Marc Schneiders wrote: > > Gnutella hasn't been around longer. It was just popular before Freenet. > > Corrected. On a side note, why is this? Do they have more > developers? I haven't really tried it, but it looks like it is there > longer. Probably the interface, really.
As for Gnutella, it was developed by the WinAmp guys, immediately after alpha-release bought by AOL(?) , who withdraw the concept and program at once. That made it popular enough in the media to get some attention. Gnutella has some really nice clients for all OS and in all languages floating around which are getting more and more sophisticated. I guess the protocoll is stable as it is no more developed. As for Freenet: It is still changing its protocol really fast, so developing sophisticated clients would make no sense at all, as it would be like being the rabbit hunting a carrot that is fixed before its nose. Having to change the implementation (and therefore the core of a client every couple of weeks is no good motivation. In my opinion, there should be a protocoll feature freeze, where the protocol is detailed described, to give authors a chance to develop their own stuff. This leads to another lack: description. There are quite a lot of documents concerning architecture, protocol spec, future idas and installation of Freenet floating around. But they are disctributed over all places, partially obsolete and bad organized. What I really miss, is a central document (Freenet Programming Guide), which will describe the Freenet, it's protocol specs and the keys in its most current form. It would be great to have such a thing for Freenet 0.3 to give outsider a chance to get familiar with the architecure of Freenet. Just now, there is no way to get familiar with it except for studying Java code. Unfortunately there a lot of programmers outside, willing to develop a client but who have no idea of Java whatsoever. These are wasted possibilities in my opinion. Another point is the organization of Freenet. There is a core team consisting of Ian, Scott, Oskar and Brandon as I guess. It is really hard for outsiders to support them or to start own developments (clients, web proxies). Have a look at Jabber.org. They provide a means of organizing projects yourself and adding them to a projects page. I guess that would help co-ordinating the work happening on non-node developments a lot. Maybe using a kind of bugzilla (see at mozilla.org) would help co-ordinating and getting an overview about the work done just now. A last comment, about negative freenet stuff I noticed. But let's first quote this: > > Hehe, you said the mailing lists were full of enthusiastic users eager to > > help. Heheh. > > You proved it :-) Despite the fact that the mailing lists are quite chaotic (is client-dev and tech used at all? why is 'general' in reality a completely off topic chat list), there are nearly no one that support potentuial Freenet users. I counted so far (including me) two Germans and one dutch trying to support the (mostly Windoze) newbies. I have to admitt they often ask bullshit, but that's still our task to support them as much as possible and maybe hack together a FAQ, answering the most often stupid questions. Which in turn leads to the really last topic: Windoze I know you guys are not using it and (partially correct) don't like it. Nevertheless Freenet is never going to be as popular a Napster or Gnutella if it doesn't offer some nice looking, easily to handle GUI clients or web proxies. Since Freenet 0.2 I "never" got a Freenet node on Windows to run smoothlessly. Is *anybody* succesfully running a recent node on Windows? Please contact me, I'd like to share the experience and I am no Newbie. You'll need a running windows node to get Windows clients, so it *is* important. Although there are some nice tries to a GUI client EZ-Freenet, FNC, they are not yet something I'd recommend an average Windows user. As you UNIX guys aren't able to program some nice Win GUI client (no offence intended), you should at least provide a central, well-organized description of Freenet nodes, and it's protocol specification (maybe including a sample conversation between node/node, node/client). OK, glad to having said this, now flame me Sebastian _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
