The collapse of many Internet dotcom companies reflects the deep unease of many consumers at trusting their credit-card numbers to the insecurity of the Internet. How, then, to bring alive the Internet's commercial potential? Micro-payments have somehow not taken off. Daedalus reckons that a consumer will trust to the Internet what he might lose on a bet — maybe US$50 or so. Rogue sites that charge more than this, in one or many payments, must be discouraged.

DREADCO Financial Services is therefore inventing a special low-limit card or cheque for Internet use. Their small limit cannot be overexploited, so the risk is low. It may feature an exact cut-off negotiated automatically with the company beforehand.

Thus the great dream — that of our own personal library from the Internet — would come true at last. The big publications, like Nature and The Times, will continue as before, for those who need them. But people who visit their websites, and those of all other publications too, will note those articles that seem interesting, and amass them for small proportional payments, controlled by the DREADCO card. A software 'gopher' could do this or haggle over it all the time. It would store its haul on a hard disk, a vast improvement on a stack of old papers. Occasionally you might glance at the screen to see what you had caught, and you could print out your results. Those for whom citations are magical, in support of personal glory, would relish the chance to amass them. The whole media scene would be one vast newspaper or magazine, with perfect indexing, no wasted time, and maybe the deletion of items no longer of interest.

Why doesn't this happen now? The academic media are deeply divided about how freely to release their papers onto the Internet, or what time delay to adopt before doing so. But once most publishers had taken the plunge, the worth of any delay would soon be shown by market forces, clear to everyone. Daily newspapers would lose value in a few days, and weeklies in a few weeks; academic journals of more lasting value should hold that value longer. Daedalus reckons that when a journal has amassed more citations than it quotes itself, its value starts to decay. But journal publication would remain useful, partly because editorial scrutiny weeds out the nonsense, and partly because anyone specializing in the area is better off subscribing to the journal.