Yeah, we left out metadata at first and then sort of forgot about it. Since your looking at the code:
I've been ripping the interface apart quite a bit while you were gone, and I have a whole load of changes to commit tomorrow morning as well (though that should be pretty much the end of it). I settled in on a new design where all the code that is executed in the node (with a few exceptions, ConnectionHandler threads, the Ticker, Conduits, the ListenSocket thread) is executed by being handled like if it was a received Message. To do this I created a MessageObject interface which messages implement, and is anything that can be sent to the MessageHandler. I rather happy with this, as it is much nicer then the old Callback interface that was being used all over the place to the execute code lumps - especially with all the handeling of the streams, I had to use a bunch of different locks all over the place, now the message id synchronization in StandardMessageHandler takes care of it all. The code in the CVS still has a bunch of Callback objects, but I ripped all of them out today, and I am almost finished patching it back together without them (Node done, just the clients left). Does this mean we'll be getting fproxy back? On Sun, Aug 27, 2000 at 06:06:38PM +0100, Theodore Hong wrote: > ok, I'm trying to catch up to the current code and I'm confused -- is > private metadata currently working or not? InsertClient seems to ignore > the -metadata switch, and there's this comment: > > rt = bc.prepare(htl, // hops to live > in, // data > plaintext.size(), > in, // no meta data > > what's the deal? > theo > > _______________________________________________ > Freenet-dev mailing list > Freenet-dev at lists.sourceforge.net > http://lists.sourceforge.net/mailman/listinfo/freenet-dev > -- \oskar _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
