Florian , I don't expect you to agree, but it may be worth reading anyway :-)
>The universal client is not secure enough for most applications. The universal client is already a fact. The entire Internet depends on it. What you are saying about non-repudiation may be correct, but I don't think this is what the primary goal is. Signatures should rather be seen as "OK buttons on steroids". Do you think that OK-buttons are better? Because they give you a better possibility to deny that you did this and that? Sure, but this may work to your disadvantage as well. With a web signature capability you get - A well defined "acceptance process" for the user - A "signature view" for the user to look and decide upon - Strong authentication in a single integrated and cryptographically secured step. In fact, for several use-cases, including 3D Secure payments, this step is the only step needed by the user, which means that there are also usage advantages compared to authentication + OK. Note: WebSign (which already have MILLIONS of users) is driven by governments and banks who think that OK is a bit primitive. At least when they with big costs have deployed PKI. If you are actually referring to the security in some operating systems, I would say that authentication is a MUCH bigger issue since there is no way you call "roll back" an incorrect authentication. Therefore I don't see that signatures add any particularly nasty problems outside of the fact that some people do not understand that general purpose on-line signatures like provided by "signText()" and upwards, are essentially useless unless the relying party is trusted. That is a limitation that I don't think is a problem for the governments, banks, and similar institutions that have on-line signatures on their agenda, as they consider themselves as trustworthy. Who wouldn't? For supporting cross-organization workflow (we are not there yet..), signatures are more or less required for data integrity purposes. A "somewhat related" and very interesting link http://software.silicon.com/security/0,39024655,39155596,00.htm shows that even e-governments can be pretty creative these days! Anders ----- Original Message ----- From: "Florian Weimer" <[EMAIL PROTECTED]> To: "Anders Rundgren" <[EMAIL PROTECTED]> Cc: <mozilla-crypto@mozilla.org> Sent: Tuesday, January 24, 2006 15:38 Subject: Re: The Browser Digital Signature Riddle * Anders Rundgren: > Somewhat surprising, the people who seem to be the least aware of > these efforts to transform the ubiquitous Internet browser from being > a "Universal Thin Client", to become a "Universal PKI-enabled Thin Client" > are actually the browser vendors and W3C! > > Comments? The universal client is not secure enough for most applications. As a result, the purported non-repudiation part of such signing mechanisms is potentially harmful to end users. I'm not sure if it makes sese to push things further in that direction, at least as far as browser vendors are concerned. _______________________________________________ dev-tech-crypto mailing list dev-tech-crypto@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-tech-crypto _______________________________________________ dev-tech-crypto mailing list dev-tech-crypto@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-tech-crypto
