Re: Re: Search by urn ??
Bill Pringlemeir <[email protected]>
| Newsgroups | gmane.network.gnutella.devel |
|---|---|
| Message-ID | <[email protected]> |
--- Arne Babenhauserheide <[email protected]> wrote: > El Friday, 21 de March de 2008 21:16:21 [email protected] escribió: > > Quoting Bill Pringlemeir <[email protected]> from > ml.gnutella.dev-forum: > > :The bootstrap is really lacking. Really people have to have the LW source > > : and the message format document. There is really not much else that is > > : helpful. > > > > I don't want to look at LW's sources, or any other servent source code for > > that matter, to be able to implement something that is meant to be > > "standard" on the Gnutella network. We need to have good documentation. > > We also need to have people creating the documentation. [snip] > But we can't just look at them and be angry that they didn't provide it for > us for free. There are several points against this. One being that the LW source tree seemed to have several versions of Mojito. For a C or C++ developer, this really makes it hard to figure out the code. I think I needed a Java class browser to figure out which was the real mojito. Another issue, Is mojito a java kadmelia library or is it a protocol to be used by gnutella? It is apparently confusing for it to be both. The other thing is the removal of the URN support. If URN requests are still forwarded, then I don't think this is problematic. If they are discarded then that really co-opted the network. The final point I was trying to make is that by not helping other developers, LimeWire users don't benefit. I think that many gtk-gnutella clients run with very high out degrees and longer up times. If gtk-gnutella peers supported the passive kadmelia mode, gtkg ultras could provide active LimeWire leaves with DHT boot strap nodes. It is also most helpful for LimeWire to say, we are too busy, you are on your own. If this was known before hand, more time could be spent reverse engineering the protocol. I also agree that other details to gnutella might be under documented, but finding alternate sources is a pretty key detail to a p2p protocol. fwiw, Bill Pringlemeir. ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs