Timestamps are in GMT/BST.
[0:10] * kukuks (i=mixderax@) has joined #freenet
[0:11] <TheSeeker> hmm, I just did an svn update, and got 10513 :/
[0:12] <CIA-5> toad * r10515 /trunk/freenet/src/freenet/node/NodeDispatcher.java: Probe request fixes.
[0:12] <toad_> TheSeeker: hmmm
[0:12] * TheSeeker tries again
[0:14] <TheSeeker> hmm, 10515 now, seems to be working.
[0:19] * MineHaunter (n=van85@) Quit ()
[0:20] * kakaks (i=mixderax@) Quit (Remote closed the connection)
[0:20] <TheSeeker> toad_: should the error on startup: "Could not initialize class com.onionnetworks.fec.Native8Code" be fixed now?
[0:21] <toad_> TheSeeker: you are getting it?
[0:21] <toad_> TheSeeker: does it include a stack trace?
[0:21] * Zhay (n=zhay@) has joined #freenet
[0:21] <Zhay> Y'all...!
[0:22] <TheSeeker> toad_: just "null" .... have always gotten it under 64 bit JVM, since I can't use the native libs.
[0:23] <TheSeeker> more accurately, I get:
[0:23] <TheSeeker> null
[0:23] <TheSeeker> Could not initialize class com.onionnetworks.fec.Native8Code
[0:23] <toad_> TheSeeker: odd
[0:23] <toad_> TheSeeker: there are 64-bit libs in the freenet-ext.jar
[0:23] <TheSeeker> Only the first time my node starts an insert
[0:23] <Zhay> With the new freenet, is communication via FCP encrypted?
[0:23] <toad_> Griffon26 was getting it on 32-bit
[0:23] <TheSeeker> toad_: not for Windows
[0:24] <toad_> Zhay: no, why should it be?
[0:24] <toad_> TheSeeker: ahhhh
[0:24] <Zhay> toad_: Like, if a node is running on a remote server over the internet.
[0:25] <toad_> Zhay: that's not generally a good idea
[0:25] <toad_> Zhay: FCP has access to the local filesystem for efficiency reasons
[0:25] <toad_> but even if it didn't, it's still not a very good idea
[0:25] <Zhay> k.
[0:25] <toad_> ssh tunnel it if you ultimately trust both ends
[0:29] <Zhay> Hmm. Through an ssh tunnel, though, isn't it appropriately anonymous? Or is there a problem with that sort of setup?
[0:29] * Werdna (n=Andrew@) Quit (Read error: 131 (Connection reset by peer))
[0:33] * Werdna (n=Andrew@) has joined #freenet
[0:36] <toad_> Zhay: appropriately anonymous? eh?
[0:37] <Zhay> toad_: client ==ssh==> node
[0:37] <toad_> yeah, so?
[0:37] <Zhay> ...
[0:37] <toad_> if you want anonymity, you control both the node and the client
[0:38] <Zhay> O...kay.
[0:40] <toad_> Zhay: you have a freenet *node*
[0:40] <toad_> that gives you anonymous access to the network
[0:40] <toad_> the server/client thing is just for convenience
[0:41] <toad_> so we can have lots of different clients
[0:41] <toad_> they'll normally run on localhost
[0:41] <toad_> so you have Freenet Client Protocol for clients to talk to the node
[0:41] <toad_> and Freenet Node Protocol for nodes to talk to other nodes
[0:41] <toad_> good night
[0:41] * toad_ (n=toad@) Quit (Remote closed the connection)
[0:41] <TheSeeker> toad: after updating, I get a "Failed to load native FEC: java.lang.ExceptionInInitializerError" along with a stacktrace, in place of the 'null' -> 'could not initialize...' lines. seems to be working as intended.
[0:42] <TheSeeker> d'oh
[0:42] <Zhay> What an ass. :P
[0:42] <Zhay> Toodles!
[0:43] * Zhay (n=zhay@) has left #freenet
[0:51] <alphanimal> anybody knows what that is all about:
[0:51] <alphanimal> http://127.0.0.1:8888/freenet:USK@I~r8qww4X-kMu0kogs2eNSWVDYcOYjo8l2y6ZxlYhRg,thL8FR2OBVchN6YSJ~8gAm6qkQQZBkxefDleY-gcMQY,AQABAAE/NMW/2/
[0:53] <alphanimal> onpe?
[1:00] * railk (n=railk@) has joined #freenet
[1:08] * Zothar_Work (n=chatzill@) Quit ("Chatzilla 0.9.74 [Firefox 1.5.0.7/2006090921]")
[1:09] * blibbet (n=blibbet_@) Quit ("Leaving")
[1:10] * JustMe (i=JustMe_@) has joined #freenet
[1:17] * auc1 (n=chatzill@) Quit (Remote closed the connection)
[1:27] <alphanimal> anyone still here to help?
[1:31] <agsarite> it doesn't look like it
[1:37] <alphanimal> forgot my question though
[1:48] <SinnerG> lol :p
[1:58] * JustMe (i=JustMe_@) Quit ()
[2:01] <alphanimal> ok, i got 14 connections right now but its still very slow... is that normal?
[2:02] * JustMe (i=JustMe_@) has joined #freenet
[2:03] <agsarite> if it's the data you're after isn't cached, then i compare freenet to the speed of dialup, with 5000x the latency.
[2:04] <agsarite> stoopid grammar.
[2:06] <alphanimal> could be true
[2:15] * railk (n=railk@) Quit (Remote closed the connection)
[2:17] * Gasi_ (n=chatzill@) has joined #freenet
[2:23] * n0ob (n=travis@) Quit ("off work")
[2:33] * Werdna (n=Andrew@) Quit (Read error: 104 (Connection reset by peer))
[2:35] * Gasi (n=chatzill@) Quit (Read error: 110 (Connection timed out))
[2:46] <MrNaz> agsarite: i hate grammer and speling to
[2:55] * alphanimal (n=Miranda@) Quit (Read error: 54 (Connection reset by peer))
[3:02] * ShoeyFighter (n=chertel@) Quit ("Ex-Chat")
[3:31] * Werdna (n=Andrew@) has joined #freenet
[3:54] * MrNaz (n=naz@) Quit (Read error: 110 (Connection timed out))
[3:54] * MrNaz (n=naz@) has joined #freenet
[4:24] * Werdna (n=Andrew@) Quit (Remote closed the connection)
[4:25] * Werdna_ (n=Andrew@) has joined #freenet
[4:26] * Werdna_ is now known as Werdna
[4:53] * smartone66 (n=boo@) Quit (Read error: 60 (Operation timed out))
[5:00] * ano is now known as Gorai
[5:22] <agsarite> damn, i'm drawing when i should be coding.
[5:23] * noggly_ (i=noggly@) has joined #freenet
[5:28] <agsarite> KSK@toad
[5:31] * Werdna (n=Andrew@) Quit (Remote closed the connection)
[5:34] * noggly (i=noggly@) Quit (Read error: 110 (Connection timed out))
[5:46] * Werdna (n=Andrew@) has joined #freenet
[5:54] * basxx (n=basxx@) has joined #freenet
[5:58] * basxx (n=basxx@) has left #freenet
[5:59] <TheSeeker> I think there's a bug in the queue page... I have a lot of inserts going, and the first 7 were at decreasing priority (starting at emergency) with the rest at 'will never finish'... when the highest priority upload finished, the priorities of the other 6 changed... at least according the the queue page... the colors match the correct order/priority, but the drop down boxes don't match, and the order of the files are reversed...
[6:00] <TheSeeker> refreshing fuqid by deactivating and reactivating doesn't show a change in priority though.
[6:01] <TheSeeker> hmm, closing firefox and reloading the page after releasding fixed it, where refreshing the page would not ... stupid caching :/
[6:06] * phrosty (n=phrosty@) has joined #freenet
[6:29] * sitharus_ is now known as sitharus
[6:47] * slinky_pocuscada is now known as slinky
[6:48] * slinky (n=slinky@) Quit (Read error: 104 (Connection reset by peer))
[6:50] * slinky (n=slinky@) has joined #freenet
[7:19] * PhrostByte (n=phrosty@) has joined #freenet
[7:23] * phrosty (n=phrosty@) Quit (Read error: 104 (Connection reset by peer))
[7:29] * PhrostByte is now known as phrosty
[7:44] * MrNaz (n=naz@) Quit (Connection timed out)
[7:45] * MrNaz (n=naz@) has joined #freenet
[7:47] * Werdna (n=Andrew@) Quit ("Leaving")
[7:53] * phrosty (n=phrosty@) Quit ("baseball is wrong: man with four balls cannot walk.")
[7:59] * SinnerG (i=SinnerG@) Quit (Read error: 60 (Operation timed out))
[8:09] * Jflesch (n=Jflesch@) has joined #freenet
[8:10] <ljn1981> Morning guys
[8:11] <agsarite> morn'in.
[8:30] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[8:37] <ljn1981> toad, nextgens, anyone with a clue about native FEC: http://dark-code.bulix.org/lj37d4-18438?raw OS X 10.3 on PPC G4
[8:38] <ljn1981> r10515
[9:04] * Werdna (n=Andrew@) has joined #freenet
[9:19] * Gruick (n=stephpom@) has joined #freenet
[9:21] * hdp (n=hdp@) has joined #freenet
[9:24] * Gruick (n=stephpom@) has left #freenet
[9:40] * sbc (n=sbc@) has joined #freenet
[10:04] <sbc> toad: Got your freemail "Hello, testing..." So something is working :)
[10:05] <TheSeeker> ljn1981: from what I can tell, if you don't have a native FEC binary and have to use the pure java version, that's the error you'll get when it tries to initialize the fec class.
[10:06] <ljn1981> Yeah
[10:07] <ljn1981> But I was under the impression I did but I might be confusing it with when I got nativbigint some months ago
[10:08] * timmy2chk (n=Vincent@) has joined #freenet
[10:13] <TheSeeker> *nod* ... I don't get either... 'cause I'm on 64 bit Windows, and nobody wants to develop for that platform. :P
[10:28] <hdp> hi, running freenet (977) over a long period of time makes my cpu load going nuts. is this a known problem? i have ~ 10 active/inactive peers, cpu is 2.4ghz
[10:49] * toad_ (n=toad@) has joined #freenet
[10:49] * ChanServ sets mode +o toad_
[10:50] <toad_> LOL @ comments on http://slashdot.org/article.pl?sid=06/08/29/0441202 - "is it anonymous?!"
[10:50] <toad_> when we get /.ed everyone says "it is anonymous, therefore it sucks"
[10:51] <toad_> when something else gets /.ed everyone says "it's not anonymous, therefore it sucks"
[10:51] <toad_> slashdotters are automatically contrarian...
[10:52] <ljn1981> Morinng toad_
[10:52] <ljn1981> toad_: http://dark-code.bulix.org/lj37d4-18438?raw OS X 10.3 on PPC G4
[10:52] <ljn1981> r10515
[10:52] <ljn1981> toad_: Well everything must suck then.
[10:52] <ljn1981> toad_: And I'm running another PROBEALL. too
[10:52] <ljn1981> So far I'm at 109 nodes
[10:53] <toad_> <ljn1981> toad_: Well everything must suck then. ------- i think that's the idea
[10:53] <toad_> ljn1981: known issue
[10:54] <toad_> ljn1981: will take it out when sorted out Griffon26's FEC-doesn't-load-on-x86 problem...
[10:54] <ljn1981> Ok
[10:58] <_ph00> can I get a direct link to a freemail.jar download plz?
[10:58] <_ph00> can't find it
[10:59] <_ph00> # networkSizeEstimateSession: 433 nodes
[10:59] <_ph00> # nodeUptime: 18h48m
[11:00] <_ph00> never got so high node estimate before
[11:00] <_ph00> must mean that freenet is growing bigger, or that my node is 'learning' about more nodes (or both)
[11:00] <_ph00> do nodes 'learn' about other nodes?
[11:08] <ljn1981> toad_: http://dark-code.bulix.org/jquxr1-18446?raw
[11:09] <ljn1981> _ph00: What do you mean learn?
[11:09] <TheSeeker> _ph00: sort of, they learn about locations. during node swap requests, you get a location of a node that wants to swap, along with the locations of all it's peers. the esitmate is based on the number of unique node locations seen in a given period.
[11:14] <_ph00> ljn: I mean like, once a node have 'found' another node, does it get easier to find that same node again after a restart?
[11:14] <ljn1981> _ph00: No.
[11:14] <_ph00> anyone knows where to download freemeil? I don't think I have it in my freenet dir
[11:14] <_ph00> ljn, so my highest-than-ever node estimates must mean that freenet is growing bigger
[11:17] <_ph00> damn
[11:17] <_ph00> gotta go
[11:17] <_ph00> bbl
[11:17] <ljn1981> _ph00: Or a lot of nodes join the net and quickly leave
[11:17] <_ph00> hm
[11:17] <_ph00> possible
[11:17] <ljn1981> Other estimates sayd about 100 nodes at any given time
[11:17] <ljn1981> But the code to get those might have bugs
[11:18] <_ph00> but nodes joining and leaving should be doing that all the time
[11:18] <ljn1981> doing what?
[11:18] <_ph00> I never had 430 nodes estimate before, that's why I assumed that freenet was growing bigger
[11:18] <sleon> # networkSizeEstimateSession: 547 nodes
[11:18] <sleon> # networkSizeEstimateRecent: 514 nodes
[11:18] <sleon> :=
[11:18] <ljn1981> It might be
[11:18] <_ph00> that's your node
[11:18] <_ph00> sleon I mean
[11:18] <_ph00> 514
[11:19] <_ph00> I never had more than 360-380 before
[11:19] <_ph00> anyway
[11:19] <ljn1981> networkSizeEstimateSession: 257 nodes
[11:19] <ljn1981> Based on just short of 3 hours
[11:19] <_ph00> got to go right now, bbl
[11:19] <sleon> _ph00: i have had over 1000 already
[11:19] <sleon> _ph00: with uptime of one week or so
[11:19] <ljn1981> sleon: That was session, not revent, right?
[11:20] <sleon> yes
[11:20] <ljn1981> recent*
[11:20] <ljn1981> Recent only count ones that have been seen in the last 48 hours, more precise.
[11:21] <sleon> ljn1981: but still, 1000 nodes! :)
[11:21] <ljn1981> Hehe
[11:27] <TheSeeker> wow, it's really been 11 hours since I upgraded to 10515? Time flies. I have a node estimate of 355 and was about to say "I've only been up a few hours but..." :P
[11:28] <ljn1981> Hehe
[11:28] <TheSeeker> Hmm... could we get a stat somewhere that keeps track of how many times we've swapped vs attempted to swap (local and remote) and/or a list of the locations we've had this session?
[11:29] <TheSeeker> I know this could be figured out from the logs... but damn that would be a lot of work :P
[11:30] <sleon> TheSeeker: not at all!
[11:30] <sleon> TheSeeker: only some zgrep/sed kung fu
[11:30] <MrNaz> _ph00: can i exchange refs with you?
[11:30] <TheSeeker> and mamoth log file sizes die to having to log on minor
[11:30] <MrNaz> i still only have one hehe
[11:30] <TheSeeker> *mammoth
[11:31] <sleon> TheSeeker: it takes only 20 minutes to compute :D
[11:35] <CIA-5> toad * r10516 /trunk/freenet/src/freenet/crypt/DiffieHellman.java: Run genParams in-thread if necessary.
[11:35] <ljn1981> Would it be useful to have on the darknet page? It can't be that much work to do but do we want to have it?
[11:35] <ljn1981> With all the stats we're starting to have, wouldn't we want to move them to a dedicated stats page instead?
[11:40] <TheSeeker> I miss all the pretty graphs that 0.5 had :P
[11:40] <ljn1981> In most cases I don't but in some cases graphs would be nice
[11:41] <TheSeeker> not understanding what half of them meant is beside the point... graphical representation of infromation over time is neat. :D
[11:41] <_ph00> back
[11:41] <_ph00> (10 min only)
[11:42] <ljn1981> wb
[11:43] <_ph00> did I get that right? Isthe right way to upgrade to latest testing renaming freenet-latest-testing.jar to freenet.jar?
[11:43] <ljn1981> Yes
[11:43] <_ph00> k
[11:44] <ljn1981> ./upgade.sh testing && mv freenet-latest-testing.jar freenet.jar && ./run.sh restart
[11:44] <sleon> TheSeeker: yea
[11:45] <ljn1981> Or something like that... Think I remembered some names wrong
[11:45] * bla (n=paul@) has left #freenet
[11:45] <ljn1981> ./upaate.sh testing && mv freenet-testing-latest.jar freenet.jar && ./run.sh restart
[11:45] <ljn1981> There
[11:45] <_ph00> thx
[11:46] <ljn1981> duh
[11:46] <ljn1981> ./update.sh testing && mv freenet-testing-latest.jar freenet.jar && ./run.sh restart
[11:46] <_ph00> I already did it 'the windoze way' (i.e. using the GUI)
[11:46] <ljn1981> I spell as bad as you today.....
[11:46] <_ph00> only the donload I did by ./update.sh testing
[11:46] <_ph00> heh
[11:46] <ljn1981> aha
[11:46] <_ph00> my problem is not really spelling, more typing
[11:47] <ljn1981> Same here
[11:47] <_ph00> but actually, one line (with &&) is better
[11:47] <_ph00> If I only remembered that it can be done, instead to go straight for the gui
[11:47] <ljn1981> Yeah much easier
[11:47] <_ph00> years of windoze do damage the brain
[11:48] <ljn1981> Very much so
[11:48] <_ph00> k, gotta go again, back in 20 min or so
[11:48] <_ph00> Freenet 0.7 Build #977 r10516
[11:49] <_ph00> tah-dah!
[11:49] <_ph00> *gone
[11:58] * kakaks (i=kakaks@) has joined #freenet
[11:58] <MrNaz> so what does freenet mean by "darknet" ?
[11:59] <_ion> You only connect to people you trust.
[11:59] <MrNaz> right...
[12:06] <MrNaz> ok so ive added a half dozen or so references from freenet-refs
[12:06] <MrNaz> but after a short while they all disconnect from me
[12:06] <MrNaz> why is that?
[12:06] <toad_> maybe they uninstalled freenet?
[12:07] <MrNaz> well its always happened... they connect and a short while later they disconnect
[12:07] <MrNaz> toad_: is your freenet running? you appear to be disconnected
[12:08] <MrNaz> heh i removed all the refs peers... you're once again my only reference haha
[12:08] <toad_> yes
[12:08] <toad_> oh
[12:08] <toad_> so it's your fault?
[12:09] <MrNaz> yeap
[12:09] <toad_> your node shows up as DISCONNECTED
[12:09] <MrNaz> must be...
[12:09] <MrNaz> yea same for yours... the funny thing is that when i add someone, they show as connected for a while and then disconnected
[12:09] <MrNaz> by a while i mean a few minutes
[12:09] <toad_> you removed me and that's why i show up as disconnected?
[12:10] <MrNaz> no, you're still there
[12:10] <TheSeeker> Pentagon proposed budget, $480Bln ... is that per year? o.O
[12:10] <toad_> TheSeeker: yes
[12:10] <MrNaz> TheSeeker: yes
[12:10] <toad_> TheSeeker: it'll be higher if they invade Iran
[12:10] <MrNaz> toad_: dont you mean "when" ?
[12:10] <MrNaz> hehe
[12:10] <toad_> well, i'm still hoping that somebody somewhere may see reason
[12:11] <toad_> MrNaz: are you behind a NAT of any sort?
[12:11] <MrNaz> i heard that hope expressed before they invaded iraq
[12:11] <toad_> MrNaz: indeed
[12:11] <MrNaz> i'm beind an IPCop box...
[12:11] <TheSeeker> sheesh. What kind of publics works projects could be done to improve lives in america if 10B per year were given to every state? :P
[12:11] <toad_> MrNaz: well, I gave them the benefit of the doubt about iraq
[12:11] <toad_> and i don't think we can pull out now
[12:11] <toad_> but invading iran at this point would be insane
[12:12] <toad_> especially as cheney had the opportunity to sort out all the problems with iraq, and refused do
[12:12] <MrNaz> TheSeeker: how about how many famine stricken countries could be fed on $2billion? for that price you could feed all of africa for a year
[12:12] <toad_> to
[12:12] <toad_> that means it's entirely his fault
[12:12] <toad_> well maybe not entirely
[12:12] <toad_> but substantially
[12:12] <toad_> s/iraq/iran doh
[12:13] <MrNaz> toad_: cheney is making too much money from his kickbacks from his halliburton buddy to have ANY interest in founding a viable state in iraq
[12:13] <toad_> cheney had the opportunity to sort out all the problems with iran, and refused to
[12:13] <MrNaz> s/buddy/buddies
[12:13] <toad_> MrNaz: i disagree on that one. fighting states is much more fun than peacekeeping. and they'll probably have to pull out of iraq in order to invade iran.
[12:13] <toad_> MrNaz: which would likely result in all hell breaking loose across the middle east
[12:13] <toad_> which presumably is exactly what cheney wants
[12:14] <toad_> MrNaz: anyway re freenet
[12:14] <MrNaz> toad_: hang on... what part do you disagree with?
[12:14] <TheSeeker> toad_: not if the draft is re-instated... :P
[12:14] <toad_> MrNaz: cheney needs a stable iraq so he can invade iran
[12:15] <toad_> but maybe he'll somehow manage to do it anyway
[12:15] <toad_> they can't reinstate the draft until AFTER the election
[12:15] <toad_> but they have to start the war BEFORE it
[12:15] <toad_> (in order to win it)
[12:15] <toad_> the election i mean
[12:16] <TheSeeker> and yet, we "need to be diplomatic with N. Korea" ...
[12:16] <toad_> TheSeeker: that's because N Korea really does have nukes
[12:16] <TheSeeker> heh
[12:16] <toad_> it's true, you don't attack anyone who really has nukes
[12:16] <toad_> that's pretty basic
[12:17] <_ion> I heard a rumour that a certain country called United States of America has weapons of mass destruction.
[12:17] <MrNaz> toad_: see i dont believe the election means anything... democrat / republicans may differ on minor budget points, but really, with regards to military decisions, defense contractors and unelected pentagon officers call the shots, not elected officials (who i believe are nothing but puppets and showmen anyway)
[12:17] <TheSeeker> they don't have the means to launch them yet... shouldn't we be doing something while their test firings are still failing? :P
[12:17] <toad_> MrNaz: i don't know about that. clinton was very reluctant to risk american lives.
[12:17] <toad_> TheSeeker: do they have to launch them?
[12:18] <toad_> how hard would it be to freight one in to mexico and smuggle it across the border?
[12:18] <MrNaz> toad_: and yet the clinton administration went to war in kosovo and led a constant low grade campaign in iraq
[12:18] <toad_> MrNaz: which killed many many civilians, and very few americans
[12:19] <MrNaz> if you ask me, the only difference between democrat and republican govts is ostensible reasoning... the real decision makers are not elected, and are not visible
[12:19] * TheSeeker looks for suitcase nuke conspiracy theories.
[12:19] <toad_> i don't think al gore would have gone into iraq
[12:19] <MrNaz> toad_: which is why he wasnt elected
[12:19] <toad_> anyone would have done afghanistan
[12:20] <MrNaz> toad_: why do you say that?
[12:20] <toad_> perpetual war ...
[12:20] <toad_> MrNaz: umm, because osama bin laden was in afghanistan?
[12:21] <MrNaz> to this day, the FBI states there is no hard evidence implicating bin laden... furthermore, plans to invade afghanistan were drawn up long before 9/11 ... personally, i think 9/11 gave the defence industry an excuse to do what theyd wanted to do for years already
[12:21] <toad_> that's nonsense
[12:21] * kakaks (i=kakaks@) has left #freenet
[12:21] <toad_> he's taken responsibility numerous times
[12:21] <toad_> and they wouldn't give you the evidence if they had it
[12:21] <MrNaz> toad_: url references to interviews?
[12:21] <toad_> that's not to say that they didn't run with it; of course they did
[12:21] <toad_> MrNaz: every video he's put out since then
[12:22] <MrNaz> you have a url to one of them? the few that were released are all questionable
[12:22] <toad_> the point is, invading iran would be deliberately escalating the situation
[12:22] <MrNaz> toad_: well... i think thats the point
[12:22] <toad_> there may be a case for invading iran on the nuclear issue, but right now they are nowhere near having nukes, and nobody has even tried sanctions yet
[12:23] <toad_> and i don't see how the americans can possibly hold both iran and iraq at the same time
[12:23] <MrNaz> starting a global war is in the intestests of a superpower that is attempting to gain hegemonic power
[12:23] <toad_> well they didn't start it
[12:23] <toad_> but they're not interested in solving it
[12:23] <MrNaz> they didnt?
[12:24] <toad_> and they seem very interested in fucking up the battle for hearts and minds
[12:24] <toad_> radicalising all muslims everywhere
[12:24] <toad_> and generally escalating the situation
[12:24] <MrNaz> well...
[12:24] <MrNaz> full disclosure: im a muslim
[12:24] <toad_> i'm not sure that abu ghraib was deliberate, but i'm not sure it wasn't
[12:24] <toad_> yay, first time i've met one through freenet :)
[12:25] <toad_> well, i don't suppose it was deliberate really ... but it was really stupid, and there was no attempt made to prevent it
[12:25] <MrNaz> my disagreement with US foreign policy has nothing to do with religion, and everything to do with abuse of power and the concentration of wealth while people starve
[12:25] <toad_> you don't have any particular issue with the US annexing muslim lands left right and center?
[12:25] <toad_> or you do, but not as a religious matter
[12:25] <toad_> fair enough
[12:26] <MrNaz> as i said, not on the grounds of religion
[12:26] <toad_> i agree 100% about concentration of wealth; the US spends 0.1% of its GDP on aid, which is ludicrous
[12:26] <MrNaz> annexing land unjustly is immoral and wrong if you're a muslim, a christian or a flying speghetti monster believer
[12:26] <MrNaz> haha
[12:26] <toad_> and it's at the forefront (with us, the UK) in trade injustice
[12:27] <MrNaz> USAID is *not* a humanitarian organisation... "we'll give you the money you need to feed your people if you let our corporations buy out your national infrastructure"
[12:27] <toad_> right, well... i did give the US the benefit of the doubt with iraq, i don't know whether it was worth doing it for humanitarian reasons, it depends how it pans out
[12:27] <toad_> MrNaz: right, that's the basic deal
[12:27] <MrNaz> if there was any doubt to give the benefot of, its gone
[12:28] <toad_> but now it looks like they're going to annex the entire middle east to get the oil before china does
[12:28] <toad_> this not only ensures that the WoT gets a lot worse, and that they're bogged down forever
[12:28] <MrNaz> the US invasion has resulted in more deaths EVERY MONTH than saddam was responsible for in his entire reign
[12:28] <toad_> it also guarantees a new cold war with china - one which the US *WILL LOSE*
[12:28] <toad_> MrNaz: that's bullshit
[12:28] <MrNaz> it really isnt...
[12:28] <toad_> MrNaz: saddam started the war in the 80s which killed a million people
[12:28] <MrNaz> now THATS bullshit
[12:28] <MrNaz> a million is hyperbole
[12:28] <toad_> saddam was also substantially responsible for the sanctions in the 90s
[12:29] <toad_> which also killed a million people
[12:29] <MrNaz> well
[12:29] <MrNaz> that depends
[12:29] <toad_> we were responsible too, sure
[12:29] <toad_> but if he hadn't invaded kuwait it would never have happened
[12:29] <MrNaz> saddam discussed the annexing of kuwait with the US and was given the go ahead... after all, kuwait was part of iraq not long before that
[12:29] <toad_> i've heard mixed messages about that
[12:29] <MrNaz> so who was responsible for the kuwait invasion?
[12:30] <toad_> but the bottom line is one UN member does not invade another UN member
[12:30] <toad_> it's outrageous
[12:30] <MrNaz> right... and the US is in iraq because...?
[12:30] <toad_> :)
[12:30] <MrNaz> hehe
[12:30] <MrNaz> besides
[12:30] <MrNaz> there's many, many precendents from history
[12:31] <toad_> most things have been done at some point or other
[12:31] <toad_> that doesn't mean they were right
[12:31] <MrNaz> the reasoning behind the iraq invasion may be difficult to divine now as it is still current and not all info has reached the public as yet
[12:31] <toad_> well, we know the reasons that were put out at the time
[12:31] <toad_> we know many of the real reasons too
[12:31] <MrNaz> but lets go back into history and look at issues where all info has reached the surface
[12:31] <_ph00> FreeWorld (tm) vs EvilArabs(tm) is a fake
[12:32] <toad_> _ph00: both osama bin laden and dick cheney are doing everything in their power to make it a reality
[12:32] <_ph00> well, it *is* real
[12:32] <_ph00> but I think that it's not 'really' real...
[12:32] <_ph00> like
[12:32] <toad_> obviously this involves sacrificing the "Free" in "FreeWorld"
[12:32] <_ph00> FreeWorld(tm) doesnt really want to stop terror
[12:33] <MrNaz> i wrote an essay that had a few examples of exploitation in history... have a look at http://www.mrnaz.com/?mode=journal&entryid=122
[12:33] <toad_> well ... it sure looks that way sometimes
[12:33] <_ph00> the 'free' in FreeWorld(tm) does not really stands for 'freedom'. we're not free at all. it's just
[12:33] <_ph00> advertising
[12:33] <_ph00> propaganda
[12:34] <toad_> MrNaz: re economics, the Irish Potato Famine is the classic example
[12:34] <_ph00> any link to irish potato famine?
[12:34] <toad_> _ph00: this is the land of the free. the free no longer have any use for you. go to iraq!
[12:34] * _ph00 searches wikipedia
[12:34] <toad_> well first there was a famine
[12:34] <toad_> natural, biological problem
[12:34] <TheSeeker> The US war machine is like a drug company ... there is certainly no desire to solve the problems that CAUSE the problems... just to make the tools to supress the symptoms just long enough to make sure that you need to buy more on a regular basis...
[12:34] <toad_> then there was capitalism
[12:35] <_ph00> toad_ I'm not saying that Iraq is more free than FreeWorld(tm). What I'm saying is that FreeWorld(tm) is not so free after all
[12:35] <toad_> result: everyone starves or emigrates, food is exported while nobody can eat, everyone sells their lands, all the peasant farmers sell up or starve
[12:35] <MrNaz> TheSeeker: yes... profits in the military industrial complex plays too large a part in US foreign policy
[12:35] <toad_> all the land is concentrated into a few rich owners
[12:35] <toad_> landgrab by means of "free trade"
[12:35] <toad_> _ph00: sure
[12:36] <toad_> argument at the time was we mustn't impede the market
[12:36] <MrNaz> toad_: and there you have anothe example of "economic rationalism" at its giddy best
[12:36] <toad_> MrNaz: right
[12:36] <_ph00> so don't givge me the old 'go to iraq' crap. you know what I'm talking about
[12:36] <toad_> _ph00: sorry, it was sarcasm
[12:36] <_ph00> oops
[12:36] <_ph00> I didn't get that
[12:36] <MrNaz> toad_: if you can be bothered... read that essay i wrote... it deals with that topic specifically
[12:37] <_ph00> as I often get that kind of answers form a lot odf people who actually believe that FreeWorld(tm) is the best possible way...
[12:37] <toad_> well
[12:37] <toad_> I like the freedom we have in the free world
[12:37] <toad_> the rule of law, the constitution that makes some kind of sense
[12:37] <_ph00> better than iraq...
[12:37] <_ph00> better than afghanistan
[12:37] <toad_> the state providing for the needs of its poorest people
[12:37] <toad_> etc
[12:38] <toad_> unfortunately it's constantly under attack
[12:38] <toad_> and thanks to the war on terror we seem to be losing at the moment
[12:38] <_ph00> but not real freedom anyway, and that little we have is rapidly disappearing before our very eyes
[12:38] <MrNaz> toad_: the rule of law is now nothing but a mirage... the rule of law has degenerated into the rule of him with the most expensive lawyer
[12:38] <toad_> MrNaz: not always
[12:38] <toad_> sometimes we win
[12:38] <MrNaz> as a businessman, i can tell you that this is the case, first hand.
[12:38] <toad_> even in parliament
[12:39] <toad_> last year e.g. we (the FFII et al) blocked the software patents directive
[12:39] <toad_> that's a very specific and narrow example
[12:39] <MrNaz> yea... but when you lose more battles then you win, you are losing the war, and to tell yourself otherwise is wishful thinking
[12:39] <_ph00> war on terror 'we' seem to be losing? who are 'we'? FreeWorld(tm) govts? they don't want the 'war' on terror to stop. they need it to pass more fascist laws
[12:39] <toad_> it was supported by all the big businesses
[12:39] <toad_> we were supported by all the small ones
[12:39] <toad_> and we won!
[12:39] <toad_> _ph00: we = the citizenry
[12:39] <_ph00> then again
[12:39] <_ph00> ok, then
[12:40] <_ph00> you are way off track
[12:40] <toad_> okay
[12:40] <_ph00> we= the people are being screwed
[12:40] <toad_> i'm gonna go get lunch
[12:40] <_ph00> well OK, that's what I think, of course
[12:40] <MrNaz> toad_: won? watch your back... corporations are legally immortal, and so are their agendas.
[12:40] <toad_> when I come back, maybe can have a look at MrNaz 's connection problem ?
[12:40] <MrNaz> haha
[12:40] <toad_> MrNaz: sure, they're bringing another pass this year
[12:40] <_ph00> you can thing what you want, I mean, you are way off trac *for me*
[12:40] <_ph00> k
[12:40] <MrNaz> yes... connection problem... talk about getting side tracked
[12:40] <toad_> MrNaz: and it's more fun than ever; go to jail for patent infringement, or even for *inciting patent infringement*
[12:41] <_ph00> and "thinK"*
[12:41] * greycat (i=rfc1413@) has joined #freenet
[12:41] <_ph00> is "inciting patent infringment" a crime in UK?
[12:41] <_ph00> "inciting crime"?
[12:41] <toad_> (IPRED2, also EPLA, which is "legalize software patents", and the London Agreement, which is "you are bound by patents written in french")
[12:41] <toad_> _ph00: not now
[12:41] <toad_> _ph00: patent infringement is normally a civil matter
[12:42] <toad_> _ph00: which is understandable given that there are patents on e.g. the progress bar, buying things over the internet
[12:42] <_ph00> in many other EU countries the"inciting" crime has been abolised as "obsolete" or something like that
[12:42] <toad_> _ph00: under IPRED2 it would be a crime, and inciting, attempting, aiding or abetting it would also be a crime.
[12:42] <_ph00> but I'm afraid it'll be back soon enough
[12:42] <MrNaz> if you ask me... criminal law in most countries has gone haywire and needs to be scrapped and started again from scratch
[12:42] <toad_> well ...
[12:42] <toad_> IPRED2 is certainly insane criminal law
[12:43] <toad_> anyway
[12:43] <toad_> bbl
[12:43] <_ph00> yeah. I did read someting about ipred2. that's pure SHIT. we should mass protest such laws. problem is: how do you get ppl to understand that they need to protest?
[12:43] <_ph00> they're get so much gov propaganda thru tv and papers, that they actually believe that ctap is *right*
[12:44] <_ph00> <MrNaz> if you ask me... criminal law in most countries has gone haywire and needs to be scrapped and started again from scratch
[12:44] <_ph00> I would go even further
[12:44] <_ph00> I would say 'countries' as national states are obsolete
[12:44] <_ph00> we don't need those any more
[12:44] <_ph00> but that way too OT
[12:45] <_ph00> we can discuss that stuff on #freenet-chat, this should be a freenet support channel
[12:45] <MrNaz> _ph00: apathy is itself a product of media corporations... when you constantly feed a population stuff like "reality tv" and barrage them with psychological assaults in the form of music videos, the result is the typical modern western citizen.
[12:45] <_ph00> right
[12:46] <MrNaz> _ph00: i agree... i've always thought the arbitrary assignment of national borders as counter-intuitive...
[12:46] <_ph00> I was talking about that yeasterday (maybe that was in pvt with another user)
[12:46] <_ph00> launching a "drop your TV set down the window" campaingn
[12:47] <_ph00> I noticed that ppl who watch no or little TV are more receptive to common sense arguments, while TV watchiing ppl will repeat the same moronic arguments all the time
[12:47] <MrNaz> take africa for instance... tribal warfare was a very low grade thing prior to colonialization, the hellhole of open warfare is a direct result of the arbitrary definition of national identity by the colonial powers and their massive social programme to institutionalise nationalism at the grassroots level
[12:47] <sbc> _ph00: THAT'S NOT TRUE!
[12:47] <sbc> :P
[12:48] <_ph00> hey. take it easy
[12:48] <_ion> "Rejecting request from xxx:yyy preemptively because >SUB_MAX_PING_TIME (906.3224078256449)" lines are added to the log file all the time (many times a second). Could that have something to do with the transfer rate being so slow?
[12:48] <_ph00> what's not true and why?
[12:48] <_ph00> OK hold it
[12:48] <MrNaz> _ph00: well i've banned myself from TV
[12:48] <sbc> _ph00: The tv thing.
[12:48] <_ion> Sorry, i meant to say many times per 10 seconds.
[12:48] <_ph00> we have a real freenet realted question, we switch the OT to #freenet-chat, OK?
[12:48] <MrNaz> i only ever watch it when there's a speccific show on, and even that happens maybe twice a year... if i want to watch something to relax i'll get a dvd
[12:48] <_ion> I don't mind the discussion. :-)
[12:49] <MrNaz> mindlessly channel surfing for hours a day is something i just dont get
[12:49] <_ph00> we shoud switch to #freenet-chat anyways: this channel is logged and much OT makes it more difficult to conduct item related searches
[12:49] <_ph00> come on, I'm already there
[12:50] <_ph00> /join #freenet-chat
[12:50] * mario69 (n=nn@) has joined #freenet
[12:51] <ljn1981> MrNaz: Same here, I haven't even had the cable plugged in the tv for the better part of a year
[12:53] * mario69 (n=nn@) Quit ()
[12:53] * Zothar_Work (n=chatzill@) has joined #freenet
[12:55] <_ion> The peaks show how much bandwidth i have available, and the remaining continuous traffic is pretty much Freenet traffic. I'm certain i am connected to clients who have enough bandwidth. http://soijabanaani.net/rrd/img/adsl.png
[12:55] <ljn1981> <TheSeeker> Hmm... could we get a stat somewhere that keeps track of how many times we've swapped vs attempted to swap (local and remote) and/or a list of the locations we've had this session?
[12:55] <ljn1981> <TheSeeker> I know this could be figured out from the logs... but damn that would be a lot of work :P
[12:55] <ljn1981> <ljn1981> Would it be useful to have on the darknet page? It can't be that much work to do but do we want to have it?
[12:55] <_ion> My client is still inserting the 40 MiB file i began inserting yesterday.
[12:55] <ljn1981> <ljn1981> With all the stats we're starting to have, wouldn't we want to move them to a dedicated stats page instead?
[12:55] <ljn1981> toad_: Any input?
[12:56] <ljn1981> I was thinking I'd try to do those things if it was wanted, would be relatively easy and help familiarize me more with the sourse I think.
[12:56] * droden (n=droden@) has joined #freenet
[12:56] * ChanServ sets mode +o droden
[13:09] <TheSeeker> I'm intersted in the frequency of swaps and the locations to see if my crappy simulations are at all accurate.
[13:11] * JustMe (i=JustMe_@) Quit (Read error: 110 (Connection timed out))
[13:17] * kvsari (n=kvsari@) has joined #freenet
[13:20] <TheSeeker> eee, ouch
[13:20] <TheSeeker> # EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x00000000083348f1, pid=3328, tid=11520
[13:21] * Bombe (n=bombe@) has joined #freenet
[13:21] * ChanServ sets mode +o Bombe
[13:23] <TheSeeker> Is the hs_err_pid####.log file at all useful for trying to track down the cause of the hard jvm crash?
[13:24] * Urs_ShPo (n=gaim@) Quit (Remote closed the connection)
[13:26] * FooDerGrosse (n=kevin@) has joined #freenet
[13:38] * Werdna (n=Andrew@) Quit (Read error: 131 (Connection reset by peer))
[13:51] <toad_> rehi folk
[13:52] * SinnerG (i=SinnerG@) has joined #freenet
[13:52] <toad_> <_ion> "Rejecting request from xxx:yyy preemptively because >SUB_MAX_PING_TIME (906.3224078256449)" lines are added to the log file all the time (many times a second). Could that have something to do with the transfer rate being so slow?
[13:52] <toad_> perhaps
[13:53] <toad_> there may be problems with pre-emptive rejection...
[13:54] <toad_> _ion: what are the numbers? max/average/min?
[13:55] <toad_> ljn1981: if you want to keep more stats, then go right ahead
[13:55] <toad_> ljn1981: try not to make it too intrusive in terms of complicating code
[13:58] <TheSeeker> toad_: I'm thinking about network organization and going over papers again... it seems that 'ideal' topology assumes a cluster of 'near' nodes that are at an equally incrementing distance from the current one, along with a certain number of 'far' nodes per number of 'near' nodes that are connected using an inverse probability of distqance to determine if they are to be connected ...
[13:59] <TheSeeker> This would mean that it is impossible to have a 'far' node that is closer than any of the connected 'near' nodes, since all nodes closest to your own are already connected.
[13:59] <Zothar_Work> toad_: they could also be because routing backoff is a soft rather than hard backoff
[13:59] <_ion> toad: In the image? Yes.
[14:02] <TheSeeker> In a 1d network, the ideal distrobution of locations is a ring with equally spaced nodes... given this, it means that your 'near' nodes 'ideally' would be equally spaced and equally balanced to the right and the left.
[14:05] <TheSeeker> so, a swap should be taken if for both sides it produces: the nearest n/2 nodes to the right and left, where n is the number of near nodes, producing an average location spacing closer to 1 / (estimated network size)
[14:06] * TheSeeker works on how to do that in code
[14:07] <_ph00> I have read somewhere that to run freemail I need to use a freemail.jar file, but I dont have it and I can't fine it
[14:07] <_ph00> find*
[14:07] <TheSeeker> http://downloads.freenetproject.org/alpha/Freemail/
[14:07] <_ion> What might be a typical insert speed with a broadband connection, when everything works properly?
[14:07] <_ph00> thx seeker
[14:08] <_ion> assuming you're connected to people who also have broadband connections
[14:08] <TheSeeker> _ion: even if your neighbors are all broadband, you know nothing of *their* neighbors.
[14:09] <_ion> Hmm, right.
[14:09] <TheSeeker> Right now though, it seems that insert speeds of 10K/s are blazingly fast :P
[14:09] <_ion> Ok, thanks.
[14:09] <_ion> Is there going to be an effort to increase that?
[14:10] <TheSeeker> I'm sure there has been one for a while now.
[14:12] <nextgens> hi
[14:12] <ljn1981> lo
[14:16] <toad_> TheSeeker: something like that ...
[14:16] <toad_> Zothar_Work: what could? the rejections?
[14:17] <toad_> Zothar_Work: backoff is soft, but we have separate mechanisms for LIMITING load
[14:17] <Zothar_Work> toad_: yeah, the rejections would be greater after backoff was softened; what mechanisms limit load other than a full waiting for ACKs queue/window?
[14:17] <toad_> TheSeeker: having fun with your simulations? sound whacky...
[14:18] <TheSeeker> toad_: I haven't tried simulating the above yet, I'm still trying to formulate how I'll implement it in code :D
[14:18] <toad_> Zothar_Work: client-node side throttling
[14:18] <toad_> Zothar_Work: we run AIMD on the source node; when a request completes successfully it increases the window, when one fails with a timeout or a RejectedOverload, it reduces it
[14:19] <toad_> Zothar_Work: might be an idea to compare it with what we actually implement
[14:19] <toad_> err TheSeeker: might be an idea to compare it with what we actually implement
[14:19] <toad_> just to calibrate it
[14:19] <Zothar_Work> toad_: perhaps I need to look at logs of the operation of the AIMD or something
[14:19] <toad_> maybe
[14:20] <toad_> we will have a new load limiting algorithm when mrogers gets around to simulating it
[14:20] <toad_> but the basic problem seems to be this:
[14:20] <toad_> we reject pre-emptively to prevent timeouts
[14:20] <toad_> AIMD works on the basis of catastrophic failures - packet loss in the original sense
[14:20] <toad_> so AIMD throttles requests far too much
[14:20] <toad_> and everything is really slow
[14:21] <Zothar_Work> and yet we get lots of pre-emptive rejects
[14:21] <toad_> right
[14:21] <toad_> we get pre-emptive rejects for several reasons:
[14:21] <toad_> 1. high ping times
[14:21] <toad_> 2. high bwlimitDelay (too many transfers happening)
[14:21] <toad_> 3. failure to pre-allocate bandwidth for requests
[14:22] <toad_> any others? /me checks...
[14:22] <TheSeeker> toad_: I get a nice looking distance distrobution histrogram with the current swap algorythm, but I haven't found an initial connection algo that makes a network that stabalizes with the currently implemented swap algo ...
[14:22] <TheSeeker> meaning, I get frequent swapping and across long distances long after the distance histogram starts looking 1/d
[14:22] <Zothar_Work> perhaps we pre-emptively reject on the wrong numbers; maybe we can reject a peer's traffic based more on that peer's ping average instead of based on the node's ping average, but also take the node's memory usage, bandwidth usage and CPU usage and such into account, perhaps more than we do now
[14:23] <toad_> because we pre-allocate bandwidth, we ought to hardly ever get high bwlimitDelay ...
[14:23] <toad_> TheSeeker: i suggeast you are either implementing it wrong or you have a network which isn't small world
[14:23] <Zothar_Work> one or two really busy peers shouldn't be able to put our peer "over the edge" as much as most of our peers being busy
[14:23] <toad_> TheSeeker: it does work in simulations
[14:24] <toad_> Zothar_Work: yeah, there are some ideas ...
[14:24] <toad_> Zothar_Work: well, we use the median of the peer averages
[14:24] <toad_> so it shouldn't in theory be pushed over the edge by a few really slow peers
[14:24] <toad_> we do take bandwidth usage into account
[14:24] <toad_> we don't take CPU usage into account; we use ping time as a proxy for it
[14:25] <Zothar_Work> toad_: it would seem that we should target the lowest 20% or something rather than the middle since the lowest would give us a better idea of our own usage of our available bandwidth and other resources
[14:25] <toad_> Zothar_Work: hmmm
[14:25] <toad_> Zothar_Work: perhaps ...
[14:26] <toad_> what're your stats?
[14:26] <toad_> right now i have ping of 409ms
[14:26] <Zothar_Work> BTW, a common question when somebody mentions OOMs is whether they've got a lot queued; perhaps we should persue the request capping idea a little
[14:26] <toad_> and bwlimitDelayTime of 335ms
[14:26] <toad_> maybe, or maybe make queued stuff more memory efficient
[14:26] <toad_> or maybe up the default memory limit; the node can happily run in 100MB, but not if you have 3GB of stuff on the queue
[14:27] <Zothar_Work> Currently: * bwlimitDelayTime: 5ms * nodeAveragePingTime: 999ms * nodeUptime: 17h9m * backedoffPercent: 11.5% * pInstantReject: 40.3%
[14:27] <toad_> one thing we can do is change the datastore code to use a single database, that would probably save a lot of memory and a fair amount of CPU
[14:28] <TheSeeker> toad_: I'm working on implementing 'small world' in the manner described above, by linking nodes by index to the left and right then selecting nodes at random in proportion to the 'near' nodes and verifying that the randomly selected node doesn't already have too many 'far' connections.
[14:28] <Zothar_Work> I upped my memory to 512MB and still got OOMs with a handful of FCP generated requests at a time
[14:28] * mozillaman (n=chatzill@) has joined #freenet
[14:28] <Zothar_Work> (though I think I was getting OOMs when I wasn't generating those requests as well...)
[14:28] <toad_> TheSeeker: you choose the long range links in a 1/d configuration?
[14:29] <toad_> Zothar_Work: you shouldn't get OOMs with 512MB
[14:29] <toad_> Zothar_Work: how much stuff had you queued?
[14:29] <Zothar_Work> BTW, how hard would it be to get a single instance running multiple nodes? it seems that's been kept in mind in the design, but I'm not sure how far away we are from implementing
[14:29] <toad_> pretty easy
[14:29] <toad_> why would you want to?
[14:30] <TheSeeker> toad_: I will be, yes. where d is determind by the distance of the linked list rather than the random locations assigned to the nodes at initialization ...
[14:30] <nextgens> toad_> I've replied to your mail regarding StS
[14:30] <toad_> TheSeeker: yes, that's the idea
[14:30] <Zothar_Work> toad_: I was queuing stuff such that I never had more than a dozen or so requests at one time, very few of them were long-lived-download-type stuff since I'd cancel them after 15 minutes or so generally
[14:31] <TheSeeker> I think that's been part of my problem with init thus far. I assign a random node location, then try to connect nodes based on that location rather than 'physical' location on the ring.
[14:31] <toad_> Zothar_Work: then i have no idea why it would OOM and suggest you do some memory profiling
[14:32] <toad_> -Xloggc:/var/log/freenet-unstable/freenet-error.prof -Xrunhprof
[14:32] <toad_> :heap=all,format=a,depth=16,lineno=y,doe=y,gc_okay=y
[14:32] <toad_> or something like that
[14:32] <TheSeeker> Hmm, I'll try setting the locations on init based on curIndex/numNodes instead before completely rewriting the thing :D
[14:32] <Zothar_Work> toad_: I'd want multiple nodes in once instance since I'm currently running 3 nodes on one box and I suspect there could be some resource usage improvements, but then maybe I'm underestimating the usage of shared memory for libraries and overestimating the savings of threads instead of a few more processes
[14:32] <toad_> -Xloggc:gc.log -Xrunhprof:heap=all,format=a,depth=16,lineno=y,doe=y,gc_okay=y
[14:32] <toad_> Zothar_Work: well it's not a priority for me
[14:32] <Zothar_Work> :)
[14:32] <toad_> Zothar_Work: if you want to work on it that's fine :)
[14:43] <toad_> MrNaz: did you fix your connectivity problems?
[14:49] <nextgens> DOH
[14:49] <nextgens> I've got a switch and a server with the same ip address :<
[14:51] <nextgens> toad_> well, if you do have time to do StS, do it :)
[14:51] <nextgens> YetSomethingElseIwon'tFinishOnMyOwn though
[14:51] <nextgens> :|
[14:54] * kukuks (i=mixderax@) has left #freenet
[14:56] * _ion (i=johan@) has left #freenet
[14:56] * railk (n=railk@) has joined #freenet
[14:59] <MrNaz> toad_: my IP is dynamic... but i have a hostname thats static
[14:59] <MrNaz> home.mrnaz.com
[14:59] <toad_> how frequently does your ip change?
[15:02] <toad_> ljn1981: would you mind taking on "give the option to queue a fetch when a page DNFs on fproxy" ?
[15:04] <MrNaz> toad_: i dont know... i havent paid attention to it before
[15:04] <ljn1981> toad_: I have no idea how to do that but if it's easier to tell me how than to do it yourself then I could do it.
[15:04] <MrNaz> but i'd assume once ever few days
[15:04] <MrNaz> toad_: cant freenet work with hostnames instead?
[15:05] <ljn1981> I'm still not all that familiar with jave or the freenet source
[15:07] <toad_> MrNaz: it can work with hostnames yes
[15:07] <toad_> MrNaz: your IP shouldn't change every few minutes ...
[15:07] <toad_> ljn1981: well, have a look at how we do it in the large file handling code
[15:07] <toad_> ljn1981: you basically want to reproduce that on the ordinary-error-handling page
[15:08] <toad_> ljn1981: possibly with a user-selectable size limit
[15:08] <ljn1981> toad_: Ok
[15:08] <ljn1981> I'll give it a shot
[15:08] * core (n=Splitfil@) has joined #freenet
[15:09] * core (n=Splitfil@) has left #freenet
[15:10] <MrNaz> toad_: every few _days_ not minutes...
[15:10] <toad_> oh
[15:10] <MrNaz> hehe
[15:10] <toad_> that's different
[15:10] <toad_> you lose ALL of them?
[15:10] <toad_> is the STUN plugin loaded?
[15:11] <toad_> if your IP changes and you lose all your connections, it should try to determine your new IP via STUN
[15:11] <toad_> on the Plugins page, is there a plugin called JSTUN?
[15:11] <MrNaz> yeap
[15:12] <toad_> when it happens again, please could you check your node reference's ip address (physical dot udp equals ... - i can't actually say it cos i'll get kicked), against your real ip address (from whatismyip.com)
[15:12] <MrNaz> already did that
[15:12] <MrNaz> its picked up the new IP
[15:12] <toad_> and?
[15:12] <toad_> it picked up the new IP yet they haven't reconnected ?!
[15:12] <toad_> but it's definitely the result of an IP change? the IP was different when they were connected?
[15:13] <MrNaz> no
[15:13] <toad_> no, it was the same, or no, you don't know?
[15:13] <nextgens> toad_> iirc, we need one more message, hence I've not done it that way
[15:13] <MrNaz> my IP would have changed when my ISP capped me last night (damn idiot downloader hehe) the disconnections happened today all with the same IP
[15:14] <toad_> MrNaz: they didn't block your UDP?
[15:14] <MrNaz> i doubt it... lemme see...
[15:14] <toad_> MrNaz: one explanation is that all your peers are behind port restricted NATs or worse
[15:14] <MrNaz> i have all port blocks turned off...
[15:14] <toad_> MrNaz: in which case if your IP changes, and you're also behind one, it won't reconnect
[15:15] <toad_> actually restricted NATs
[15:15] <MrNaz> well you're not are you? you're my only peer at the moment heh i deleted all the disconnected ones, they were all from -refs
[15:15] <toad_> MrNaz: get some folk who are directly connected
[15:15] <toad_> well, when did you add me?
[15:15] <MrNaz> yesterday, before the IP change
[15:16] <toad_> you still have me in your peers list?
[15:16] <toad_> nextgens: STS is 3 stages. currently we use 4. what's the problem?
[15:16] <toad_> MrNaz: it's not set to BURST ONLY or LISTEN ONLY ?
[15:17] <MrNaz> yea you're still in my peers list... toad/dark
[15:17] <MrNaz> whats not set to burst only?
[15:17] <toad_> as 82.32.17.1?
[15:17] <toad_> the peer ...
[15:17] <nextgens> toad_> that we don't have the "nounce" ensuring the PFS in the first/two stages iirc
[15:17] <toad_> you can set peers to burst only or listen only to save bandwidth ...
[15:17] <nextgens> toad_> meaning that we have to break backward compat. anyway
[15:17] <nextgens> hence I was willing to reimplement the whole thing
[15:18] <nextgens> but if you don't bother and add a new message, it's fine
[15:18] <toad_> nextgens: what nonce?
[15:18] <toad_> nextgens: i'm looking at wikipedia/STS, what nonce?
[15:18] <nextgens> let me have a look
[15:18] <nextgens> the random value ensuring PFS
[15:18] <MrNaz> toad_: i didnt change it... but i dont even see that option? is it in the box labelled "My Peers" ?
[15:19] <toad_> alice->bob: g^x, bob->alice: g^y, e(s(g^y, g^x)), alice->bob: e(s(g^x,g^y))
[15:19] <toad_> whereas basic DH is pretty much the same without the e(s(...))'s
[15:19] <toad_> MrNaz: the dropdown box at the bottom
[15:20] <MrNaz> i only have 2 options in that: send n2ntm, update changed and remove
[15:20] <MrNaz> err... 3
[15:20] <MrNaz> heh
[15:21] <toad_> ok
[15:21] <toad_> send me your peers file (it only has me in it now, right?)
[15:21] <MrNaz> yeap
[15:21] <toad_> you know my email address?
[15:21] <toad_> encrypt it with my gpg key if you like ...
[15:22] <agsarite> toad_: did you get my freemail?
[15:22] <toad_> agsarite: yup
[15:22] <toad_> i didn't get sleon's yet
[15:22] <agsarite> yay
[15:23] <nextgens> toad_> sorry I gtg now
[15:23] <toad_> MrNaz: just privmsg me it then
[15:23] <nextgens> bbiab
[15:27] <toad_> MrNaz: okay, so it's not set to burst only / listen only
[15:47] <SinnerG> grr, anyone have xp with the 'shop' system for webhost related products? /query me :P
[16:09] * rah (n=rah@) has joined #freenet
[16:18] * Torched (n=torched@) has joined #freenet
[16:23] * droden (n=droden@) Quit ("Leaving")
[16:26] <CIA-5> mrogers * r10517 /trunk/apps/load-balancing-sims/phase6/ (10 files in 2 dirs): Routing fixes
[16:32] * XTorchedX (n=torched@) Quit (Connection timed out)
[16:50] * chris69 (n=chris@) has joined #freenet
[16:52] * chris69 (n=chris@) has left #freenet
[16:53] * Urs_ShPo (n=gaim@) has joined #freenet
[16:58] <CIA-5> ljn1981 * r10518 /trunk/freenet/src/freenet/clients/http/ (FProxyToadlet.java StatisticsToadlet.java):
[16:58] <CIA-5> Make a statistics page.
[16:58] <CIA-5> For now it show the same info as is shown on the top of the darknet page.
[16:58] <CIA-5> We should have a talk about which pieces should still be on the darknet page,
[16:58] <CIA-5> if any and the rest should be removed.
[16:58] <CIA-5> The layout of the status page should also be improved to take advantage of having a whole page for stats now.
[17:01] <ljn1981> TheSeeker: I'll look into adding a counter for successful/failed swaps in a bit, got a postit on my screen about it so I won't forget.
[17:01] <ljn1981> Now to do that queue on DNF thing toad_ asked for :)
[17:05] <TheSeeker> thanks ljn1981 :D
[17:05] <toad_> rehi
[17:05] * EdwardH (i=edwardh@) Quit (Remote closed the connection)
[17:06] <ljn1981> wb toad_
[17:06] <agsarite> it's because he didn't have a monocle, isn't it?
[17:07] <ljn1981> oops just remembered that commit message was a bit wrong, I also added a net size estimate with a 24 hour tresshold
[17:10] <agsarite> you will look at KSK@toad and you will like it, damn it.
[17:11] <toad_> :)
[17:15] <TheSeeker> So summer is over... what exactly has become of the various SoC projects?
[17:15] <toad_> most of them were partially successful
[17:16] <toad_> freemail is working, nextgens did a lot of good stuff, dunno about thaw, more work to be done on mrogers' simulations
[17:17] <toad_> i don't understand my node
[17:17] <ljn1981> Thaw is useable but I don't know if there have been any improvements worth upgrading for in the past month. It got a few issues but overall it's good
[17:18] <toad_> # bwlimitDelayTime: 321ms
[17:18] <toad_> # nodeAveragePingTime: 290ms
[17:18] <toad_> # pInstantReject: 0.0%
[17:18] <toad_> Total Output: 66.7 MiB (8.08 KiBps)
[17:18] <toad_> Payload Output: 45.0 MiB (5.45 KiBps) (67%)
[17:18] <toad_> Total Input: 51.3 MiB (6.22 KiBps)
[17:18] <toad_> ljn1981: does the index sharing functionality work?
[17:18] <toad_> my node is not rejecting any requests
[17:18] <toad_> yet it's not approaching its capacity
[17:18] <ljn1981> toad_: I think so but I never tried it
[17:18] <toad_> why?
[17:19] <toad_> not sufficiently easy to use yet then :)
[17:19] <toad_> it ought to be able to take on freemule
[17:19] * soniku is now known as ChaosControl
[17:20] <ljn1981> KSK@toad is scary......
[17:21] <toad_> anyone got any stats for me?
[17:21] <TheSeeker> I've got a bwlimitDelayTime of 0ms and a pInstantReject of 0%, but I'm still 55% backed off from my peers, and using almost no bandwidth :/
[17:22] <toad_> my node seems to be rejecting no requests yet using very little bandwidth
[17:22] <toad_> TheSeeker: are your peers backed off from you though?
[17:22] <toad_> i'm not backed off from you ...
[17:23] <ljn1981> http://dark-code.bulix.org/bsc2sy-18469?raw
[17:23] <toad_> there's something wrong here, but i dunno what it is...
[17:23] <TheSeeker> toad_: I'm not from you either.
[17:23] <ljn1981> bandwidth limit36/9KiBps
[17:23] <toad_> hmmm
[17:23] <toad_> 9kB/sec output limit?
[17:23] <toad_> but you only have 4 connections
[17:24] <toad_> i have 9+3
[17:24] <TheSeeker> toad_: hmm... nodes I am backed off from, are at 99.5% or higher backoff
[17:24] <toad_> and 40 disconnected
[17:24] * ChaosControl is now known as soniku
[17:24] <toad_> TheSeeker: hmmm, same here ...
[17:24] <TheSeeker> <flood>
[17:24] <TheSeeker> * Connected: 10
[17:24] <TheSeeker> * Backed off: 12
[17:24] <TheSeeker> * Disconnected: 1
[17:24] <TheSeeker> * Listening: 7
[17:24] <TheSeeker> * Listen Only: 6
[17:24] <toad_> all 3 backed off nodes are 100.0% backed off
[17:24] <TheSeeker> </flood>
[17:25] <toad_> could mean they have serious local problems...
[17:25] <toad_> TheSeeker: you have 12 backed off nodes, each at 99.5% or more backoff? that is interesting...
[17:26] <toad_> so ...
[17:26] <toad_> half the network is permanently backed off, and we don't route to it
[17:26] <toad_> it's not clear why
[17:27] <toad_> the other half of the network has slow as ever inserts, not spectacularly fast requests
[17:27] <toad_> and uses way too little bandwidth
[17:27] <toad_> despite rejecting no requests
[17:27] <toad_> i.e. it's being throttled too heavily at source, but there's no clear indication why
[17:28] <toad_> the 3 folk i have at 100% backoff are sleon MrNaz and Zothar_Work
[17:28] <toad_> sleon: MrNaz: Zothar_Work: give me your statistics, your nodes are 100% backed off and i'd like to know why...
[17:30] <Zothar_Work> toad_: * bwlimitDelayTime: 0ms * nodeAveragePingTime: 8258ms * networkSizeEstimateSession: 291 nodes * nodeUptime: 2h46m * missRoutingDistance: 0.0468 * backedoffPercent: 76.3% * pInstantReject: 94.9%# Requests: 1# ARK Fetch Requests: 9 * Backed off: 15 * Disabled: 1 * Listening: 8 * AcceptedTimeout 1 * ForwardRejectedOverload 1
[17:30] * Zothar_Work checks bandwidth usage
[17:30] <toad_> Zothar_Work: high local cpu usage? heavy bandwidth usage from other apps? any other obvious reason why the ping time would be through the roof?
[17:31] <Zothar_Work> CPU is relatively idle (I'm profiling, but CPU idle says it shouldn't be a profiling caused problem)
[17:32] * Gasi_ (n=chatzill@) Quit ("Chatzilla 0.9.67+ [SeaMonkey 1.0.5/2006091003]")
[17:32] <CIA-5> toad * r10519 /trunk/freenet/src/freenet/node/Node.java:
[17:32] <CIA-5> Minor changes to shouldRejectRequest:
[17:32] <CIA-5> - The once-every-ten-minutes requests we still need to clear for bandwidth usage.
[17:32] <Zothar_Work> I've noticed this sort of thing in the past, but usually the node's been up longer before this happens
[17:32] * timmy2chk (n=Vincent@) Quit (Read error: 60 (Operation timed out))
[17:32] <toad_> what i've found when i've investigated is that occasionally we get a network glitch which produces a really long ping time
[17:33] <toad_> and then it takes a while to get it out of the system
[17:33] <toad_> maybe we're averaging too much
[17:33] <Zothar_Work> bandwidth usage is normal and well below limit and capacity
[17:34] <Zothar_Work> I have noticed on my MRTGs that bandwidth usage tends to be better after a node restart, then generally taper off
[17:35] * Zothar_Work just got the nodeAveragePingTime too high userAlert, so it's been this way for awhile, but wasn't as of the data I copy/pasted
[17:37] * Zothar_Work updates from #977 r10501 to r10519
[17:37] <toad_> hmm
[17:38] <Zothar_Work> toad_: I'll hold off on updating if you need me to leave the node running in this state for now...
[17:38] <toad_> well i think the basic problem is that our load limiting algorithm isn't compatible with our pre-emptive rejection
[17:38] <toad_> Zothar_Work: upgrade
[17:39] <Zothar_Work> we need to poke mrogers to do the low-level write up :)
[17:40] <toad_> it's not the low level that's the problem
[17:40] <toad_> it's the mid-level - the load limiting algorithm
[17:40] * railk (n=railk@) Quit (Remote closed the connection)
[17:40] <toad_> maybe i should just implement it, simulations or no simulations
[17:40] <Zothar_Work> true, but at least it'd be another step towards where we need to be :)
[17:40] <agsarite> now. . . what sort of dark voodoo magic do i need to know in order to sign a frost message without using frost?
[17:41] * agsarite things he's going to need to read its source
[17:41] <Zothar_Work> agsarite: dunno, but Frost's source should be authoritative, yes
[17:43] <toad_> Subject: [freenet-dev] Token passing load limiting
[17:43] <toad_> any comments?
[17:43] <ljn1981> toad_: I had a similar situation on my node when I did those memory profiles for you
[17:43] <ljn1981> Similar to Zothar_Work's that is
[17:43] * soniku (n=areeves@) has left #freenet
[17:43] <ljn1981> toad_: I think I'm going blind.... I can't find the right place in the code for the DNF thing you wanted me to do or where the code is I'm supposed to make it work like
[17:43] <ljn1981> toad_: I thought the time of voodoo coding was over
[17:44] <toad_> ljn1981: yeah, but the current system plainly isn't working ... and mrogers seems to be taking forever ...
[17:44] * felev41 (n=m293@) has joined #FREENET
[17:44] <toad_> and the proposal just fits together, it *makes sense*
[17:44] <felev41> http://code.bulix.org/gvp3h2-18472?raw
[17:45] <toad_> it works on the principle that 1) you don't start requests any faster than they can complete, and 2) you route requests to (more or less) their best nodes
[17:45] <ljn1981> felev41: #freenet-refs
[17:45] <toad_> and ian's probably going to force me to do opennet without proper simulations anyway
[17:45] <agsarite> fuck.
[17:45] <toad_> and it'll probably be quicker to implement token passing than to chat about it frankly
[17:46] <agsarite> stupid software
[17:46] <felev41> ok thanks
[17:46] * agsarite (n=agsarite@) has left #freenet
[17:46] <Zothar_Work> toad_: my gut says do it, not that my gut is an oracle or anything
[17:46] * felev41 (n=m293@) has left #FREENET
[17:46] <Griffon26> toad_: my node is currently using < 20% cpu
[17:46] <toad_> Griffon26: yay
[17:46] <toad_> Griffon26: rejecting no requests?
[17:47] <Griffon26> toad_: I saw that before as well, it's just a phase
[17:47] <Griffon26> toad_: 0.0% instant reject
[17:47] <Zothar_Work> toad_: does the current load limiting mechanism interact with routing any? If so, perhaps the same mechanism can be used, reducing coding?
[17:47] <toad_> Griffon26: how many requests is it rejecting now?
[17:47] * Apophis2 (n=asdf@) has joined #freenet
[17:47] <toad_> Zothar_Work: yes, it interacts with routing; that's one of the complications
[17:47] * agsarite (n=agsarite@) has joined #freenet
[17:47] <Griffon26> toad_: I thought that was the pInstantReject value. What do you need?
[17:48] <toad_> Griffon26: so right now it's rejecting no requests? and using very little bandwidth?
[17:48] <Zothar_Work> native FEC fails unless I have semi-latest freenet-ext.jar I suppose
[17:48] <Griffon26> it's using between 0 and 4K/s, just like it was before when it used 100% cpu
[17:48] <toad_> okay
[17:48] <toad_> and your limit is?
[17:49] <Griffon26> output 50K, input -1
[17:49] <toad_> the input=-1 may be a problem
[17:49] * agsarite (n=agsarite@) Quit (Remote closed the connection)
[17:49] <toad_> we don't interpret it specially
[17:49] <toad_> just set it to 1000000
[17:50] <Griffon26> so the description of the field is incorrect?
[17:50] <Griffon26> "Input bandwidth limit (bytes/sec); the node will try not to exceed this; -1 = 4x set outputBandwidthLimit"
[17:50] <toad_> i dunno
[17:50] <toad_> oh ok
[17:50] <toad_> maybe we do
[17:50] * toad_ not perfect...
[17:50] <Griffon26> =)
[17:50] <Griffon26> I bet if I start frost, it starts to go wrong soon enough
[17:50] * agsarite (n=agsarite@) has joined #freenet
[17:51] <ljn1981> toad_: Too bad eh? Otherwise we'd have had 1.0 a few years ago :P
[17:51] <toad_> :)
[17:52] <ljn1981> Could have skipped all the funny stuff with NGR on 0.5 :P
