Welcome to the Gnutella Developer Forum. Some suggested topics of conversation just to get things rolling: * Why do Gnutella downloads fail? Is it a case of...
2
info@...
Jan 17, 2001 3:09 am
Hello everyone, this is Vincent Falco from Free Peers, Inc. I would like to introduce myself, I'll be participating in the forum and providing my thoughts and...
3
info@...
Jan 17, 2001 3:15 am
... Incorrect configuration of the software is a major problem. Another problem is the connection backlog limit for certain Windows operating systems. The...
4
Greg Bildson
gbildson@...
Jan 17, 2001 4:56 am
Hi all. Greg Bildson from LimeWire here. Our original goal was to have a private forum for active Gnutella developers without distractions. I don't know how...
5
Greg Bildson
gbildson@...
Jan 17, 2001 5:01 am
So, the number one problem with the Gnutella net right now seems to be download failures. For reasons that I won't repeat at this time (see...
6
kane@...
Jan 17, 2001 12:09 pm
hi, this is josh from toadnode i've got just a few suggestions for the forum that might help eliminate some confusions. is there a way we can list the clients...
7
Steve Bryan
steve_bryan@...
Jan 17, 2001 5:35 pm
... functioning ... Hi, this is Steve Bryan from the Mactella development effort at cxc. I'd like to report that I have witnessed the push capability working ...
8
info@...
Jan 17, 2001 6:53 pm
File indexes that change can be a problem. There are a couple of solutions to this. One technique, is to check the filename on the request against the index....
9
swabby@...
Jan 17, 2001 8:12 pm
Hi all, I'm John, Gnucleus is my clone, and I'm interested in working with everyone to keep our clients compatible. The number one reason downloads fail is...
10
crohrs@...
Jan 17, 2001 9:50 pm
... Here's the LimeWire proposal. Let the bytes of a GUID G be labelled G [0]..G[15]. All bytes are unsigned. Let a "short" be a 2 byte little-endian**...
11
info@...
Jan 17, 2001 11:46 pm
... I don't consider that a failure. Usually when you get a busy response, this is a good thing it means the computer is responsive. At that point the user...
12
Christopher Rohrs
crohrs@...
Jan 18, 2001 12:05 am
... from ... do ... I realize my proposals (2) and (3) contradict; ignore (3). The high 4 bits of the checksum should be reserved for future use. Current ...
13
Oso
osokin@...
Jan 18, 2001 12:54 am
... Hey, Christopher, you think it is easy to write such a long reply? :-) ... Hmm.. A first day of the forum and already we have a protocol change proposal -...
14
info@...
Jan 18, 2001 2:15 am
... The result doesn't have to be bulletproof, because the penalty for collision is that the old client gets a search result. No big deal, it will be quite...
15
Serguei Osokine
osokin@...
Jan 18, 2001 3:19 am
... Oh, that's right - I sort of assumed that I'm still on the gnutellang forum and forgot that this forum was created from the different mailing list and that...
16
swabby@...
Jan 18, 2001 3:59 am
From my understanding, this forum should be dedicated to our client's compatibility with each other. We should not be modifying the protocol, especially here,...
17
Vincent Falco
info@...
Jan 18, 2001 4:45 am
... Yes we should. I would like to see some cosmetic modifications to the protocol so that some vendor-specific data can be added and passed in an agreed upon...
18
Greg Bildson
gbildson@...
Jan 18, 2001 4:57 am
Okay, we have been hinting at a proposal. I don't think we are ready for it unless we can agree on the magnitude of the problem though. Users are having...
19
Vincent Falco
info@...
Jan 18, 2001 5:17 am
... side ... And the best part about 3) is that it is the easiest to code with the least risk of introducing a defect. It can also bring back those users that...
20
GDF Sysop
gdf_sysop@...
Jan 18, 2001 7:16 am
... so ... It appears that after the first day or so, there are representatives present in this forum from the following companies/products BearShare Clip2 ...
21
Clip2
gdf@...
Jan 18, 2001 7:43 am
... is ... several ... Steve, thanks for your comments. For the sake of those (including us) who may not have heard it elsewhere, can you relay the reasoning ...
22
Clip2
gdf@...
Jan 18, 2001 7:47 am
... Is this guaranteed to happen, or is it just a possibility? In other words, if there are no changes to the upload directories, why does the index differ /...
23
Ben Houston
ben@...
Jan 18, 2001 7:52 am
... Why not let the old clients connect but simply return to each of their search requests just two hits such as "Gnutella v0.56 clients are not supported" and...
24
lambla@...
Jan 18, 2001 9:36 am
Hi guys, I'm happy to see that most of you work on this forum for immediate enhancements of the gnutella network and are interested in working on gPulp for the...
25
crohrs@...
Jan 18, 2001 4:03 pm
... serving ... Also, if your download and shared directories are the same, Gnutella 0.56 does not share downloaded files until you rescan or restart. This...
26
Serguei Osokine
osokin@...
Jan 18, 2001 4:48 pm
... Sounds like a good idea to me. Of course, I'm NG/gPulp-biased to begin with... :-) Best wishes - Oso....
27
Serguei Osokine
osokin@...
Jan 18, 2001 5:20 pm
... Okay, I'm not Steve (maybe he has other reasons), but for me personally, the reason I thought the original idea was to broadcast the PUSH is that it has...
28
Serguei Osokine
osokin@...
Jan 18, 2001 5:40 pm
... Oh, absolutely - perhaps I did not make myself sufficiently clear. In fact, I rather like these extensions of yours myself. I was arguing against first, a...
29
Greg Bildson
gbildson@...
Jan 18, 2001 6:02 pm
We wrote up some of our thoughts on what makes for healthy behaviour of a Gnutella servent. It is more focused on medium term changes but perhaps worth a...
30
Greg Bildson
gbildson@...
Jan 18, 2001 6:09 pm
Hi Sebastien. We are sensitive to your wishes. I think us developers would like to think and talk for a while before we start coming out with press releases. ...
