[19:36:31] <GargoyleMT> excellent 
[19:36:49] <GargoyleMT> who here has access to files on AS? 
[19:37:58] <fnordpojk> ancient? 
[19:38:55] <GargoyleMT> yeah 
[19:39:14] <GargoyleMT> re: possible tth collisions -
http://shroom.myftp.org/tthbug/TTH.htm
[19:39:33] <fnordpojk> i do. 
[19:40:25] <GargoyleMT> Shroom just PMed me, or I just got back to
this client, so I'm still trying to figure it out.... it could be a
bad client, or a bug in dc++ itself
[19:41:20] <fnordpojk> interesting. 
[19:41:29] <fnordpojk> i heard rumors of tth poisoning, though. 
[19:41:56] <GargoyleMT> I'm thinking it's got a decent chance of being
xayide's client that picks up hashes from searches or lists.
[19:43:38] <fnordpojk> i get the same results from that users share anyway. 
[19:44:09] <fnordpojk> and i'm using fuldc svn, based off 4034. 
[19:44:58] <fnordpojk> strange though that the file list should
contain the wrong tth.
[19:45:21] <GargoyleMT> ah, well no local bug should cause those kinds
of problems, but if they have the wrong hash, they shouldn't be
downloadable either, since they should fail the check at the end.
[19:46:03] <fnordpojk> they probably do. 
[19:46:53] <fnordpojk> i get two users who have several files with that TTH. 
[19:46:56] <Sedulus> and yes MT: about banning xay, that's worth it 
[19:46:56] <fnordpojk> both are using 4033. 
[19:48:03] <fnordpojk> the users with only one file with that TTH are
all running 401.
[19:48:05] <GargoyleMT> Unless he's taking advantage of the fact that
dc++ doesn't check on resume...
[19:48:43] <GargoyleMT> hmm, that's interesting, but the users with
multiple files with the same hash are also running identical versions?
[19:48:48] <GargoyleMT> (or close?) 
[19:48:56] <fnordpojk> yes. 
[19:49:11] <fnordpojk> 4033. 
[19:50:18] <GargoyleMT> hmm, then I'd suspect xayide's handiwork,
especially since shroom says that fulDC generated unique/presumably
correct hashes for him locally
[19:50:33] <fnordpojk> what's interesting is.. 
[19:50:51] <fnordpojk> i queued two files with that TTH from the two
different users.
[19:50:58] <fnordpojk> they both had no slots. 
[19:51:09] <fnordpojk> but fuldc did an autosearch and downloaded
those two files from other users.
[19:51:14] <fnordpojk> and did not complain about tth mismatch. 
[19:51:55] <GargoyleMT> :-/ hmm, ok, I'll go download the fulDC source
to check something...
[19:52:11] <fnordpojk> the source on the webpage is 4033 though. 
[19:52:43] <GargoyleMT> oh, that's ok, I'm just curious whether the
tth checking is intact or not.
[19:52:58] <Trem> it should be, haven't touched it 
[19:53:31] <Trem> but it's changed a bit in sharemanager, might be
some bug there that's causing it to retreive the wrong tth
[19:53:32] <GargoyleMT> ok, maybe there was no tthl?  would dc++ check
against the root hash?
[19:54:00] <fnordpojk> trem, then why do i get that same hash from
users running 401?
[19:54:05] <fnordpojk> on different files. 
[19:54:36] <Trem> ohh 401? then i have no idea, but i only get that on
rar files that's 15 000 000B
[19:55:00] <fnordpojk> these were rar files. 
[19:55:42] <fnordpojk> go to AS, do a tth search for
MECQIGZTRE3NGWROK5ZQ4CHMTFRUTLIFDV4TGSA, see for yourself. :)
[19:55:54] <Trem> i did =) 
[19:56:28] <fnordpojk> the users with single search replies are all
running 401, the users with multiple replies are running 4033.
[19:56:33] <fnordpojk> 'tis very weird. 
[19:57:08] <Trem> check sork's list he's got a folder full of them 
[19:57:49] <fnordpojk> wow, yeah. 
[19:57:59] <fnordpojk> and xay isn't even on that hub. :) 
[20:00:26] <Trem> let's blame him anyway =) 
[20:00:56] <GargoyleMT> well, it was AS he was in here defending the
other day, wasn't it?
[20:01:43] <fnordpojk>  
- [01:48:40] <[VIP]fnordpojk> håller xayide på och leker med tth
poisoning på as
 
[01:59:51] <[Maximum]Mono> Tror inte det. Han postade precis den där
länken (http://shroom.myftp.org/tthbug/TTH.htm) och skrattade gott åt
TTH-buggen dock :)
[20:02:50] <Trem> even if he did he shouldn't be able to mess up their lists 
[20:04:51] <Trem> ahh it's empty fake files, i get 2-40MiB/s when
downloading them
[20:05:16] <fnordpojk> from who? 
[20:05:22] <Trem> from 4 different users 
[20:05:38] <Trem> averaging 500-1000KiB/s from each user 
[20:05:44] <fnordpojk> hm. yeah. i got weird speeds, too. 
[20:06:48] <Trem> i'll see if they match the sfv soon 
[20:09:09] <Trem> bah the sfv was empty 
[20:09:44] <GargoyleMT> if you have the file locally, is the tth legit? 
[20:09:56] <Trem> haven't hashed them yet 
[20:11:04] <GargoyleMT> from what you say, it sounds like an ordinary fake? 
[20:11:45] <Trem> yup can't open the files, they're note valid rar
archives, they downloaded way too fast, empty sfv file
[20:13:04] <Trem> they're in my share now 
[20:13:07] <GargoyleMT> [20:11:32] <[BBB]Shroom> dont know yet. One
file I just downloaded was apperently fake (empty), but the ones I
downloaded before were not...
[20:13:18] <GargoyleMT> [20:10:17] <GargoyleMT> any fakeshare checking
in xdc?  it could just be an ordinary fake file?
- [20:10:34] <GargoyleMT> (seems odd to jump to the conclusion that a
fake file is a TTH bug, though)
- [20:11:21] <xAyiDe> that guy aint faking. Its a friend of mine 
- [20:11:54] <xAyiDe> Ist real files if you download them 
- [20:11:57] <xAyiDe> Shrrom told me 
[20:13:32] <Trem> try downloading them =) 
[20:14:22] <Trem> filled with zeros 
[20:14:45] <fnordpojk> yup. 
[20:14:57] <Trem> so let's blame xay for spreading fake files? =) 
[20:15:12] <fnordpojk> haha. 
[20:15:22] <GargoyleMT> or picking up the hash of faked ones. 
[20:15:44] <fnordpojk> i'm getting those files at 5-6MB/s, fascinating. ;) 
[20:15:56] <GargoyleMT> hmm, so here's the rest of the PM, from the beginning: 
[20:15:58] <GargoyleMT> 19:46:25] <GargoyleMT> will you talk about
your client that picks up TTHes from elsewhere?
- [20:02:24] <xAyiDe> elsewher e= wherE? 
- [20:04:25] <GargoyleMT> you talked about discovering TTHs for files
by looking at lists and search results, instead of actually hashing
the files... you coded that client, didn't you?
- [20:04:38] <xAyiDe> Yepp 
- [20:04:50] <xAyiDe> its xdc actualy 
- [20:05:14] <GargoyleMT> so there are a bunch of people on AS with it? 
- [20:05:21] <xAyiDe> Nope 
- [20:05:25] <GargoyleMT> re: the PM I got and the message in chat
about the tth "bug"
- [20:05:35] <xAyiDe> No not related 
- [20:05:39] <xAyiDe> I dont have any share in AS 
- [20:05:43] <xAyiDe> And none else is using this 
- [20:05:46] <xAyiDe> yet. 
- [20:06:30] <GargoyleMT> well it seems some 0.4033 clients are
reporting the same hash for multiple 15,000,000 byte files in their
share
- [20:06:38] <GargoyleMT> you've seen the screens, what does XDC
detect them as?
- [20:09:25] <xAyiDe> 40x 
- [20:09:33] <xAyiDe> 401 - 4034 
- [20:10:17] <GargoyleMT> any fakeshare checking in xdc?  it could
just be an ordinary fake file?
[20:18:13] <fnordpojk> they are. 
[20:18:26] <fnordpojk> i downloaded all of them, and they're all just
a bunch of 0's.
[20:23:37] <fnordpojk> anyway, why am i up at 2.20AM? 
[20:23:55] <Trem> because you love spending time with us? =) 
[20:24:02] <fnordpojk> oh. 
[20:24:03] <fnordpojk> right. 
[20:24:43] <GargoyleMT> thank you, gentlemen. 
[20:25:09] <GargoyleMT> so it looks not like xayide at all, but just
some ordinary fake files.
[20:25:17] <fnordpojk> yes, in this case. 
[20:25:22] <fnordpojk> for that specific tth. 
[20:25:25] <GargoyleMT> I couldn't have done... anything without you
both.  thanks.
[20:25:42] <GargoyleMT> yeah, if there's really a bug, I'd like to see
it so we can fix it.
[20:26:41] <fnordpojk> haven't seen any evidence in favor of that yet, though. 
[20:31:34] <fnordpojk> anyway. bedtime. bbl. 
[20:32:18] <GargoyleMT> g'night 
[22:16:04] <cologic> -history