Timestamps are in GMT/BST.
[0:01] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[0:25] * Hadaka (naked@naked.iki.fi) Quit (Read error: 54 (Connection reset by peer))
[0:25] * Hadaka (naked@naked.iki.fi) has joined #freenet
[0:27] * oierw` is now known as oierw
[1:24] * CoolerQ (~quentins@cpe-66-189-88-43.ma.charter.com) has joined #freenet
[1:28] <CoolerQ> I just launched freenet for the first time in a looong time... I took care to run update.sh, but it still seems like it's not connecting to the network (Route Not Found for everything); any suggestions?
[1:36] * salahx (salahx@sc1-24.217.174.147.charter-stl.com) Quit ()
[1:38] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit (".")
[1:39] * CoolerQ (~quentins@cpe-66-189-88-43.ma.charter.com) Quit ("Leaving")
[1:56] * mazzanet (~mazzanet@mazzanet.registered) Quit ("Read error: Connection reset by Telstra")
[2:02] * mazzanet (~mazzanet@mazzanet.registered) has joined #freenet
[2:47] * Ash-Fox (Hal-9000@aba29.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[2:47] * Ash-Foxeus (Hal-9000@aak68.neoplus.adsl.tpnet.pl) has joined #FreeNET
[3:06] * m[a]zzanet (~mazzanet@mazzanet.registered) has joined #freenet
[3:07] * mazzanet (~mazzanet@mazzanet.registered) Quit (Nick collision from services.)
[3:07] * m[a]zzanet is now known as mazzanet
[3:37] <iip_i2p> <Newsbyte> toad, are you on?
[3:39] * mazzanet slaps a large trout around a bit with a Newsbyte
[3:43] <iip_i2p> <Newsbyte> well, in any case, I suspect you log the chan and search for your name, so, you'll find this, toad ;-) If you set a flag with a fixed 20% failure-chance, doesn't it become trivial to determine if an insert comes from a certain node? Everytime node a inserts an update of his freesite, there is a 20% failure; if node b forces it (through an adapted seednodes.ref) to use node c, also under his control, he can see if there is a flag working.
[3:44] <iip_i2p> <Newsbyte> If you can repeat that, say, 10 times, you'll become increasingly sure that node a is the one that inserts it. No?
[3:52] <iip_i2p> <Newsbyte> /me prods mazza with a long stick (with a needle on top)
[3:53] <mazzanet> ow
[4:33] * aum (~aum@port-204-54-210.fastadsl.net.nz) has joined #freenet
[4:35] <aum> hi - anyone here who knows freenet code?
[4:38] * nextime (~nextime@danex.i-m-c.it) Quit ("nextime has no reason")
[4:41] * nextime (~nextime@danex.i-m-c.it) has joined #freenet
[4:48] * mazzanet (~mazzanet@mazzanet.registered) Quit (Read error: 60 (Operation timed out))
[4:48] * mazzanet (~mazzanet@mazzanet.registered) has joined #freenet
[5:21] * m[a]zzanet (~mazzanet@203-217-48-65.dyn.iinet.net.au) has joined #freenet
[5:23] * mazzanet (~mazzanet@mazzanet.registered) Quit (Read error: 60 (Operation timed out))
[5:23] * m[a]zzanet is now known as mazzanet
[5:32] * mazzanet (~mazzanet@mazzanet.registered) Quit (Read error: 54 (Connection reset by peer))
[5:38] <Iakin5> aum: What is the question?
[5:54] * mazzanet (~mazzanet@mazzanet.registered) has joined #freenet
[6:08] <Iakin5> Can someone rebuild the snapshots?
[6:08] <Iakin5> There was an evil NPE in 60201
[6:09] <Iakin5> Which is fixed in 60292
[6:09] <Iakin5> or well..
[6:21] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[6:37] * m[a]zzanet (~mazzanet@mazzanet.registered) has joined #freenet
[6:39] * mazzanet (~mazzanet@mazzanet.registered) Quit (Nick collision from services.)
[6:39] * m[a]zzanet is now known as mazzanet
[6:49] * Fisu (~fisu@hoas-50dd0932.hoasnet.inet.fi) has joined #freenet
[6:51] <Fisu> Hmm: ERROR): Encountered an unexpected error when locking the lockfile 'lock.lck'. Shutting down.
[6:54] <Fisu> Also there is after that: java.io.IOException: Value too large for defined data type at xxx at xxx etc.
[7:10] * Ash-Foxeus (Hal-9000@aak68.neoplus.adsl.tpnet.pl) Quit (Client Quit)
[7:25] <Iakin5> Fisu: No idea what that means
[7:28] <Fisu> Well in anycase, the result is that current freenet-latest.jar won't start at all
[7:28] <Fisu> 5091 that is, probably
[7:37] <Iakin5> Fisu: can you see a lock.lck file somewhere in the freenet/store folder?
[7:37] <Iakin5> And what platform are you on?
[7:42] <Iakin5> Fisu: This is what fails for you: lockFileLock = new FileOutputStream(lockFile).getChannel().tryLock()
[7:42] <Iakin5> or rather 'new FileOutputStream("lock.lck").getChannel().tryLock()'
[7:43] <Iakin5> What that code is supposed to do is to lock the file for exclusive use..
[7:43] <Fisu> Yes, it creates an empty lock.lck file in the storeFile dir
[7:43] <Iakin5> (using the locking facilities of the underlying filesystem)
[7:43] <Iakin5> Fisu, right it should be empty
[7:43] <Fisu> I'l try different JVM
[7:44] <Iakin5> What JVM did you use?
[7:44] <Iakin5> This is a post 1.4 feature
[7:44] <Fisu> Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2-gentoo-1075231362)
[7:44] <Fisu> 15:02 < Iakin5> This is a post 1.4 feature
[7:44] <Fisu> ups
[7:44] <Iakin5> Hmm..
[7:45] <toad_> Newsbyte: I don't get it
[7:45] <Fisu> sure enough blackdown seems to work
[7:45] <Iakin5> ok, good
[7:45] <Fisu> Java(TM) 2 Runtime Environment, Standard Edition (build Blackdown-1.4.1-01)
[7:46] <toad_> insert is: first N hops - 20% chance of going to second stage. then HTL hops, till htl -> 0.
[7:46] <Iakin5> Something to report to the gentoo team maybe?
[7:46] <toad_> then M hops, 20% chance of terminating on each
[7:46] <Fisu> Still I've never had any problems with the other JVM and so someone else *might* get this same problem
[7:46] <Fisu> any = any freenet related problems
[7:46] <toad_> i don't see what Newsbyte's problem is...
[7:46] <Iakin5> Fisu, you are the first reporter of that particular issue
[7:46] * Ash-Fox (Hal-9000@aax159.neoplus.adsl.tpnet.pl) has joined #FreeNET
[7:47] <toad_> malicious nodes can end it prematurely? well duh!
[7:47] <Iakin5> And that feature has been in unstable for almost a month now
[7:47] <Iakin5> and in stable ever since 5091 or so
[7:47] <Fisu> Iakin5: Well I've been using stable. I guess the JVM is not-good then
[7:47] <toad_> per node FT is best way to work around such attacks
[7:48] <Iakin5> toad, output from my updated messageaccounter:
[7:48] <Iakin5> QueryAborted 29/77
[7:48] <Iakin5> QueryRejected/Aborted query 25/5
[7:48] <Iakin5> QueryRejected/All routes failed. 2/0
[7:48] <Iakin5> QueryRejected/Cache failed 0/1
[7:48] <Iakin5> QueryRejected/DataInsert never received 15/9
[7:48] <Iakin5> QueryRejected/Looped request 104826/1469
[7:48] <Iakin5> QueryRejected/No route found 69805/4116
[7:48] <Iakin5> QueryRejected/Node overloaded 0/720
[7:48] <Iakin5> QueryRejected/You sent bad data! for freenet.node.states.request.TransferInsert@12410997 0/0
[7:48] <Iakin5> QueryRejected/You sent bad data! for freenet.node.states.request.TransferInsert@19691482 0/0
[7:48] <Iakin5> QueryRejected/You sent bad data! for freenet.node.states.request.TransferInsert@4455407 0/0
[7:48] <Iakin5> QueryRejected/You sent bad data! for freenet.node.states.request.TransferInsert@654449247 0/0
[7:48] <Iakin5> QueryRejected/null 16/0
[7:48] <toad_> there is a question of whether to count a collision as a fatal failure for ft purposes...
[7:48] <toad_> Iakin5: woah
[7:48] <toad_> LOTS of looped request
[7:49] <toad_> like i said...
[7:50] <toad_> that may be a problem for me and kenman...
[7:50] * Hory (~Miranda@81.196.25.110) Quit (Operation timed out)
[8:04] * osh___ (~chatzilla@h88n2fls307o1039.telia.com) has joined #freenet
[8:06] <toad_> Iakin5: that's sent/received or received/sent ?
[8:07] <Iakin5> sent/recieved
[8:08] <Iakin5> I had 16 cache failed before a restart earlier
[8:08] <Iakin5> And something like 20 storedata
[8:09] <Iakin5> And those 'DataInsert never received' seems to always happen
[8:09] <Iakin5> 16/10 of them now
[8:09] <Iakin5> toad, is storedata the actual data-carrier?
[8:18] * osh__ (~chatzilla@h88n2fls307o1039.telia.com) Quit (Read error: 110 (Connection timed out))
[8:27] * osh___ is now known as osh
[8:32] * osh (~chatzilla@h88n2fls307o1039.telia.com) Quit ("ChatZilla 0.9.61 [Mozilla rv:1.8a1/20040520]")
[8:35] * verl (~verlverl@h165n2fls33o877.telia.com) Quit (Read error: 104 (Connection reset by peer))
[8:35] * Ash-Fox (Hal-9000@aax159.neoplus.adsl.tpnet.pl) Quit (Client Quit)
[8:35] * Ash-Fox (Hal-9000@aax159.neoplus.adsl.tpnet.pl) has joined #FreeNET
[8:45] <aum> hi - anyone here who knows freenet code?
[8:47] * verl (~verlverl@h165n2fls33o877.telia.com) has joined #freenet
[9:03] <Iakin5> 11:57] <Iakin5> aum: What is the question?
[9:07] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[9:07] * hirvox (~hirvox@cs181027153.pp.htv.fi) has joined #freenet
[9:18] <toad_> hmmm
[9:18] <toad_> anyone know anything about broadband tech?
[9:19] <hirvox> whee, my ISP lowered its prices and raised its speeds. 3Mbit down, 500kbit up now for 59e :)
[9:19] <toad_> Speed: 512/256 Kbps (Download/Upload)
[9:19] <toad_> Rate Adaptation: Actual speeds may vary depending upon line length/quality.
[9:19] <toad_> Worst-Case Speeds: 512/57 Kbps (after rate adaptation)
[9:19] <toad_> RADSL sacrifices uplink for downlink?
[9:19] <toad_> does cable?
[9:20] <hirvox> not really, but latency goes thru the roof at approximately 90% uplink usage
[9:20] <toad_> in other words, is cable inherently better for uplink ~= downlink applications?
[9:21] <hirvox> bbl, -> post office
[9:21] <toad_> aha
[9:21] <toad_> some of them are ADSL not RADSL
[9:21] <toad_> ADSL is better, right?
[9:26] <toad_> oh i see
[9:26] <toad_> |Your download speed should be identical i.e. 512kbps. Unless your line quality is very poor (downstream attenuation greater than 52dB), you should see full speed upload also.
[9:26] <toad_> not likely in a residential urban area..
[9:42] <hirvox> back
[9:51] * lolo-laptop (~lostlogic@68.251.84.226) has joined #freenet
[9:53] * hobx (~hobx@h196n1fls21o1077.bredband.comhem.se) Quit (niven.freenode.net irc.freenode.net)
[9:53] * leex-Zzzz-gone (~bbtt@cpc2-warr1-5-0-cust27.bagu.cable.ntl.com) Quit (niven.freenode.net irc.freenode.net)
[9:53] * vsalento (~vsalento@cs136227.pp.htv.fi) Quit (niven.freenode.net irc.freenode.net)
[9:53] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) Quit (niven.freenode.net irc.freenode.net)
[9:53] * oierw (mathew@cpe-69-75-112-240.hawaii.rr.com) Quit (niven.freenode.net irc.freenode.net)
[9:53] * xena (~xena@espians.com) Quit (niven.freenode.net irc.freenode.net)
[9:53] * toad_ (toad@toad-with-underline.active.supporter.pdpc) Quit (niven.freenode.net irc.freenode.net)
[9:53] * Ash-Fox (Hal-9000@aax159.neoplus.adsl.tpnet.pl) Quit (niven.freenode.net irc.freenode.net)
[9:53] * mazzanet (~mazzanet@mazzanet.registered) Quit (niven.freenode.net irc.freenode.net)
[9:53] * eMBee (~mbaehr@sirius.iaeste.tuwien.ac.at) Quit (niven.freenode.net irc.freenode.net)
[9:53] * bluephile (bluephile@69-160-205-193.clvdoh.adelphia.net) Quit (niven.freenode.net irc.freenode.net)
[9:53] * aum (~aum@port-204-54-210.fastadsl.net.nz) Quit (niven.freenode.net irc.freenode.net)
[9:53] * thelema (~thelema@adsl-65-65-203-187.dsl.stlsmo.swbell.net) Quit (niven.freenode.net irc.freenode.net)
[9:53] * hobx_ (~chatzilla@ankh.math.chalmers.se) Quit (niven.freenode.net irc.freenode.net)
[9:53] * dystopia (admin@gaia-wan0.ipv6.freeshell.bofx.net) Quit (niven.freenode.net irc.freenode.net)
[9:53] * hirvox (~hirvox@cs181027153.pp.htv.fi) Quit (niven.freenode.net irc.freenode.net)
[9:53] * hobx (~hobx@h196n1fls21o1077.bredband.comhem.se) has joined #freenet
[9:53] * hirvox (~hirvox@cs181027153.pp.htv.fi) has joined #freenet
[9:53] * Ash-Fox (Hal-9000@aax159.neoplus.adsl.tpnet.pl) has joined #freenet
[9:53] * mazzanet (~mazzanet@mazzanet.registered) has joined #freenet
[9:53] * aum (~aum@port-204-54-210.fastadsl.net.nz) has joined #freenet
[9:53] * leex-Zzzz-gone (~bbtt@cpc2-warr1-5-0-cust27.bagu.cable.ntl.com) has joined #freenet
[9:53] * vsalento (~vsalento@cs136227.pp.htv.fi) has joined #freenet
[9:53] * eMBee (~mbaehr@sirius.iaeste.tuwien.ac.at) has joined #freenet
[9:53] * xena (~xena@espians.com) has joined #freenet
[9:53] * toad_ (toad@toad-with-underline.active.supporter.pdpc) has joined #freenet
[9:53] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) has joined #freenet
[9:53] * oierw (mathew@cpe-69-75-112-240.hawaii.rr.com) has joined #freenet
[9:53] * hobx_ (~chatzilla@ankh.math.chalmers.se) has joined #freenet
[9:53] * bluephile (bluephile@69-160-205-193.clvdoh.adelphia.net) has joined #freenet
[9:53] * thelema (~thelema@adsl-65-65-203-187.dsl.stlsmo.swbell.net) has joined #freenet
[9:53] * dystopia (admin@gaia-wan0.ipv6.freeshell.bofx.net) has joined #freenet
[9:53] <hobx> fuck
[9:53] <hobx> finished
[9:54] * sanity (~ian@81-178-89-164.dsl.pipex.com) has joined #freenet
[9:54] * ChanServ sets mode +o sanity
[9:54] <sanity> ooh, two of me
[9:54] <sanity_> hello other sanity
[9:54] <sanity> hi hobx
[9:54] <hobx> dorkus McMaximus strikes again
[9:54] <hobx> morn
[9:56] <sanity> hobx: going by dorkiness = 1 / numGirlfriends
[9:56] <sanity> hobx: i think that makes you dorkus McInfiniteus
[9:56] <hobx> which means that having one is full dorkiness?
[9:57] <sanity> nope, its a scale between 0 and infinity
[9:57] <sanity> anyway, how is life?
[9:57] <sanity> hobx: i got a mac
[9:58] <hobx> you told me.
[9:59] <hobx> as long as you don't turn into a mac user I'm fine with it.
[9:59] <sanity> well, i use a mac, doesn't that make me a mac user?
[10:00] <hobx> no.
[10:00] <toad_> oooh
[10:00] <hobx> level of exposure to the Jobs RDF makes you a Mac user.
[10:00] <toad_> it appears my ISP no longer has the 10-server-connections-only no-public-servers clause!
[10:00] <toad_> this means we can run a web server, and my freenet usage is no longer in technical breach
[10:02] <sanity_> hobx: Jobs RDF?
[10:02] <hirvox> Reality Distortion Field
[10:03] <sanity_> ah
[10:03] <sanity_> well, he definitely knows how to make pretty computers
[10:03] <hirvox> afk
[10:03] <sanity_> and the UI makes me rather depressed for the prospect of Linux catching up, Gnome and KDE still seem to be playing catchup with Windows 95 in many respects
[10:03] <hobx> You feeling a strong urge to start paying for OS updates and bugfixes?
[10:04] <sanity_> you get what you pay for *ducks*
[10:04] <toad_> actually technically we ARE in breach because it says "your PC must meet the Minimum Spec" i.e. run windoze, but lets ignore that, the installation dude did
[10:04] <toad_> sanity: I like KDE
[10:04] <hobx> sanity: Not true for sex, not true for software
[10:05] <toad_> granted Windows XP is a little more ummm verbose?
[10:05] <toad_> it's more something
[10:05] <hobx> I like windowmaker
[10:05] <toad_> but KDE is nice in other ways
[10:05] <sanity_> sanity: well, i have never used a hooker, so for all i know that could be better
[10:05] * sanity_ used to use WindowMaker way back in 96-97
[10:05] <toad_> lol
[10:06] <sanity_> of course, WindowMaker is a pale imitation of NextStep
[10:07] <sanity_> oh shit, i think i might be entering the Jobs RDF - noooooooooo.....
[10:07] <hobx> So I finished writing my simulator. This really sucks.
[10:07] <hobx> I don't want to run it.
[10:08] <hobx> Anybody want to run a betting pool on what will happen when I do. The odds are:
[10:08] <hobx> NPE: 4
[10:08] <hobx> AIOB: 3.5
[10:08] <hobx> INFINITE LOOP: 7
[10:09] <hobx> WORKS: 100000
[10:10] <hobx> NPE! We have a winner!
[10:10] <sanity> what kind of simulator?
[10:11] <hobx> any DHT really
[10:11] <sanity> why?
[10:11] <hobx> first algorithm is implemented is "Symphony"
[10:11] <hobx> I want to try some variants see if they work.
[10:17] <hobx> Winner #2 is infinite loop!
[10:17] <sanity_> does your simulator account for variations in latency between peers?
[10:20] <hobx> you sound like toad
[10:20] <hobx> I'm not interested in your petty real world problems!
[10:21] <sanity_> you can hear me? freaky!
[10:21] <sanity_> hobx: duh, your a mathematician
[10:21] <hobx> There is some interesting math here. Having a simulator helps show when it is worth exploring.
[10:22] <sanity_> what is so interesting about O(log(N)) ?
[10:22] * lolo-laptop (~lostlogic@68.251.84.226) Quit ("Client exiting")
[10:23] <hobx> http://citeseer.ist.psu.edu/kleinberg00smallworld.html
[10:23] * lolo-laptop (~lostlogic@68.251.84.226) has joined #freenet
[10:23] <sanity_> smallworld is sublinear, not logarithmic, right?
[10:23] <hobx> noop
[10:24] <hobx> log(n)?
[10:24] <hobx> so poly-log
[10:24] <sanity_> oh wait
[10:24] <sanity_> small world assumes optimal routing decisions - right?
[10:25] <hobx> read that paper
[10:25] <sanity_> i don't have time to read papers
[10:25] <hobx> log(n)? is for pure greedy routing
[10:25] * sanity_ reads the abstract
[10:25] <hobx> ie, route to the node that is closest in id to that you seek
[10:25] <sanity> gah
[10:25] <hobx> though he deals with ids as positions in a 2D grid, their are variants
[10:26] <hobx> for one dimensional numerical difference.
[10:26] <sanity> that assumes you already know the ID of the server you are looking for! why not just use its IP address
[10:27] <hobx> no
[10:27] <hobx> Same as all DHTs
[10:27] <hobx> when you insert data, you have some key, place it on the node whose id matches closest.
[10:27] <hobx> When you want to find something based on the key, find the node whose id matches closests
[10:28] <sanity> cool, i want the node with the id closest to that illegal scientology stuff
[10:28] <hobx> It isn't that different from freenet except Freenet never explictely states what keyspace regions a node should specialize in.
[10:28] <hobx> sanity: Yes, DHTs cannot prevent that form of attack, obviously.
[10:29] <hobx> I have serious doubts whether freenet can either though :-(
[10:29] <sanity> hobx: well, freenet makes it harder
[10:29] <sanity> hobx: too many people waste time trying to make things impossible, when it is sufficient just to make them harder
[10:29] <hobx> wasp
[10:29] <Redb3ard> heh
[10:29] <hobx> catch it, come back
[10:30] <Redb3ard> scientology is a good test case for freenet, id think
[10:30] <sanity> hobx: "wasp"?
[10:30] <sanity> oh, there is a wasp
[10:30] <sanity> do make sure you really piss it off, then it will leave you alone
[10:30] <sanity> :-)
[10:31] * sanity has a kitten for dealing with such annoyances
[10:31] <hobx> seems to have found it's own way out.
[10:31] <Redb3ard> damn
[10:31] <sanity> spiders, moths, bees, wasps, nothing he won't eat
[10:31] <KenMan> hey, that's cool that toad can fully communicate now (with a webserver)
[10:31] <Redb3ard> those white anglo-saxon protestants are always causing trouble
[10:31] <Redb3ard> it probably left for an ivy league school though
[10:32] <hobx> sanity: enforcing id = SHA1 of IP address probably does it better in practice than freenet.
[10:32] <Redb3ard> so you should be safe until spring break
[10:32] <sanity> hobx: hobx: yeah, until your ISP changes your IP address
[10:32] <hobx> Anyways, the algorithm that Kleinberg proves to work is very close to freenet, so I think you should be more interested and less dismissive.
[10:33] <sanity> hobx: old routing I assume, not NGR?
[10:33] <hobx> It is a very interesting result, especially if one hopes to analyse anything about freenet (I don't particularly)
[10:33] <sanity> ok ok, i will print it out and put it beside the toilet (only guaranteed way to ensure that i read something)
[10:35] <sanity> well, the result doesn't particularly surprise me since freenet is topologically similar to a binary search tree
[10:36] <sanity> although it surprises me that someone proved it
[10:36] * lolo-laptop (~lostlogic@68.251.84.226) Quit (Remote closed the connection)
[10:37] <sanity_> 44433333333333333333333333i8888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888888l;\\\\\\\\\\\\\\\\\\\\\\\\\\\\
[10:37] <sanity> turns out the cat is fascinated by printers
[10:37] <sanity> lateral motion really gets them excited
[10:37] * lolo-laptop (~lostlogic@68.251.84.226) has joined #freenet
[10:37] <sanity> he looks like a spectator at a tennis match
[10:38] <hobx> get a dog
[10:39] <sanity> why?
[10:39] <sanity> i can take the cat for a walk
[10:39] <sanity> he hangs out with us at the local pub
[10:39] <sanity> everything good about a dog, without the slavish sycophancy
[10:42] <sanity> and, to top it all, he scored 8.2 out of 10 on ratemykitten.com! (see http://www.ratemykitten.com/ratemy/kitten?image=151083)
[10:42] <toad_> LOL
[10:43] <toad_> |The Service may comprise software, services, technical information, training materials or other technical data which are subject to the United States of America Export Control Regulations or the laws or regulations of another country. You may not download or otherwise export or re-export the Software or any underlying information or technology except in full compliance with all United States and other applicable laws and regulations.
[10:43] <toad_> this is a UK ISP !
[10:44] <sanity> ah, sounds like that excellent legal firm, Cut & Paste Bros.
[10:44] <toad_> very likely
[10:44] * toad_ is just making absolutely sure of what I discovered earlier
[10:47] <toad_> well actually iirc there's some we-won't-trade-with-megacorps-that-export-to-cuba thingy
[10:47] <toad_> so i suppose it makes sense
[10:50] <Redb3ard> guys, how deep does the encryption go on freenet? can i actually see what request a node other than my own is making?
[10:51] <toad_> Redb3ard: you can see what requests are being made of your node
[10:51] <toad_> but you only see the key
[10:51] <Redb3ard> in such a way that i could request it myself, and see what pops up in the browser window
[10:51] <toad_> the key is of course encrypted
[10:51] <toad_> so no
[10:51] <toad_> you cannot decrypt it
[10:52] <Redb3ard> i couldnt play that back, and get anything?
[10:52] <Redb3ard> hows that done?
[10:52] <Redb3ard> each node has its own public/private key going on?
[10:55] <sanity> Redb3ard: you could play it back, but you couldn't decrypt the resulting data
[10:56] <sanity> Redb3ard: essentially you have Kp - the plaintext key, lets say it is the string "dog"
[10:56] <sanity> Redb3ard: you hash it using SHA1 to get Kc - the crypto key
[10:56] <sanity> you then request Kc from the network
[10:57] <sanity> when you get the resulting data, Dc - you decrypt it using a symmetric crypto algorithm (AES or Twofish, can't remember) where the key is Kp
[10:57] <sanity> note that only the requestor knows Kp, nodes relaying the request only see Kc and Dc
[10:58] <toad_> where he says key read data - content of the file :)
[10:58] <sanity> of course, given Kc you could use a dictionary attack to find Kp, but generally speaking that would be difficult
[10:58] <toad_> bbl
[10:58] <sanity> toad_: thanks for obfuscating that
[10:59] <sanity> erm, i mean explaining :-)
[11:04] * caddy (caddy@pD952A397.dip.t-dialin.net) has joined #freenet
[11:07] <caddy> well ... im a bit of a noob concerning the finer details of netiquette, so do you think it's okay for me to ask a question that prolly has already been answered somewhere in a thread somewhere around the net; and making a suggestion that is probably absolute nonsense, although I can't find any obvious mistake in it... ?
[11:07] <caddy> oh, btw, hi everybody.
[11:07] <sanity> caddy: why not, everyone else does :-)_
[11:08] <caddy> thx, so as I see it, you've got the problem that dividing a file into fixed size parts reduces anonymity for big file dls?
[11:09] <caddy> could this be solved by increasing the probability that continous parts of the file are requested from the same client, and for said client increasing the priority of such requests?
[11:09] <sanity> i don't think so
[11:09] <caddy> i dont too
[11:09] <caddy> but i dont see y not.
[11:10] <caddy> too easy to not have an obvious mistake in it, i suppose-
[11:10] <sanity> the issue is that if someone sees you requesting a bunch of things, and they know that they are all parts of the same file, then they can infer that it is you requesting the file, rather than someone downstream of you
[11:10] <caddy> but if all clients do this, aren'
[11:10] <sanity> i think your proposal would make that worse, not better
[11:10] <caddy> sry
[11:10] <caddy> t we back to plausible deniality?
[11:10] <caddy> deniability
[11:11] <sanity> yes, and you have less plausible deniability that you were requesting a given piece of data when it can be shown that you requested many pieces of that data
[11:11] * emacsen (~serge@pcp09296048pcs.arlngt01.va.comcast.net) has joined #freenet
[11:11] <emacsen> Hey all
[11:11] <sanity> caddy: unfortunately I need to run, you should email the tech mailing list about it
[11:11] <sanity> emacsen: hi bye
[11:11] <caddy> ok
[11:11] <caddy> hi
[11:12] <emacsen> bye?
[11:12] <emacsen> oh ok
[11:13] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) Quit ("no reason")
[11:19] * interrupt (~chatzilla@64.122.23.213) has joined #freenet
[11:21] * Hory (~Miranda@81.196.25.110) Quit (Read error: 104 (Connection reset by peer))
[11:23] * salahx (salahx@sc1-24.217.174.147.charter-stl.com) has joined #freenet
[11:26] <Redb3ard> too bad you cant make freenet always traverse at least 2 hops for any request
[11:26] <Redb3ard> then you could have provable deniability
[11:26] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[11:26] <Redb3ard> "hey, the source code here says it couldnt have been me making the requests"
[11:28] <salahx> well actually to some extent its already donwe
[11:29] <salahx> THe HTL isn't always decremented, the lower the HTL, the less liekly it is to get decrmemnted
[11:29] <thelema> would it suffice to have a noDS flag on requests so that nodes recieving such a request wouldn't check their datastore before forwarding the request (with a fixed chance of removing that flag)?
[11:30] * Brutha- (~x@pD9587E22.dip.t-dialin.net) has joined #freenet
[11:31] <hirvox> request success rates are lousy already, why do we want to make it worse?
[11:40] <KenMan> to keep things interesting ? :p
[11:41] <salahx> perhaps to avoid giving users a heart attack when this things astarts working well? :
[11:42] <salahx> the README already recomend watching paint dry of a hour a day to adjuste your ssytem to the blazing speeds and successes of Freenet
[11:44] <hirvox> :)
[11:45] <KenMan> for the past 13 hours, my node has sent out 2.5X the number of data requests that it has taken in :(
[11:45] <KenMan> how can a network sustain itself if all the nodes run in a similar manner ?
[11:46] <KenMan> check localQueryTraffic/hr and compare to outboundAggregateRequests so measure your own node...
[11:46] * TLF (francisco@31.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[11:47] <KenMan> oh, wait. It is iakin's node that becomes the "great equalizer" - it sucks up that 1.5X excess from all nodes, to make the ledger balance out.
[11:48] <thelema> KenMan: it can't.
[11:48] <thelema> but not all nodes do this.
[11:48] <hirvox> my node's ratio is 1.5x
[11:48] <KenMan> i know. I am still stuck trying to understand the 'math'
[11:49] * KenMan believes that most nodes do overproduce, and few big suckers have to compensate
[11:49] <thelema> hirvox: we'd do noDS as an alternative to real onion routing.
[11:51] <KenMan> i think iakin's node is a "rubberhole" - it sucks in so many queries that they can't help but to collide on the way back out
[11:52] <hirvox> thelema:then we'll have to disable looping checks for nodes that forward the request while noDS is set
[11:52] <hirvox> because if the data was on those nodes, we're screwed
[11:52] <KenMan> a blackhole eats things - they don't come back out. So perhaps what he has should be termed a 'whitehole'
[11:53] <Redb3ard> heh, hey toad, you guys want a favicon.ico for the freenet page?
[11:55] <thelema> hirvox: if a request comes back DNF to a node with the data, just have the node check then.
[11:55] <Redb3ard> http://24.125.12.101/freenet.ico
[11:55] <Redb3ard> what can i say, i was bored
[11:56] <hirvox> thelema:fair enough
[11:58] * thelema is now known as thelema|away
[12:00] <KenMan> holy smokes - i finally hit 80 peers in stable (my limit) !!!
[12:13] <sanity_> KenMan: how is stable working for you?
[12:16] <KenMan> well, seeing as i don't actually *use* the network, pretty well.
[12:17] <KenMan> it is not horrible, judging by the stats...
[12:17] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) has joined #freenet
[12:17] <KenMan> but i am using the eye of someone who has been tracking the performance stats over many months...
[12:18] <KenMan> it is not much worse or much better than 5084, just 'different' in its behavior
[12:20] <KenMan> i am still uncomfortable with mRI / load balancing. I patched 5091 with Toad's latest mRI improvements, and it may be a little better as a result.
[12:22] <KenMan> my biggest concern is that we do not consider the amount of resources allocated to us (mRI) in a direct manner when computing load / our outgoing resource allocations. It is a bear of a problem to figure out.
[12:23] <KenMan> it is damned near impossible when routes are changing mRI values to us multiple times for each single request we send to them.
[12:25] <KenMan> there is a nasty feedback type of problem resulting from this behavior... one that would not be so dramatic if mRIs updates were less frequent than the rate of queries
[12:27] <KenMan> something like "only send one mRI out for every 5 requests taken in" would probably smooth query rates very nicely, when looking at small time intervals of around 1 minute.
[12:28] <KenMan> For instance, when I view the queries per minute graph, it has a very sawtooth behavior, regardless of the amplitude. What I mean is that the direction changes almost every minute (one down, the next up)...
[12:28] <KenMan> I dunno, maybe I'm just spewing nonsense...
[12:32] * supermannn (~kdfjk@216.32.201.125) has joined #freenet
[12:32] <supermannn> people were suggesting using 90 connections for splitfiles... is there a max limit? why not use 300 or something real high?
[12:32] <KenMan> supermannn: go for it.
[12:33] <supermannn> is there any reason not to?
[12:33] <KenMan> i think that when you begin using somewhere around 8000 peers, that level is starting to become questionable.
[12:33] <supermannn> lol
[12:34] <KenMan> no i'm dead serious. You shouldn't get much resistance around here.
[12:35] <KenMan> are you referring to number of peers, or are you talking about the 'simultaneous connections' setting for splitfiles ?
[12:35] <supermannn> sim conn :)
[12:36] <KenMan> yeah, set it kind of high, maybe 5000 or so. It may take a few years to reach that level, but after that point you should have good performance.
[12:36] <supermannn> im watching a graph of dl rate, and experimenting with diff settings
[12:36] <supermannn> at 90 connections i got around 15KB/sec.....as i increase the connections it jumps faster. im up to 70KB/sec using 227 conn
[12:37] * TLF (francisco@31.Red-81-40-113.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[12:37] <KenMan> cool . Trouble is that you can't diddle the knobs on the other nodes :(
[12:37] <KenMan> if we could get everyone to use many more connections, the transfer rate should be almost unlimited...
[12:39] <supermannn> hmm where does fuqid store temp files
[12:40] * nextgens (~nextgens@ALyon-252-1-12-10.w82-122.abo.wanadoo.fr) has joined #freenet
[12:40] <supermannn> 80KB/sec
[12:40] <supermannn> nice
[12:41] <supermannn> if all of freenet, including browsing could do 80KB id use it more
[12:41] <supermannn> and insert sites
[12:41] <nextgens> hello !
[12:41] <nextgens> downloading what? child P0rn ? ;-)
[12:41] <KenMan> yeah, and all that filthy animal porn too
[12:41] <KenMan> :p
[12:42] <supermannn> some file called, 'george bush screwing a donkey.mpg'
[12:42] <supermannn> kidding
[12:42] * caddy (caddy@pD952A397.dip.t-dialin.net) Quit ()
[12:42] <nextgens> I've got an ask : Does some dev. work on UP&P ?
[12:43] <supermannn> i wonder if freenet could ever have a commercial potential
[12:43] <KenMan> you wouldn't be the first ;)
[12:43] <nextgens> that's why I'd like to know ... :-)
[12:44] <supermannn> id like to make some original video and insert it as a freenet exclusive
[12:44] <nextgens> Toad talk about it 2 month ago on the devl. list... and since ... nothing ...
[12:44] <KenMan> nextgens: it was talked about, but i don't think anyone actually did anything re: upnp
[12:45] <KenMan> there were one or two references to some UPnP code from some other projects
[12:46] <nextgens> ok, I'll try do to something and if it works I'll publish a patch ...
[12:46] <nextgens> I've read Azurus Sources
[12:46] <supermannn> what political party are most freenet users in?
[12:46] <nextgens> a java bittorent client using upnp ...
[12:46] <emacsen> the one with the girls?
[12:47] <emacsen> and beer
[12:47] <supermannn> girls and beer/
[12:47] <supermannn> ?
[12:47] <emacsen> yeah, a good party
[12:47] <supermannn> we need some good original content on freenet. lets make a documentary
[12:48] <supermannn> any ideas?
[12:48] <KenMan> the People Against Globally Administrated Nationalistic Systems seem to be quite frequent in here...
[12:48] <supermannn> pagan?
[12:49] <emacsen> Judean People's Liberation Front
[12:49] <supermannn> someone should go film george bush shaking hands with osama and insert it into freenet :) then we'll be all over the news
[12:50] <KenMan> that footage (from the late 80's) was banned
[12:50] <supermannn> really?
[12:50] <KenMan> yeah, you won't be getting your hands on it anytime soon...
[12:50] <supermannn> is it on freenet?
[12:52] * mikeDOTd (~mikedotd@plutonium.nexthop.net) Quit (Read error: 113 (No route to host))
[12:52] <supermannn> how many users does freenet have?
[12:52] <supermannn> there should be a little population counter on the interface page
[12:53] <KenMan> maybe we could build a central server that each node would check in with everytime it starts up...
[12:53] <supermannn> nah
[12:53] <supermannn> whys it have to be central to get pops tats
[12:54] <KenMan> it worked very accurately for napster
[12:54] <emacsen> KenMan: yeah a central server would be good
[12:55] <emacsen> maybe a few outer-circle servers
[12:55] <supermannn> id just like to get an idea of how many people are there
[12:55] <emacsen> paint the outer ones white, the inner ones blue and the central one red :)
[12:56] <KenMan> :)
[12:56] <supermannn> itd be cool to see what keys are the most popular too
[12:57] <KenMan> again, we could collect that data with a central server
[12:57] <emacsen> KenMan: there's no way to preserve anonymity then
[12:58] <KenMan> we wouldn't have to report exactly WHEN those keys went through... but yeah, i guess it might not be a superb idea :p
[12:58] <emacsen> even if we trust the server- I can't trust the line
[12:58] <supermannn> wheres the problem
[12:58] <emacsen> a logger
[12:59] <supermannn> if you have a cluster of nodes, just have them pass the data from one to the next, and finally one of them tells it to the server
[12:59] <emacsen> a computer at the ISP who is run by someone we don't like
[12:59] <supermannn> that way you dont know who it came form
[13:00] <emacsen> no, you still have a collection then
[13:00] <emacsen> you remove the onion routing
[13:01] <emacsen> one MIGHT make the argument for every node to take its own cache and submit that but then you still know too much
[13:06] <toad_> <supermannn> someone should go film george bush shaking hands with osama and insert it into freenet :) then we'll be all over the news
[13:06] <toad_> <KenMan> that footage (from the late 80's) was banned
[13:06] <toad_> <supermannn> really?
[13:06] <toad_> <KenMan> yeah, you won't be getting your hands on it anytime soon...
[13:06] <toad_> uhh, sure
[13:07] <toad_> it's not possible to ban something like that in the US, is it? Of course the networks might decide they don't like it...
[13:08] <supermannn> ya sounds fishy to me
[13:08] <supermannn> if they banned such a thing itd be worse PR than if they showed it and just explained it
[13:09] <supermannn> honesty is the best policy :)
[13:11] <supermannn> wouldnt there be some way to have the data of how many people youre connected to, etc, travel from node to node, changing depending on that nodes data, and slowly building a total population number
[13:11] <supermannn> it wouldnt be realtime accurate, but why cant it be done
[13:11] <Iakin> [18:06] <KenMan> check localQueryTraffic/hr and compare to outboundAggregateRequests so measure your own node...
[13:11] <Iakin> [18:06] * TLF has joined #freenet
[13:11] <Iakin> [18:07] <KenMan> oh, wait. It is iakin's node that becomes the "great equalizer" - it sucks up that 1.5X excess from all nodes, to make the ledger balance out.
[13:11] <KenMan> it is only possible if you OWN the footage, and can destroy it. And of course, assuming that no one taped the Oct 23, 1988 airing of the ABC World News.
[13:12] <Iakin> My node's lQPH is about 60-70k whilst outboundAggregateRequests hover at around 7-8k/hour
[13:12] * KenMan 's ass hurts, from producing so much shit ... :p
[13:13] <supermannn> has freenet ever tried to get funding from the US govt?
[13:14] <KenMan> it would provide an excellent opportunity to set up a sting operation. That might sway their evaluation...
[13:14] <supermannn> sting operation?
[13:14] <supermannn> why
[13:14] <KenMan> so, FreeNet could prosper from the unforunate plight of a few hundred new residents at that cuban hotel...
[13:15] <KenMan> GauvaHomino Bay
[13:15] <supermannn> if theyre funding tor, why wouldnt they fund freenet
[13:15] <supermannn> they could use it
[13:22] <Iakin> iakin.poweruser.org/freenetstuff/bwandCPU.jpg
[13:22] * Iakin really needs to figure out what causes those dips..
[13:25] <supermannn> during splitfile transfers i get dips to 0kb/sec
[13:25] <supermannn> well i did
[13:25] <KenMan> iakin, one more time... how many queries per hour are you pushing ??
[13:30] <KenMan> what is the colors for the two graphs ? is the top graph network, bottom CPU ? are they aligned vertically (in time) ?
[13:31] <KenMan> Iakin : do your dips/drops show *any* correlation to the outgoingRequestInterval/minute diagnostics graph ?
[13:33] * supermannn (~kdfjk@216.32.201.125) has left #freenet
[13:33] <KenMan> hey toad, come on and post an image already. Now that you won't be in ISP violation of anything...
[13:45] * TLF (francisco@67.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[13:46] <KenMan> where'd you get that gun, johnny ?
[13:56] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[14:02] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit (",.")
[14:05] <KenMan> do i see a silhouette of somebody pointing something from a tree ? Click Bang Oh what a hang, your daddy just shot poor me...
[14:07] * nextgens (~nextgens@ALyon-252-1-12-10.w82-122.abo.wanadoo.fr) Quit (Read error: 104 (Connection reset by peer))
[14:26] * greycat (~wooledg@192.35.79.70) has joined #freenet
[14:36] * nextgens (~nextgens@ALyon-252-1-12-10.w82-122.abo.wanadoo.fr) has joined #freenet
[14:38] * TLF (francisco@67.Red-81-40-113.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[14:39] <sanity_> morpheus won their appeal with the RIAA. decentralised p2p is legal.
[14:41] <hirvox> did it go all the way? or will we see round 3? or renewed lobbying for INDUCE?
[14:41] <oierw> sanity: link?
[14:42] <oierw> because I had never heard that they had found it legal
[14:42] <sanity_> http://p2pnet.net/story/2193
[14:42] <sanity_> well, induce was the contingency against them losing this
[14:43] <sanity_> but it is hard to imagine induce going through as written given the mounting opposition
[14:46] <Redb3ard> induce is a good idea
[14:47] <Redb3ard> i dont want these pedophiles fondling kids on p2p networks
[14:47] <Redb3ard> wont someone please think of the children!?
[14:48] <sanity_> Redb3ard: ironically induce wouldn't prevent that - it only cares about copyright infringement of Britney Spears dressed up as a 14 year old schoolgirl girating all over the place
[14:48] <greycat> Not that there's anything wrong with that.
[14:49] <Redb3ard> can we kick someone for suspected pedophilia?
[14:49] <Redb3ard> because only a pedophile has any reason to be against induce
[14:49] <Redb3ard> heck, even the name of the bill says quite clearly that its to protect children from dangerous violent sexual predators
[14:50] <Redb3ard> and mp3s
[14:50] <Redb3ard> quite clearly, mp3s are the invention of pedophiles, used as bait so they can get their slimy pedophile hands on young bodies
[14:51] <Redb3ard> if you insist on courtroom-level proof, i introduce this as evidence: mp3s were invented by the germans
[14:53] <Redb3ard> the sickest part, is i dont even think they get top dollar
[14:53] <Redb3ard> congress should auction legislation on ebay, to bring in more revenue
[15:05] <iip_i2p> <NextGen> test
[15:12] <salahx> lol
[15:12] <nextgens> test to #freenet-opn ...
[15:13] <Iakin> [19:45] <KenMan> iakin, one more time... how many queries per hour are you pushing ??
[15:13] <Iakin> 60-70kqph
[15:13] * TLF (francisco@67.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[15:13] <Iakin> [19:50] <KenMan> what is the colors for the two graphs ? is the top graph network, bottom CPU ? are they aligned vertically (in time) ?
[15:13] <Iakin> red = bw out, yellow = bw in, green = CPU
[15:13] <KenMan> i think red is cpu in both , right ?
[15:13] <KenMan> bah !
[15:14] * nextgens (~nextgens@ALyon-252-1-12-10.w82-122.abo.wanadoo.fr) Quit ("going to IIP")
[15:14] <Iakin> And, yes, they are *almost* correctly lined up in time
[15:14] <KenMan> i concluded that also. But you made the CPU *red* in the upper pic
[15:14] <KenMan> oh, now i see...
[15:15] <KenMan> well, tell me what the red is in the lower pic, then i will fully understand it
[15:15] <KenMan> kernel CPU or something ?
[15:18] <KenMan> Iakin: that is crazy, your node is dropping / terminating / RNFing 90% of the requests coming in, huh ? ( cause only 10% ever get propagated back out )
[15:19] <KenMan> I wonder if that ratio would go down if you tried setting an outboundBandwidthLimit of 20k or something, just as a control experiment...
[15:20] * KenMan concludes that iakin's node is definitely some kind of hole... blackhole whitehole rubberhole
[15:21] <KenMan> i know, "instant RNF hole" . Yeah, one of those. Shame on you for running such a node... tsk, tsk.
[15:22] <KenMan> but what else could you possibly do in this case ? all your peers are meager nodes with normal capacity for queries. This seems to be a fault of the load calculations as they translate into outgoing mRIs. The best you could do is adapt to a "normal" rate, and hopefully add value by offering higher transfer rates in some cases.
[15:25] <KenMan> the only way you could offer a higher (than most nodes) query rate with comparable success to them would be to have MORE peers than the normal maxConns setting.
[15:25] <KenMan> and even that appears to be circular logic to me... but...
[15:31] <KenMan> ...it would probably help out with the current implementation. Try using 500 peers or more. Toad - can he do that ? Is there a limit ?
[15:34] <hobx> What is worse than not seeing the results you wanted in an experiment?
[15:35] <KenMan> seeing them in someone else's experiment ??
[15:35] <greycat> Having the experiment get crushed by a meteor?
[15:35] <KenMan> ;)
[15:36] <KenMan> ahh, drawing conclusions that are supposed to explain the discrepancy. That is definitely worse.
[15:36] <hobx> seeing results that are too good to be true.
[15:36] <KenMan> that can be very stressful
[15:36] <greycat> getting the bill.
[15:37] <KenMan> i tip my hat to greycat
[15:38] <hobx> he
[15:38] <KenMan> btw that has a perfect analogy to dating, which might explain hobx's aversion to women.
[15:38] <hobx> Try the mathematical sciences for a change.
[15:39] * TLF (francisco@67.Red-81-40-113.pooles.rima-tde.net) Quit (Read error: 104 (Connection reset by peer))
[15:39] <KenMan> i don't want to lose even *more* hair, now...
[15:39] <Redb3ard> thank god i dont have a balding gene
[15:39] <hobx> bald is sexy!
[15:39] <Redb3ard> um, yeh
[15:39] <Redb3ard> i bet the wife tells you 4" is too big, also
[15:40] <Redb3ard> hey guys, how much bandwidth do you personally use on freenet in a week's time?
[15:40] * oierw` (mathew@cpe-69-75-112-240.hawaii.rr.com) has joined #freenet
[15:41] <greycat> right now none... my node's been down for a while
[15:41] <greycat> but how much do I use on file sharing? "all of it".
[15:41] <Redb3ard> not what i mean
[15:41] <greycat> http://209.142.155.49/greycat.yi.org/graphs.cgi
[15:42] <Redb3ard> my 24 hours of freenet, i think i probably used less than 200k for myself
[15:42] <Redb3ard> the various index pages, and a few images that were on them
[15:42] * emacsen (~serge@pcp09296048pcs.arlngt01.va.comcast.net) has left #freenet
[15:42] <greycat> oh, you wanted to know "how much of the bandwidth that freenet utilized was for other people instead of for your personal gratification"?
[15:43] <Iakin> [19:51] <KenMan> Iakin : do your dips/drops show *any* correlation to the outgoingRequestInterval/minute diagnostics graph ?
[15:43] <Iakin> Nope, that one is _very_ smooth..
[15:43] <greycat> that's very difficult to answer, in general, because it doesn't log anything like "user made 17 requests, 5 of them were DNF, 12 were satisfied, total of 17.5 MB downloaded, 12.5 MB cached, 5.0 MB destroyed by pcaching"
[15:44] <hobx> http://www.math.chalmers.se/~ossa/temp/bb.jpg
[15:45] <Iakin> [21:35] <KenMan> kernel CPU or something ?
[15:45] <Iakin> yea, exactly.. forgot about that one ;)
[15:46] <Redb3ard> greycat, your personal gratification, actually
[15:46] <Redb3ard> and i didnt need an exact count of bytes
[15:46] <Redb3ard> just wanted a guess
[15:46] <Redb3ard> was curious
[15:47] <Iakin5> DataRequest 25156/527942
[15:47] <greycat> if you assume a "normal" week when I was using my node? I probably use a few megs a day worth of bandwidth, compared to a continuous 12kBytes/sec outgoing and god knows how much incoing.
[15:47] <Iakin5> QueryAborted 58/232
[15:47] <Iakin5> QueryRejected/Aborted query 95/19
[15:47] <Iakin5> QueryRejected/All routes failed. 2/0
[15:47] <Iakin5> QueryRejected/Cache failed 3/0
[15:47] <Iakin5> QueryRejected/DataInsert never received 31/7
[15:47] <Iakin5> QueryRejected/Looped request 296872/1601
[15:47] <Iakin5> QueryRejected/No route found 207746/7646
[15:47] <Iakin5> QueryRejected/Node overloaded 0/3026
[15:47] <Iakin5> QueryRejected/You sent bad data 0/1
[15:47] <Redb3ard> yeh, thats what i wanted to know
[15:48] <KenMan> poor hobx, he is worse off than me. But it makes me feel a little better. Thanks.
[15:48] <hobx> it is a sign of verility!
[15:49] <greycat> virility, but yes.
[15:49] <KenMan> yeeeaaaahhhh, that sounds good :(
[15:49] <Redb3ard> hobx, actually, its not a sign of virility, either
[15:49] <greycat> baldness is caused by testosterone
[15:49] <KenMan> shh, shut up. I am sticking with his theory, as it means my sex life will only improve...
[15:50] <Redb3ard> and virility is a specific balance of hormones, not a PPM of T in your bloodstream
[15:50] <hobx> Redb3ard: Be careful. People with big beardgrowth almost always go bald.
[15:50] <Redb3ard> older women have quite a bit of T, sometimes grow mustaches
[15:50] <Redb3ard> never heard of them being considered virile
[15:51] <hobx> ever fucked a woman with a mustache?
[15:51] <KenMan> that's cause you don't have to be around any of the butch types
[15:51] <Redb3ard> no, should i? are you suggesting she might get me pregnant?
[15:51] <KenMan> heh, she might f--- you, in any case...
[15:52] <hobx> Redb3ard: You worry a lot about that?
[15:52] <KenMan> yes, definitely rubberhole - iakin attracts so many queries in, they have no choice but too loop or RNF...
[15:53] <Iakin> [22:00] <Redb3ard> hey guys, how much bandwidth do you personally use on freenet in a week's time?
[15:53] * KenMan had assumed Redb3ard was male, but i suppose it could be a she with a red 'beard' ... :o
[15:53] <Iakin> Somewhere between 1 to 2MBit/s.. split over two nodes
[15:57] <Iakin> rubberhole is a good name.. seems like quite a bit of the unstable network's requests bounces around my node
[15:57] * oierw (mathew@cpe-69-75-112-240.hawaii.rr.com) Quit (Read error: 110 (Connection timed out))
[15:57] <Iakin> ..now and then hitting it and getting reflected
[15:58] <Iakin> Heh
[16:01] <Iakin> 'The National Geographic survey had thrown up the sad fact that only 23 out of 56 young americans knew the whereabouts of the Pacific Ocean'
[16:01] <Iakin> ..fascinating
[16:14] <Redb3ard> i find that hard to believe
[16:14] <Redb3ard> even as stupid as people are
[16:14] <Redb3ard> the numbers seem fudged so that they can sound sensationalist
[16:15] <salahx> heh
[16:15] <salahx> you'd be surpised
[16:15] <salahx> No more fudged than Freenet's estimators :)
[16:16] <salahx> by taking out/subsituting the the identifting words in the Declaration of Independs, peopel woudl probably think its some communist document
[16:17] <Redb3ard> i cant claim ive memorized it
[16:17] <Redb3ard> but i at least would recognize it
[16:17] <salahx> nor have I, but again, I'd recognize it
[16:28] * greycat (~wooledg@192.35.79.70) Quit ("Lick Bush in '04")
[17:13] * TLF (francisco@157.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[17:25] <leex-Zzzz-gone> unstable working better
[17:25] * leex-Zzzz-gone is now known as leex-R
[17:28] <hobx> 128000 nodes, max degree 20, with average search of 18.36
[17:29] <hobx> that's a pretty good search
[17:30] <leex-R> ? 128000 nodes on unstable
[17:30] <hobx> no
[17:31] <hobx> simulation of a much simpler algorithm.
[17:37] * Hory (~Miranda@81.196.25.110) Quit (Connection reset by peer)
[17:54] * Jarod (~kdx@cpe-024-211-212-102.ec.rr.com) Quit ()
[18:05] * mikeDOTd (~mikedotd@plutonium.nexthop.net) has joined #freenet
[18:12] * Brutha- (~x@pD9587E22.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
[18:13] * interrupt (~chatzilla@64.122.23.213) Quit ("Chatzilla 0.9.64f [Mozilla rv:1.7/20040707]")
[18:28] <leex-R> heh
[18:28] * leex-R is now known as leex-H
[18:34] * lolo-laptop (~lostlogic@68.251.84.226) Quit ("Client exiting")
[18:35] * Chinpokomon (sdf32es@81-6-220-3.dyn.gotadsl.co.uk) has joined #freenet
[18:37] * Chinpokomon (sdf32es@81-6-220-3.dyn.gotadsl.co.uk) Quit (Client Quit)
[18:38] * Chinpokomon (sdf32es@81-6-220-3.dyn.gotadsl.co.uk) has joined #freenet
[18:39] <iip_i2p> <lucky> not to troll, but do you think I2P might wipe out freenet? =/
[18:40] <salahx> naaah
[18:40] <salahx> the 2 have different goals
[18:40] <salahx> similar; but different
[18:40] <salahx> in fact most likely the two will work together
[18:41] <iip_i2p> <lucky> yea, but I2P has a distributed store (which in my understanding is similar to freenet.) doesn't that sort of have the ability to take over freenet's goal?
[18:42] <salahx> Nor sure, though the 2 projects will likely "borrow" a lot of ideas from each toher
[18:43] * Chinpokomon (sdf32es@81-6-220-3.dyn.gotadsl.co.uk) Quit ()
[18:43] <iip_i2p> <lucky> hm, yea. maybe this store will be more like gnunet?
[18:44] <iip_i2p> <lucky> actually, i sort of got gnunet working yesterday. heh, nice idea, but no content.
[18:47] * moskau23 (~Miranda@dsl-213-023-249-157.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[18:51] <leex-H> :)
[19:33] * TLF (francisco@157.Red-81-40-113.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[19:56] * bluephile (bluephile@69-160-205-193.clvdoh.adelphia.net) Quit (Client Quit)
[19:57] * bluephile (bluephile@69-160-205-193.clvdoh.adelphia.net) has joined #freenet
[20:08] * toad_ here, for half an hour or so...
[20:09] <toad_> <iip_i2p> <Newsbyte> well, in any case, I suspect you log the chan and search for your name, so, you'll find this, toad ;-) If you set a flag with a fixed 20% failure-chance, doesn't it become trivial to determine if an insert comes from a certain node? Everytime node a inserts an update of his freesite, there is a 20% failure; if node b forces it (through an adapted seednodes.ref) to use node c, also under his control, he can see if
[20:09] <toad_> <iip_i2p> <Newsbyte> If you can repeat that, say, 10 times, you'll become increasingly sure that node a is the one that inserts it. No?
[20:09] <toad_> no, you're misunderstanding... i think i explained...
[20:15] <toad_> <sanity_> and the UI makes me rather depressed for the prospect of Linux catching up, Gnome and KDE still seem to be playing catchup with Windows 95 in many respects - okay, that is simply not true 90% of the time. there are issues that aren't fully worked out yet sure, but there's TONS of stuff where we're close to or ahead of XP
[20:18] <toad_> <hobx> There is some interesting math here. Having a simulator helps show when it is worth exploring.
[20:18] <toad_> <sanity_> what is so interesting about O(log(N)) ?
[20:18] <toad_> <hobx> http://citeseer.ist.psu.edu/kleinberg00smallworld.html
[20:18] <toad_> <sanity_> smallworld is sublinear, not logarithmic, right?
[20:18] <toad_> <hobx> noop
[20:18] <toad_> <hobx> log(n)?
[20:18] <toad_> <hobx> so poly-log
[20:18] <toad_> <sanity_> oh wait
[20:18] <toad_> <sanity_> small world assumes optimal routing decisions - right?
[20:18] <toad_> <hobx> read that paper
[20:18] <toad_> <sanity_> i don't have time to read papers
[20:18] <toad_> <hobx> log(n)? is for pure greedy routing
[20:18] <toad_> sanity_ reads the abstract
[20:18] <toad_> <hobx> ie, route to the node that is closest in id to that you seek
[20:18] <toad_> <sanity> gah
[20:18] <toad_> <hobx> though he deals with ids as positions in a 2D grid, their are variants
[20:18] <toad_> <hobx> for one dimensional numerical difference.
[20:18] * toad_ DUHHH !
[20:18] <toad_> obvious solution to premix routing connectivity difficulties
[20:19] <toad_> does it still get log(n)^2 for largely random routes and one dimensional space?
[20:22] <toad_> <Redb3ard> too bad you cant make freenet always traverse at least 2 hops for any request
[20:22] <toad_> <Redb3ard> then you could have provable deniability
[20:23] <toad_> no, only if the first 2 hops don't know the request details
[20:23] <toad_> premix routing would do 2 hops
[20:23] <toad_> or maybe 3
[20:23] <toad_> <thelema> would it suffice to have a noDS flag on requests so that nodes recieving such a request wouldn't check their datastore before forwarding the request (with a fixed chance of removing that flag)? - umm, what for?
[20:24] <toad_> inserts?
[20:25] <toad_> <thelema> hirvox: we'd do noDS as an alternative to real onion routing. - uh, why would it help?
[20:28] <toad_> <KenMan> my biggest concern is that we do not consider the amount of resources allocated to us (mRI) in a direct manner when computing load / our outgoing resource allocations. It is a bear of a problem to figure out.
[20:28] <toad_> <KenMan> it is damned near impossible when routes are changing mRI values to us multiple times for each single request we send to them.
[20:28] <toad_> it is a horrible problem EVEN IF WE DON'T
[20:28] <toad_> but we have no chance while we are sending 1.5-2.5x as many queries as we receive and instantly RNFing many
[20:28] <toad_> <KenMan> there is a nasty feedback type of problem resulting from this behavior... one that would not be so dramatic if mRIs updates were less frequent than the rate of queries
[20:29] <toad_> i don't think so...
[20:29] <toad_> <supermannn> people were suggesting using 90 connections for splitfiles... is there a max limit? why not use 300 or something real high?
[20:29] <toad_> <KenMan> supermannn: go for it.
[20:29] <toad_> <supermannn> is there any reason not to?
[20:29] <toad_> there's a limit... i think it's 100
[20:30] <toad_> <supermannn> at 90 connections i got around 15KB/sec.....as i increase the connections it jumps faster. im up to 70KB/sec using 227 conn
[20:30] <toad_> woah
[20:30] <toad_> that's seriously slow... lots of retries?
[20:32] <toad_> <supermannn> id just like to get an idea of how many people are there
[20:33] <toad_> actually, we could probably do a reasonable distributed estimate with some code
[20:34] <toad_> based on the idea i spewed a while ago... pick a random node from your RT, whose ID isn't too close to yours, send a packet to another node, tell it to forward it to that node, by routing through its own links (closest ID), and the next node doing the same and so on until it reaches the destination
[20:34] <toad_> do this a bunch of times from different nodes and count the hops
[20:35] <toad_> according to something hobx showed, it the hops should be proportional to log(n)^2
[20:35] <toad_> where n is the total number of (active?) nodes
[20:35] <toad_> of course in practice some nodes are more active than others but it may still be a possibility...
[20:39] <toad_> <Iakin> My node's lQPH is about 60-70k whilst outboundAggregateRequests hover at around 7-8k/hour - woah...
[20:40] <toad_> <supermannn> if theyre funding tor, why wouldnt they fund freenet - iirc tor is centralized and unscalable... i could be wrong...
[20:41] <toad_> <sanity_> http://p2pnet.net/story/2193 - woah, has this been posted to chat yet?
[20:42] <toad_> <sanity_> but it is hard to imagine induce going through as written given the mounting opposition - umm, we'll see... there's overwhelming opposition to software patents in the EU, and we'll probably still get them eventually...
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.
