top services.

    November 2012 saw Google tightly integrating its storage locker service to its email product, so that up to 10GB of files can be quickly and simply shared with friends, family, and colleagues, as email attachments. One simple and obvious use would be an email revival of the tape trees that <irony> nearly saw off the recorded music business in the 1980s </irony>. So what the right hand giveth with the very innovative and exciting new music service in Google Play, the left hand might not exactly take away, but it certainly helps music fans level the music distribution playing field a bit.

    The commentary on Google’s move is about competition between services, how Dropbox and Box might struggle against such convenient integration, and how email is a natural sharing medium. It illustrates how fierce that competition is, and how much churn there is as one service displaces another. This is also true in music, both licensed and unlicensed.

    But let’s imagine for a while that we can move the business end of music copyright out of the services market and onto the networks. For an ISP, music is currently just data. Charging for it as music, and using the ability to build library management and playlists, and offer a coherent delivery path to consumer electronics, all this would add up to a considerable win. For services of course, authorised music is just a more expensive form of unauthorised music. Licences buy the right to innovate around the identity of the consumer and a knowledge of what they listen to, but the benefit seems not to be there for many businesses, or at least not significant enough to go through the licensing pain and cost.

    My 2003 startup – Playlouder MSP – now operates a white label service enabling networks to offer music tightly integrated into their customer management and billing systems. Record companies have not yet been bold enough to offer true network licences, preferring the comfort of a managed service environment. Meantime competition rages over the attention and data that network users create, without really discovering the services they might actually directly pay for. ISPs and other network owners are music’s ideal partner in the quest to turn data back into music, while services push back by selling music by the GigaByte.