In order to be useful, a protocol for the transfer of value for the internet must be on most desktops, in most browsers, in many web servers, yet many the existing proposals involve proprietary software, and many of them involve a banking monopoly for the issuer of the software. It seems unlikely that many other financial intermediaries would wish to participate in such a grotesquely uneven playing field.
Many of these proposals require the issuer and protocol developer to be intimately involved in every transaction, with the notable and important exception of the IBM and SOX proposals. They require everyone using cash to buy or sell on the net to have an account with the issuer and protocol developer, and the protocol developer's copyrights give it a total monopoly of net money issue.
For a protocol to become widely used, a wide variety of players must play along. There has to be something in it for everyone.
This is the most advanced, best funded, and most widely deployed transmissible form of cash. It is a smartcard, a tiny tamper proof computer the size and shape of a credit card, that keeps a truthful account of its owners balance; designed to be used in a wholly cashlike manner, rather than a credit card like manner. If it is indeed tamperproof, and if it comes into wide use, it will trivially solve the problems of cashlike transactions through the internet and many other problems as well. Its great weakness is that by design it must be deployed through the banks and with the cooperation of the banks, and the banks are somewhat ambivalent. Many banks all around the world have signed on, but most of them are promoting it in a way likely to guarantee failure, making it impossible to achieve critical mass.
Another big problem is that every person who uses it must make a significant investment in hardware, which no one wants to do until it has already achieved critical mass, making the critical mass problem even more severe than with a software only solution. People are reluctant to install new software, but they are even more reluctant to install new hardware.
Some people, myself among them, doubt that the cards are truly tamperproof. A tampered card could appear to be always full of Mondex money, no matter how much the owner spends, generating perfect counterfeit Mondex money and placing it in circulation.
There are numerous smart cards and smart card proposals and smart card alliances in Europe, and they are moving towards interoperability. Other programs also claim to be the leading smart card program, but right now the bottom line is that smart cards are not particularly useful for buying or selling stuff over the internet, despite numerous grandiose announcements by alliances representing vast amounts of wealth and market power. Other such alliances have come and gone.
Despite curious power of this relic, and the comparative success of this scheme (or rather lesser failure), this form of net money has as yet failed to achieve anything like critical mass, and its success is merely comparative. It is used within narrow circles, a group of inconveniently small islands that have yet to merge.
IBM of course, has no plans to give away the software, but if the protocol is truly open, as promised, they will not achieve or seek any monopoly power over internet money issue, since any other software writer will be able to write to the IBM standard without needing permission, thus limiting IBM to merely huge profits should the scheme succeed.
Instead of having to change all browsers to handle a new cash protocol, we would simply write the protocol in Java, and the browser would download the protocol implementation from the web site issuing the instruments of exchange.
This Java extension seems to me to have been designed by an excessively large committee with incompatible and conflicting desires, but if the later versions turn out to be usable, and come to be widely distributed with browsers, this will greatly reduce the critical mass problem.
Millicent has some big money partners and support, such as KDD, and is going ahead with its plans. It is targeting uncontroversial material such as dictionary searches and pictures. In their trial, they claimed total turnover of about ten thousand dollars, about as much as e-gold claims to turn over in two or three days. They employed only very respectable vendors in their trial. If their product is limited to highly respectable vendors, it appears to me that it is inherently incapable of achieving critical mass.