keywords: Crypto-policy, digital signatures, Anti-Horsemen, secure-DNS, DOJ-web-prank, Ayn Rand, hydrazine Will Day <willday@rom.oit.gatech.edu> responed to Frank Stuart's cogent and insightful posting on the DOJ hack and a possibility of using the tawdry event to some advantage.... If you missed Frank Stuart's original message, much of it is in Will Day's response. I have made some tedious reformatting here with prepended 'Day' and 'FS' marks to separate the two and clipped some ephemeral lines. Will Day signed his message with Pretty Good Privacy, but I clipped the signature from that off, also, having already hosed it up with my editing of line lengths, &c., so that you couldn't check it anyway, even if you wanted to. The gist of Will Day's question is how can an argument be made that supports the greater availability of strong crypto from the fact that someone swapped spoofed-up (and inane) pages for the DOJ's own on the DOJ's own web site? Let's start by trying to imagine a future world of geodesic networks based on robust strong protocols that are ubiquitous, easy to use, and embedded in the chips of even the most mundane devices. In this world, authentication of data, such as web pages, is required (or at least it could be) for every packet we receive. For data retrieved from remote sites we may require multiple signatures, and certificates signed by someone in whom we have placed a high reputation value. When (or if) freely available legal authentication technology becomes ubiquitous and transparent, we will be able to use it for even low risk, trivial, applications - like remote controls for televisions. We sure as hell won't have to stay up at night worrying that some punk is going to change any of the bits on our web server and that such changes might go undetected by our code-signed auto-gunsel. We are a long way from crypto-Chaumian-utopia, and it appears as though the US federal government wants to make it harder for us to get there. (As an aside, I think the Clinton administration gets more of the blame for this than they should, since it was entrenched policy before they got out of Arkansas. I think the roots of Clipper and GAK are back in Reagan's stint, about the time of NSDD 145. But then, the current leaders aren't doing much to develop meaningful *public* cryptography policy, so they have to take the heat now.) Strong crypto helps people protect what is theirs. That is part of what Frank Stuart is saying, and he's right. That is a good point, and it deserves some attention. Strong crypto can help big powerful organizations like the DOJ, and it can help regular folks, too. Frank Stuart's next point is a beauty - at least the DOJ site that was attacked didn't have copies of everyone's secret keys stored on it. We all may know that even if GAK were ever passed, no one would be stupid enough to store the keys on a web server sitting out on the big bad Internet, but the cleverness of this spin-vector is that it raises the issue of GAK-riskiness in the context of DOJ's computer security. The last point is that another law on the books isn't needed, and wouldn't be effective anyway. I have gone on way too long already, but to sum up, the DOJ being abused may serve to help the cause, if the proper angle can be seen - and Frank Stuart is off to a good start. The specific answer to Will Day's question, which is a good one - how does crytpo protect your poor little Linux box in the corner that serves up web pages? - is left up to the student as an exercise. -- Day>A short time ago, at a computer terminal far, far away, Frank Stuart Day>wrote: FS> however, I think those in a position to do so should start with FS> the spin control. Some suggestions: FS> FS> The fact that even the U.S. Justice Department is unable to FS> adequately protect it's own site from intruders underscores the FS> need for widely-available strong encryption. FS> FS> While this is certainly a major embarrassment for the Justice FS> Department, at least the mandatory "key escrow" program the Clinton FS> administration is insisting upon has not yet been implemented; FS> no private citizens' data appears to have been compromised this time. FS> FS> It's doubtful that a new law or government bureaucracy would have FS> prevented this from happening but it's entirely possible that tools FS> such as strong encryption could have. It's ironic that the U.S. FS> Government is focusing on the former while fighting use of the latter. Day> I understand how it affects their claim for the security of escrowed Day> keys, but I'm afraid I don't follow the other argument. How would Day> the wide availability of strong encryption have helped prevent the Day> breakin? How would encryption in general prevent breakins? I'd Day> love to use this as an argument for strong encryption, but I Day> don't see how it really applies.