you are talking about a 140k jar file containing the index too ... do u mean everone who wants to query the key index has to have it entirely locally ?
the applet would just search within a file actually that's it ? but u have to download the new version of the index each day (or each how many u want ...) ? how can u call it a distributed search ? tell me David ----- Original Message ----- From: "Brandon" <[email protected]> To: <devl at freenetproject.org> Sent: Monday, April 09, 2001 11:08 PM Subject: Re: [freenet-devl] distributed search > > > Not a bad idea after all. I don't have a clue how to implement FProxy or > > node plug-ins though - is there a (stable) API? > > A node plugin uses the standard Java Servlet API to deal with connections > and either Client or SimplifiedClient to talk to the node, depending on > your preference. FProxy is a good example of a node plugin. > > There isn't a FProxy plugin API as of yet as it hasn't come up. My plan > was to have pluggable MIME type handlers. So if, for instance, you wanted > to give a key index summary file a particular mime type then you could > configure FProxy, on receiving a file of that type, to format it as HTML > rather than streaming it directly to the browser. > > > > That's more sane then a Java applet as Java applets trip the security > > > filter. > >
