Two recent events have underlined the importance of the basic distribution methodiology implemented by freenet and e-donkey.

Firstly the confirmation of what many of us suspected, namely that the music companies are now employing external parties to sabotrage filesharing networks by floodiing them with dud files that masquerade as the works of 'popular artists'.

Quite simply systems employing cryptographic hasjoing are not susceptible this type of attack, as the hash keys are unique, so that as long as you receive the pointer to the key from a trusted source there need be no concern about wasted downloads.

Secondly, the news that Ranger are being employed to scan network traffic by the MPAA to identify peers sharing movie files, connect their IP to an ISP and then send threatening letters to force compliance indicates that industry has now begun to identify the weak points of second generation networks. The freenet/edonkey model is robust to such detection attempts because of both the lack of an intuitive naming protocol and the modular assembly of the file's components where a full download never takes plave from one peer to another.

Decentralised databases using encryption or hasjing is the way to go. We just need some nicer user interfaces and more GPL's code.