Felten spills the SDMI beans
No arrests, yet....
USENIX 10 It came off without a hitch. Princeton University Professor Edward Felten, who led the team of researchers which successfully cracked the SDMI challenge, delivered his group's findings at the tenth annual USENIX conference in Washington Wednesday, and was not arrested.
The SDMI watermarking technologies his team examined were remarkably easy to attack, he said. If their purpose is to prevent professional pirates from copying digital media, then the schemes are a failure. If their purpose is to 'keep honest people honest', then they're ridiculously overdone. "A 'no-trespassing' sign is adequate for that," he noted.
Felten added that the basic approach of using watermarks to enforce content access controls was foolish from the start, though he did say that fragile watermarks could be handy in detecting whether a file had been tampered with, so the technology isn't by any means useless.
He cautioned that regulations which might punish the release of signals-processing research to accommodate the legislative convenience of a single industry could have broad, unintended consequences. Seismology, for example, depends upon such research to fine-tune its equipment and techniques, he observed.
The controversy over his presentation began last year, when the Recording Industry Association of America (RIAA) sent him a nastygram hinting that he could be sued under the Digital Millennium Copyright Act (DMCA) for making his findings public.
He therefore canceled his plan to release the paper at the fourth annual Hiding Workshop back in April. The Electronic Frontier Foundation (EFF) saw in him a living example of speech chilled by a draconian law, and has since taken up his cause.
Felten filed for a declaratory judgment in court, seeking permission to deliver his paper at USENIX unmolested. But the RIAA has since denied any intention to impede Felten's presentation, and recently asked the court to dismiss his suit because the controversy no longer exists.
EFF has rebutted, arguing that the case remains relevant, especially in consideration of Dutch cryptographer Niels Ferguson's reluctance to publish proof of his claim that the High-bandwidth Digital Content Protection (HDCP) standard for DVI is flawed.
The group has recently posted a great deal of fresh material related to the suit here. ®