Steve Furlong writes:

Does anyone know of a reference implementation for Stefan Brands's digital certificate scheme? Alternatively, does anyone have an email address for Brands so I can ask him myself? (I haven't gotten anything back from ZKS's "contact us" address. But I don't know if Brands is still at ZKS.)

I will give you some pointers, but in exchange, Steve Furlong, you are commanded to fix your cypherpunks archive at http://archives.abditum.com/cypherpunks/index.html. Despite your comments on the web page as of February 13, none of the year 2004 links work. You were already pointed to www.credentica.com, but it's not a very informative page, is it? Just a logo and a mailto. Well, there is at least some "hidden" content, at http://www.credentica.com/technology/technology.html. This has a paper on Brands' technology and some information on his book. In late 2002 I saw email from Brands with two return addresses, brands@credentica.com and sbrands@videotron.ca. As for implementations, there was library called NCash created by Swedish student Niels MC6ller, described briefly at http://www.lysator.liu.se/~nisse/hacks.html: "For my master of science thesis, I implemented an offline digital cash system invented by Stefan Brands. I haven't done anything about it since I got my MSc degree, but the information about it is still available." Well, the information isn't available, not there, anyway. But if you hunt around, you can find an announcement about NCash at http://www.privacy.nb.ca/cryptography/archives/coderpunks/new/1998-02/0008.h... and the source code plus a technical paper are at http://munitions.vipul.net/dolphin.cgi?action=render&category=09. The good news is that the source code is available; the bad news is that it is written in Pike. No, I'd never heard of it, either. Apparently it is some Swedish language that was all the rage back in 1997. However it seems to be a pretty straightforward language, kind of Pythonish with a C like syntax, so it might be feasible to port it to something more modern, probably Python. Or you could get the latest Pike interpreter and see how well they've maintained backwards compatibility all these years. According to the announcement, you'll also need the Pike crypto library, but my guess is that it's been incorporated into the Pike package by this time. I'm sure you're aware that Brands credentials are patented up the wazoo. Like Chaum, Brands prefers to publish his work in the pages of the U.S. Patent Office. That's why they're both so rich.