gtoal writes:

One thing this shows, even if the application of the technique turns out to be too difficult to be practical, is that Dorothy Denning's evaluation of the design was worthless. That team should have found this themselves. No wonder she was trying to play it down in the NYT.

Dorothy's group was highly worthwhile; you just have to understand the objectives :-) It produced a fine interim report describing how strong the strong part of the Clipper system was, making it politically easier for the NIST to declare their stuff to be a standard blessed by experts. The evaluation of the whole Clipper system, beyond just the SkipJack algorithm itself, will supposedly be in the final report, which will supposedly be out Real Soon Now, according to a conversation I had with her in March. I suspect that report is either being hastily revised (:-), or else they had already discovered it and were stalling to see if anyone else had, or they had seen Matt's draft and have already written the revisions but were waiting to see if he could get it published. On the other hand, maybe they've discovered one of the other N technical weaknesses in Clipper, or had given the system a negative review and aren't getting support from NIST to release it, or some other amusing variant. Meanwhile, way to go, Matt! Both for doing the analysis, and for getting the material to the press effectively - and also way to go John Markoff! (Out here in the San Francisco Chronicle, it only made the bottom half of Page 1 Column 1; haven't seen the San Jose Murky News yet.) Bill