I'm not sure that Paypal has met the needs of any enduser yet, so I'd question whether it "succeeded." The rest of what you say seems to be to the point that modern "IP" laws tend to discourage creativity. To some extent this may be so, as with interminable term extensions, but the most immediate discouragement lies in the DMCA, not an "IP" law at all but a grant of rather unprecedented monopoly power, irrespective of creative labor, among other items of note. Macon -----Original Message----- From: Tim May [mailto:timcmay@got.net] Sent: Thursday, July 31, 2003 4:35 PM To: cypherpunks@lne.com Subject: Re: Digicash Patents PayPal apparently met the needs of its customers, which was for a low-tech, low-security, no anonymity online payment system. While I'm not saying I predicted it, neither is it surprising that something like it succeeded. (Members of this very list had some discussions with the guy who started PayPal...I wasn't in on this, but I gather that he used some of the ideas, but not the high security/untraceability ideas...just the online payment low-hanging fruit part. The same is true of EBay, by the way, where some of our crowd developed an online system very much like what E-Bay became, but several years _after_ the AMiX system. C'est la vie.) In any case, there will be many successes and failures in Internet-related business. This list is about certain kinds of these systems, but not really about "online payments" in their general form. I'm not saying folks can't or shouldn't talk about Mondex or PayPal or FastTrack, just that they have little to do with the obvious themes of the group.

The beauty of a marketplace is that many different parties get to try every which way of satisfying a need. Most will fail. Even the first several attempts can fail, disguising a real opportunity as a guaranteed failure.

Software patents and the difficulty of "metering" usage has made this kind of experimentation, this kind of evolutionary learning, much harder to do. For example, when Intel sold the 4004 microprocessor 30 years ago, it owned a bunch of patents and trade secrets about how the chip was made, what it's design was, etc. But it didn't force potential customers to lay out their business plans, to sign "no compete" clauses, and it did not have to devise complex ways of knowing how many chips a customer was using. The chips metered themselves, metered the patents and other IP, and Intel could sell them to guys in garage shops, companies in Boise and Peoria, and even to distributors to resell and resell. Such is not the case with, say, the Digicash software. Since Chaum, then the Canadians, then the Indians, then First Data, etc., wanted to maximize the payoff and "get a piece of the action," they could not simply sell the technology, even bits of it, to guys in garages and people with bizarre and untested ideas. And such software is usually not sold to unidentifiable customers. Digicash (or its descendants) will not sell one copy of its core technology to a company without draconian safeguards and audits. This is part of why a paper trail back to the users of various technologies exist when above-board licensing is used. (And why many of us would obviously then favor simply _taking_ the technology, ripping it off. This cuts the paper and liability trail. Yes, there are minor issues with "theft of intellectual property," but this is mostly smoke and mirrors anyway. No one in the Western world seems to think ideas in general are patentable, so how did RSA get patented? We've had this discussion many times over the past 11 years. And more.) This is replicated all over the digital landscape, where software packages have complicated licensing schemes and where vendors want to see only "staid and conservative" business plans. No Digicash for BlackNet, in other words. This whole phenomenon has dramatically slowed down exploratory developement and weird, new ideas. A couple of guys in a tilt-up in Silicon Valley simply cannot just go down to Fry's to buy some parts and try out some ideas, not in this world of licenses, audits, lawyers, and generally pointless efforts to meter usage. Which is why doing things without benefit of any patent licenses is the best strategy. That this also requires no nexus, no trail back to a corporate office, is of course part of why it's the best strategy. --Tim May