In our 2004 year-end summary, we called into question the relevance of the DRM components of the MPEG-21 suite of standards.  MPEG-21 had two primary components related to DRM: the MPEG Rights Expression Language (REL), a standard for expressing content rights in a machine-readable language, derived from ContentGuard's XrML language; and the Rights Data Dictionary (RDD), a standard for expressing fundamental or "atomic" components in rights operations (e.g, "copy") in order to facilitate interoperability among different rights management schemes, based on semantic data modeling work by the UK firm Rightscom Ltd. 

We see no production implementations of either MPEG-21 rights component at this point, and we are ready to call MPEG REL irrelevant to the market, which has passed it by.  There have been one or two academic reference implementations of MPEG REL, and the Digital Media Project (see below) has incorporated it into its IDP-1 spec.  Microsoft uses a version of XrML in each of its DRM technologies (as do a handful of other companies), but the company has not indicated that it will migrate them to MPEG REL, nor does there appear to be any particular reason for it to do so.  As for RDD, it is an elegant approach to rights interoperability but one that is less practical than other approaches such as Coral (discussed below). 

OMA DRM, also discussed further below, uses a "rival" rights expression language, ODRL (Open Digital Rights Language).  ODRL is purview of the loose-knit ODRL Initiative, which is led by Renato Iannella (the language's original inventor), now at an Australian research lab, and Susanne Guth, now at the wireless carrier O2.  The ODRL Initiative made a provocative and useful statement in July, when it published a mapping of the Creative Commons content licensing scheme to the ODRL language -- thus demonstrating that Creative Commons is a rights management scheme after all, albeit one that does not use encryption or other means of content protection.

By far the most significant implementations of ODRL are the subsets used in the OMA DRM 1.0 and 2.0 standards.  The Open Mobile Alliance has been positioning OMA DRM (and its companion, OMA Download) as open-standard alternatives to proprietary DRM technologies such as Microsoft's Windows Media DRM and Apple's FairPlay (for iTunes), apparently as a way for companies like Nokia to keep Microsoft and Apple out of their back yards.

OMA DRM 1.0 is now well established within the mobile industry, with more and more services based on it being launched.  However, OMA DRM 1.0 is a narrow spec, designed for "lightweight" content such as screen savers and ringtones, and for low-functionality handsets.  It only provides "forward lock" functionality, the ability to keep users from forwarding digital content to others.  OMA DRM 2.0 is a much more powerful, flexible standard, designed for full-featured multimedia handsets and containing features such as domain authentication that apply to home entertainment networks (HENs) as well as mobile devices. 

The mobile industry has rallied around the OMA standards over the past couple of years and made great progress.  But during 2005, momentum around OMA DRM 2.0 slowed dramatically.  The mobile industry failed to produce handsets and content services based on OMA DRM 2.0 in time for the year's consumer electronics product introduction cycle, i.e., summer or early fall.  Finally, in December, France Telecom's wireless carrier Orange launched an OMA DRM 2.0-based content service, initially in Switzerland, with DRM technology from the French conditional access vendor Viaccess and digital music distributor Musiwave, and handsets from Nokia.

OMA DRM is taking hold primarily in Europe; the standard's loss of momentum is jeopardizing its chances for adoption across the Pond in North America.  Two companies with proprietary DRM, Groove Mobile and Melodeo, have emerged as leaders in the mobile music distribution scene in the US and Canada by launching services with such carriers as Sprint, Bell Canada, and Rogers Wireless.  Another carrier, Cingular, launched a mobile music service in the US with Apple/Motorola devices and FairPlay DRM. Elsewhere, the proprietary mobile DRM vendor SDC is working with Telstra to launch a mobile music service in Australia.

One reason for the OMA DRM slowdown has been the still-unresolved wrangling over DRM patent licensing terms, as we discuss in our year-end DRM Technology summary.  The intellectual property at issue includes patents on rights expression language technology that are part of ContentGuard's broad DRM patent portfolio; although ContentGuard created XrML, it asserts that its patents apply to any REL implementation in DRM.  The backers of ODRL have always positioned their language as royalty-free, and this past April, they went so far as to publicly challenge the validity of ContentGuard's IP.  As ContentGuard reorganizes under a new ownership structure and new management, it will be interesting to see how it approaches this issue in the coming year and whether it decides to include litigation in its strategy.

The flap over OMA DRM patent royalties makes news because the mobile industry has used the public press as a way of gaining sympathy for its side of the argument, but it is not an isolated occurrence.  The issue of technology licensing, and fees associated with it, pervades just about every DRM-related standards initiative -- so much that it calls the term "standard" into question.  Most DRM standards bodies are now really consortia that have IP licensing pools attached to them.  Sun Microsystems is attempting to buck this trend with its DReaM Project, which it announced back in September: Sun intends to create an open DRM standard through collaborative community source development that "invents around" the existing patents.  We believe this effort to be naive and unrealistic, and we do not expect it to succeed in its proposed form.

Efforts to create standards and consortia in the HEN environment ran into trouble in 2005 because there were simply too many of them -- a surfeit of efforts and good intentions.  At least three such efforts originated within the set-top box market, including SmartRight (led by Thomson), Secure Video Processor (SVP, led by a number of players including NDS, Thomson, and STMicroelectronics), and TIRAMISU (Nagravision and Optibase); there are others that are in earlier stages and have not been publicly announced. 

Of the publicly known consortia, SVP (which launched in 2004) is gaining the most traction, as semiconductor makers Broadcom and ST have actually introduced SVP-compliant chips for set-top boxes.  SmartRight appears to be dormant.  The others are just injecting ambguity and confusion into the market; meanwhile, proprietary technologies from the likes of Microsoft, TiVo, Digital 5, and Mediabolic are progressing. 

The consortia with the most promise for 2006 are the closely-related Coral and Marlin -- both of which owe their existence to technology from InterTrust, the DRM technology pioneer now owned by Sony and Philips.  Marlin is a joint development project that was announced in January 2005 by a consortium that includes InterTrust, its two owners, Matsushita/Panasonic, and Samsung.  The objective is to create interoperable DRM for portable digital media devices.   

Sony is reputed to be using Marlin as the basis for its revamped connected portable device and content services strategy.   This is a radical departure for Sony, not only from the proprietary OpenMG DRM technology it uses in its current Sony Connect music service and MiniDisc-based music players, but because it diverges from Sony's typical strategy of premium proprietary technology rather than collaboration with other vendors.

Marlin is expected to announce its first complete spec later this month at the CES trade show.  If it succeeds, it will essentially create a portable music/video player camp around Sony, Panasonic, and Samsung to rival the three other axes in the portable device market: Microsoft, Apple, OMA.  (Philips plays in all but the Apple camp.) 

The Coral Consortium, which began in late 2004, is an attempt to create DRM interoperability via a service provider architecture -- the essential idea being that service providers should be able to find market niches where specific types of interoperability have value and provide such interoperability.  We believe that this is the right approach to interoperability, because it involves actual market dynamics rather than theoretical elegance or idealistic platitudes.  We look forward to seeing the full Coral spec, which is expected by June 2006.

Coral is one of several standards initiatives that are currently approaching the problem of DRM interoperability from quite different directions.  Sun's Project DReaM also attempts to address interoperability, by breaking rights expressions down into atomic "play once" operations and passing permissions for those to various devices around a personal network. This is also a practical approach -- and was actually implemented in 2003 as part of the EURESCOM project in Europe -- but it is limited to "play" semantics and does not work with other types of rendering (e.g., on-screen viewing of documents) or with "transport" operations such as copy and lend.

Another approach to interoperability that made progress in 2005 is the Digital Media Project (DMP), the ad-hoc standards effort led by MPEG founder Leonardo Chiariglione.  DMP's approach involves codifying so-called traditional rights usages (TRUs) and supporting them in an interoperable DRM platform. 

DMP released the first version of its Interoperable DRM Platform (IDP-1) in May; the spec covered portable audio and video players to which a PC (or other more powerful connected device) would feed content -- a different approach from that of Marlin, which is based on identification and authentication of users rather than devices.  Although IDP-1 was technically impressive, it essentially -- and disappointingly -- put off the issue of interoperability.  The next version of the spec, IDP-2, to be released in early 2006, is supposed to finally address interoperability, and it will be interesting to see how DMP approaches it.

Standards initiatives in DRM have needed to address interoperability for several reasons.  The market has advanced to the point at which a small number of proprietary DRMs have taken hold and cannot really be supplanted by open standards.  Consumer complaints have moved beyond overly restrictive DRMs to lack of interoperability among them. 

Yet interoperability is a difficult problem to solve.  As the lack of interest in MPEG REL demonstrates, interoperability of rights expressions is not enough.  It is not even enough to map the broader semantics of DRM systems to one another, as Coral (and presumably DMP) are trying to do; and it is hard enough to map the mechanics of encryption and authentication schemes to one another.  As InterTrust itself has noted, interoperability cannot be solved by standards alone; the problem can't even really be solved by technology alone.  It will ultimately be up to rights holders to determine what content transfer operations they will allow. 

Yet the trend among media companies now is to (cautiously) embrace interoperability for consumers.  As this trend increases, technology vendors may emerge as the ones who are limiting consumers' options.  Not only may proprietary vendors like Apple emerge as the new digital content bogeymen, but entire market segments may annoy consumers in their attempts to form axes and compete with one another: for example, a Marlin-based portable music player segment will surely find itself more and more directly competing with an OMA-based mobile device segment as the differences between the two types of devices become fewer and fewer. 

In other words, the need for DRM interoperability standards will persist for some time.  Let's hope that the standards initiatives finally catch up to or anticipate the market in 2006 and beyond.

Get DRM Watch Newsletter

Click here to subscribe to DRM Watch