Hi... Recently I'm often seeing problems in Phex to inflate Bearshares QRT. I usually only see this problem when a connection gets interrupted during QRT ...
... QRT. I ... during QRT ... recovering. I have not ... Bearshare ... I am very confident this is a problem on your end, or else we would have noticed it a...
... Yes could be, at least I'm only seeing it with Bearshares and I'm sure it was working when I tested against older releases. ... I'm very sorry if I have...
... would ... sure it ... not ... is that ... data ... english term for ... need to ... unparseable ... Perhaps there is a bug in BearShare where we corrupt...
... safer to ... Query... This algorithm will work provided that the remote UltraPeer can accept .... as multiple patches. A clever algorithm may be used by ...
... such as ... value ... the ... PATCH ... Look, I really dont give two hoots about "recovering". Don't spread bad advice around on this group. When a host...
... No. If the guy writes a buggy servent that sometimes loses sync on QRT messages (or whatever) it receives, I'd rather have him resync within the same...
... What if there is a bug at the other end which causes a message, perhaps a QRT patch, to be shorter than it ought to? The parsing will fail and the good...
Ok... first I'm not trying to resync or at least I'm not having any kind of special algorithm that in error case tries to find how and where to sync the ...
... fine with ... wrongly ... No one was offended. I see what you mean about dropping messages - I misunderstood you. Your behavior is perfectly legitimate....
The message below after the '---' was sent on the gtk-gnutella-devel list. It seems to demonstrate that LimeWire and gtk-gnutella don't use the same physical...
Raphael Manfredi
Raphael_Manfredi@...
Sep 6, 2003 7:06 am
17216
The QuerySpeed field is a old legacy field, and as such, it was defined to be little-endian as documented in Gnutella 0.56. We used the bit #15 of this...
I'm pretty sure we did this correctly, according to our interpretation of the MinSpeed spec. The MinSpeed spec *implicitly* states that the new interpretation...
Looking at your e-mail again, I think this is the cause for confusion. As I wrote the MinSpeed document, bit #15 refers to what you seen on the wire, not what...
... Make that "protocol design" classes ;-) Never use little-endian when designing protocols. So, I take it that bit #15 is actually bit #7, right? You must ...
It's neat that LimeWire sends a 416 when it has part of the file, but not the requested range. However, in this request/response, I was hoping to get Alt Locs...
I just checked out the code, and it seems OK. However, it's possible that the uploader you connected to did not have any alternate locations to report. at that...
Swapper uses 2 distinct bytes for this information. The Query Packet is persisted with the offsets: protected sealed class Offsets { private Offsets(){} public...
Just a thought: There are not enough GWebcaches, my cache gets thousands of LimeWire 3.4.7 requests which make up more than 80 percent of the requests. I think...
Hi guys, I'm doing a research project to identify the searching habit of P2P users. Is it possible to distinguish queries from diferent users by looking at the...
Jason, I don't know this .NOT stuff, so how are you writing this stuff on the wire? FlagsA is the 0th byte in your example (which is the *correct* way ;) ),...
Hi Frank, Generally I'd say you can't use a query GUID to distinguish anything about a user (besides knowing which way to route hits for that query). In the...
Hi, We try to encourage our Open-Source community to run GWebCaches. Check out the top-right of this page: http://www.limewire.org/ Thanks! Susheel ... ...
You can set up your node as an ultrapeer an you see the query of your leaf. You can do your stat with that. Take care if you don't do that way. The queries are...
... confusion. ... on ... my 'Gnutella ... when ... READ ... would ... a ... will ... Raphael, Susheel is in fact describing the minspeed field as a big-endian...
