If the government really were to commit to removing export controls on all civilian cryptography, and to ensuring that Clipper remains truly voluntary, then I could see how the DPSG compromise *might* have some merit. With or without such an agreement, Clipper is a non-starter -- as long as better alternatives are readily available. This is already true for general purpose computers. People can easily run whatever encryption software they like on their PCs, whether the NSA, FBI or anyone else likes it or not. Unfortunately, it is NOT true for voice services like digital cellular. Although they contain computers, digital phones are not as easily reprogrammed as PCs. And even if they were, you still need the cooperation of the carrier to decrypt your call at the base station (unless, of course, you encrypt end-to-end, but then you need the right kind of data service from the carrier, plus a compatible secure telephone on the land side of the call). In theory, at least, under this "deal" the cellular vendors would be free to support either Clipper or some other, better encryption scheme, without fear of export controls destroying much of their market. In reality, of course, the cellular carriers and manufacturers are large and conservative enough (and do enough business with the government) to be easily browbeaten into installing only Clipper, if indeed they install any real encryption at all. Indeed, the NSA hardly had to lift a finger to browbeat the industry into installing a trivial "voice privacy" scheme in TDMA cellular that can be broken by any undergrad CS student in a few minutes. So I'm not worried about the effect of this deal on general purpose computer applications; it may even help, by getting rid of export controls. But the big loser will inevitably be voice privacy. Phil