Timestamps are in GMT/BST.
[1:31] * moskau23 (~Miranda@dsl-082-082-232-104.arcor-ip.net) has joined #freenet
[2:55] <Oggy_> hmm
[2:55] <Oggy_> surfing freenet is interesting
[2:55] * Oggy_ is now known as Ogredude
[2:55] <Ogredude> you click the link, and then you go off and do other stuff for 15 minutes
[3:16] <i2p_iip> <gott> cunt
[3:19] * goatee_ (~goatee@ip216-239-88-19.vif.net) has joined #freenet
[3:32] * goatee (~goatee@ip216-239-84-104.vif.net) Quit (Read error: 110 (Connection timed out))
[3:41] <Ogredude> gott: What the hell's your problem?
[3:44] <i2p_iip> <gott> what ?
[3:44] <Ogredude> or were you not actually referring to me when you said "cunt"
[4:19] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) has joined #freenet
[4:37] <i2p_iip> <gott> stop trolling, Ogredude
[4:46] <Ogredude> Not trolling, just saying that surfing freenet reminds me a lot of the old 300 baud BBS days. I think it's cool.
[4:47] * kers (~kers@152.ppp144.rsd.worldonline.se) has joined #freenet
[5:21] <KenMan> gott - i realized the missing numbers after i posted last night. About half of my outgoing requests are QR'ed, the other half are DNF'ed, not timeouts!! At least, I ams pretty certain that is how it is working.
[5:22] <i2p_iip> <gott> you cunt
[5:22] <i2p_iip> <gott> now toad is crying
[5:24] <KenMan> in case anyone likes pictures to puzzle over, here is a graph of 5096 -> http://mywebpages.comcast.net/jkcorson/FreeNet/5096.png
[5:24] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[5:25] <KenMan> for those running MS browsers, it may contain a DoS - it is ~4000 pixels wide...
[5:25] <cbreak> picture!
[5:25] <KenMan> but for any modern, competent browser, it is safe to view.
[5:27] <cbreak> shouldn't connected peers and routes be the same?
[5:27] <KenMan> if you think so, take it up with toad. If he refuses, create a petition. I'll sign it, for one.
[5:28] <i2p_iip> <gott> Hey, Danny Bo
[5:28] <i2p_iip> <gott> *Boy
[5:28] <KenMan> wait, isn't it generally equal in that picture ? or do you see some red things along the light blue line ?
[5:28] <cbreak> only small pixels at the start.
[5:29] <KenMan> sometimes they are *different* and sometimes they are not so different.
[5:30] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[5:30] <KenMan> for this session, they stayed pretty darned equal :)
[5:31] <KenMan> see 5091c.png for a comparison - they stayed pretty close for that run.
[5:32] <KenMan> 5089c.png is pretty funny, since we are now past THAT experience...
[5:32] <cbreak> strange...
[5:35] <cbreak> there still seems to be a back off problem...
[5:42] <hobx_> You really think you will learn anything from that?
[5:52] <cbreak> why does backing off exist? Load Reduction on Peer Nodes?
[6:06] <hobx_> That, and not wasting time trying to send queries to nodes that can't handle them
[6:07] <cbreak> Most of my peers are backed off (on Unstable)... Wonder why...
[6:18] <hobx_> Because they haven't come up with a way to deal with the actual problem, only to try to defer the effects.
[6:19] <i2p_iip> <gott> KenMan, you gotta chill.
[6:30] * tyesinclair (~tyesincla@170.252.64.1) has joined #freenet
[6:30] * tyesinclair (~tyesincla@170.252.64.1) has left #freenet
[6:35] * d-ArkAngel (~Rob@213-131-104-86.onyx.net) has joined #freenet
[6:36] <d-ArkAngel> Morning all.
[6:36] <i2p_iip> <gott> GOOD MORNING
[6:36] <cbreak> yes. morning...
[6:54] * kers (~kers@152.ppp144.rsd.worldonline.se) Quit ("Leaving")
[7:09] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[7:16] * Hory (~Miranda@81.196.25.110) Quit (Read error: 104 (Connection reset by peer))
[7:54] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[7:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[7:55] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[7:56] * Ash-Fox (~N@aap86.neoplus.adsl.tpnet.pl) Quit (Read error: 60 (Operation timed out))
[8:10] * guido^pe (~unknown@dsl-213-023-238-121.arcor-ip.net) has joined #freenet
[8:11] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) Quit (Read error: 113 (No route to host))
[8:11] * Ash-Fox (~N@aap86.neoplus.adsl.tpnet.pl) has joined #FreeNET
[8:23] <mikeDOTd> should ERROR messages in the log file be sent to the devel mailing list?
[8:23] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[8:24] <mikeDOTd> my node just returned RNF, with 0 attempts made
[8:24] <mikeDOTd> which as i understand it, shouldn't happen anymore
[8:24] <cbreak> I doubt it, or the list will be useless due to overload...
[8:25] <cbreak> if you have no live connection, your node can not find a route.
[8:25] <mikeDOTd> i certainly have a live connection :-)
[8:26] <cbreak> not only backed off nodes?
[8:26] <mikeDOTd> 0 backed off nodes
[8:26] <mikeDOTd> 0 everything
[8:26] <mikeDOTd> which leads me to believe it's a bug
[8:27] <mikeDOTd> since rate-limiting should prevent RNF's for the most part
[8:27] <cbreak> only if you have possible routes.
[8:29] <cbreak> on http://localhost:8888/servlet/nodestatus/nodestatus.html you should see the number of Backed off nodes and the number of contacted node references.
[8:29] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[8:30] <mikeDOTd> contacted node references = 88, backed off nodes == 65
[8:31] <mikeDOTd> so you may be right, backed off nodes must have hit 88
[8:31] <mikeDOTd> my node has never been in contact with more than ~115 nodes, even though it's been on freenet for years and max connections is set to 200
[8:31] <mikeDOTd> oh well
[8:31] * mikeDOTd --> work
[8:39] * Ash-Fox (~N@aap86.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[8:39] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[8:39] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) has joined #FreeNET
[8:40] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[8:53] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[9:03] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) has joined #freenet
[9:19] * greycat (~wooledg@192.35.79.70) has joined #freenet
[9:25] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[9:35] <d-ArkAngel> The estimators, how "tight" are they? as I understand them they're variable width "buckets" with success probability assigned to them, would that be a valid description?
[9:35] <i2p_iip> <gott> *HUS toad*
[9:35] <i2p_iip> <gott> *HUGS*
[9:48] * tyesinclair (~tyesincla@170.252.64.1) has joined #freenet
[9:49] * FallingBuzzard (~srademach@207.152.112.129) has joined #freenet
[10:07] * tyesinclair (~tyesincla@170.252.64.1) has left #freenet
[10:12] <toad_> d-ArkAngel: yes, for the probability ones
[10:12] <toad_> [ot] eeek re nuclear batteries...
[10:12] <d-ArkAngel> do they have a fixed number of buckets?
[10:15] <toad_> yes
[10:16] <d-ArkAngel> Just a thought, but for the simulator, could we reduce the size of the keyspace? Since the simulation doesn't need such a large keyspace the performance improvements might be worth it.
[10:17] <toad_> we did
[10:17] <cbreak> estimator precision was already reduced to 53 bits
[10:17] <toad_> @_
[10:17] <toad_> :)
[10:17] <d-ArkAngel> :-)
[10:17] <d-ArkAngel> so what are the BigIntegers used for in the simulator?
[10:18] <d-ArkAngel> (I was assuming keys but I guess not)
[10:19] * guido^pe (~unknown@dsl-213-023-238-121.arcor-ip.net) Quit (Remote closed the connection)
[10:19] <toad_> not much now
[10:19] <d-ArkAngel> Sorry, I'll stop hastlening you and read the code :-)
[10:19] <toad_> unless fastestimators are turned off
[10:19] <toad_> well yes, i think they still are used for keys
[10:20] <toad_> that could be optimized out, would require interface changes/additions on fastSBKE
[10:20] <toad_> that would help to speed things up
[10:21] <toad_> as would using a PRNG instead of Core.randSource
[10:21] <toad_> as would optimizing Node.connect()
[10:21] <toad_> I'm going to do the latter soon
[10:21] <toad_> bbl
[10:21] <d-ArkAngel> ok.
[10:21] <toad_> keep in contact if you're going to change anything
[10:21] <toad_> i'm toad@amphibian.dyndns.org
[10:21] <d-ArkAngel> ok, will do.
[10:22] <d-ArkAngel> I think I'm just playing with my local copy at the moment to get my bearings, but if I do anything I feel is worthwhile I'll let you know.
[10:22] * toad_ hrrm http://www.ldodds.com/foaf/foaf-a-matic.html - like orkut except it's based on standards and not commercial sucky crap
[10:23] <toad_> of course it's very early
[10:23] <toad_> but it is the Right Way to do it
[10:23] <d-ArkAngel> indeed.
[10:23] <toad_> IF it's useful to do it...
[10:24] <toad_> which i'm still not entirely convinced of
[10:24] <d-ArkAngel> it's an XML Old Boy's network :-)
[10:24] <toad_> :)
[10:25] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[10:25] <toad_> hey, Mr RIAA! Here are all my friends, you can have this list for free, now you can go find them ab
[10:25] <toad_> and blackmail them and infiltrate and bribe and kidnap and generally use them to get to me..
[10:26] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[10:26] * d-ArkAngel thinks too much time on the freenet IRC group could lead to paranoid delusions....
[10:26] <toad_> ;)
[10:27] * toad_ suspects some of the intel orgs spend a great deal of money building this sort of data
[10:27] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[10:28] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[10:28] <d-ArkAngel> I wonder if it's them behind that XML tool :-)
[10:28] <toad_> lol
[10:28] <d-ArkAngel> getting us to do thier work for them :-)
[10:28] <d-ArkAngel> sneaky bastards :-)
[10:28] <toad_> more likely orkut ;)
[10:29] * d-ArkAngel wonders if he can write a game where people work for an auction house developing software and get people do do his job for him
[10:29] <toad_> heh, sure
[10:29] <d-ArkAngel> a man can dream can't he? :-)
[10:30] <d-ArkAngel> hmmm, maybe SimTerrorist has posibilities...
[10:30] <d-ArkAngel> :-)
[10:31] <cbreak> aka C&C Generals
[10:31] <d-ArkAngel> lol
[10:31] <d-ArkAngel> so that's how the US army is commanded.
[10:31] <d-ArkAngel> No wonder they keep attacking each other every now and again... :-)
[10:38] * toad_ thinks such a game could only be published on freenet...
[10:38] <toad_> developed, for that matter :)
[10:39] <d-ArkAngel> yeah I suspect the Cencors would probably object to the final level where you're required to pilot a plane hijacked plane... :-)
[10:39] <d-ArkAngel> anyway, enough banter. (before some of my really black humour starts to surface.
[10:40] <cbreak> I have read about a game, where you play some palestiean, and have to throw stones, fire rockets and the like to scare away israely settlers...
[10:40] <d-ArkAngel> lol
[10:40] <cbreak> Wasn't that popular...
[10:41] <d-ArkAngel> I wonder how many hits I could get if I hosted a flash game where you play the iraqi information minister defending iraq from the infidels :-)
[10:41] * lolo-laptop (~lostlogic@68.251.84.226) has joined #freenet
[10:41] <cbreak> with a skin pack for darl and his sco?
[10:41] <d-ArkAngel> lol
[10:42] <d-ArkAngel> yeah, swap missiles for Summary Judgements... that would work :-)
[10:42] <d-ArkAngel> which you have to deflect with requests for discovery :-)
[10:56] <d-ArkAngel> the random routing is a fixed % chance isn't it?
[10:56] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[10:57] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[10:59] <d-ArkAngel> was just thinking that since random routing is a good tool for a node to learn about it's peers, maybe that could be inital node learning could be improved by haveing an increased chance of random routing for areas of the key space that it the node thinks it has a low chance of finding...
[11:00] <toad_> d-ArkAngel: it's not actually random
[11:00] <toad_> the 5% probe requests, with probeinexperience on, go to the least experienced node
[11:00] <toad_> one idea is to route them to the least experienced node for that key
[11:01] <toad_> but this is all immaterial if i can't find some way to prevent nonfullyconnected causing such a vast slowdown
[11:01] <toad_> anyway bbiab, food
[11:01] <d-ArkAngel> it's easy to see why it slows down tho...
[11:02] <d-ArkAngel> it's the routing that's slowing it down, as the network stops being fully connected the number of HTL average increases.
[11:04] <d-ArkAngel> to some extent that's ineviateble. it's routing that needs to be sped up in the simulator to make the performance more reasonable.
[11:07] * goatee_ is now known as goatee
[11:10] <d-ArkAngel> in fact when running the simulator there's a direct corelation between the time the cycle takes and the number of Routed by estimators
[11:45] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[11:57] * Hory (~Miranda@81.196.25.110) Quit (Read error: 104 (Connection reset by peer))
[12:08] <d-ArkAngel> Toad, in the simulator, Node.java
[12:08] <d-ArkAngel> there are two copies of this "if(h != null && h.contains(p.n)) continue;"
[12:09] <d-ArkAngel> and h dosn't change in between
[12:09] <d-ArkAngel> so dropping the second one saves a has search per route (reasonable saveing)
[12:09] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[12:09] <cbreak> what does continue do?
[12:09] <cbreak> jump back at the top of a loop?
[12:10] <d-ArkAngel> moves to the next itteration of the loop
[12:12] <d-ArkAngel> dropping that extra statment saves something like half a second per 10,000 routes on my box.
[12:15] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[12:17] <hobx_> ah
[12:17] * seb- (~seberino@seberino3.spawar.navy.mil) Quit (Remote closed the connection)
[12:17] <hobx_> good work optimizing the right places
[12:19] <cbreak> I wonder what that is in %
[12:23] <toad_> lol
[12:23] <toad_> hobx_: I know what the big cost is in terms of optimization
[12:23] <toad_> it's the Node.connect() method
[12:24] <toad_> d-ArkAngel: that's not it, it wasn't THAT much lower HTL before
[12:24] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[12:24] <toad_> <d-ArkAngel> in fact when running the simulator there's a direct corelation between the time the cycle takes and the number of Routed by estimators
[12:24] <toad_> really? i didn't think it had got that high...
[12:24] <d-ArkAngel> yeah, it's because the HTL increases as the network becomes less and less connected.
[12:25] <hobx_> toad: Indeed
[12:25] <d-ArkAngel> so the number of times route is called increases
[12:25] <d-ArkAngel> connect is called a constant number of times in each itteration, so shouldn't be a problem. it's routing that's takeing more time.
[12:26] <d-ArkAngel> you can limit it by forcing a lower max HTL but then your success ratio starts to fall (as you have to limit to something like HTL 3-5 to have an impact.
[12:27] <d-ArkAngel> Interestingly when you think about it, how many nodes are there in the live network? 10,000? with max HTL of 25 and we're simulating 400 with a HTL of 15
[12:28] <toad_> [OT] hrrm, iraqbodycount.net reports total still under 15K... many demonstrators claim 30K. both numbers include people killed by the militants...
[12:28] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[12:28] <toad_> d-ArkAngel: no
[12:28] <toad_> connect is called a LOT of times
[12:28] <toad_> because it's called every time a request succeeds
[12:29] <d-ArkAngel> I must have missed that :-)
[12:29] <d-ArkAngel> I would have thought that it was the calculation of the estimates that was the big drain on the situation.
[12:29] <cbreak> so to speed things up, requests have to fail... hmm...
[12:29] <toad_> d-ArkAngel: it WAS
[12:29] <toad_> HOWEVER
[12:30] <toad_> we should have the same number of estimates on 400x25 as on 100x100
[12:30] <toad_> but 400x25 is a LOT slower
[12:30] <d-ArkAngel> because the average HTL increases
[12:30] <toad_> THAT MUCH?
[12:30] <toad_> it was about 3 last time
[12:30] <toad_> now it's about 4, right?
[12:31] <toad_> even if it's 20, there has to be some superlinearity going on with it to explain the increased slowness
[12:31] <d-ArkAngel> during the 25x25 phase it's down at 0.9, and it slowly climes, was at 3 by 150x25
[12:33] <d-ArkAngel> and isn't connect only called when the data is found? that's still only once per successful request isn't it?
[12:36] <cbreak> is a request successfull once on each node allong the path?
[12:36] <cbreak> (since it should not be recognizable if it's a local request or not)
[12:36] <toad_> d-ArkAngel: during the 100x100 sims, it was about 3
[12:37] <toad_> 3 htl with 150x25 is a LOT slower, here, than 3 with 100x100
[12:37] <toad_> so that's not it
[12:37] <greycat> does your simulation have nodes disappear from the network randomly?
[12:37] <toad_> no, it's quite simplistic at present
[12:39] <d-ArkAngel> oh I see, it's the performance inside connect you think is being non-linear, not it's influence... I was thinking about comparative increases. my mistake.
[12:39] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[12:40] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[12:40] <d-ArkAngel> but performance is performance :-) so improvements to routing help too, in thier own small way :-)
[12:42] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[12:42] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) has joined #FreeNET
[12:54] <d-ArkAngel> Out of interest, how many routes by estimator were there per 10000 requests in the 100x100?
[12:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[12:54] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has left #freenet
[12:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[13:03] * d-ArkAngel (~Rob@213-131-104-86.onyx.net) has left #freenet
[13:17] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) Quit (Client Quit)
[13:17] * dystopia (admin@gaia-wan0.ipv6.freeshell.bofx.net) Quit (niven.freenode.net irc.freenode.net)
[13:22] * setre (~jussi@h136n1fls301o291.telia.com) Quit (Remote closed the connection)
[13:22] * setre (~jussi@h136n1fls301o291.telia.com) has joined #freenet
[13:23] * dystopia (admin@gaia-wan0.ipv6.freeshell.bofx.net) has joined #freenet
[13:23] <sanity> hi toad
[13:23] <sanity> wazzup?
[13:26] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) has joined #FreeNET
[13:26] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) Quit (Read error: 54 (Connection reset by peer))
[13:26] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) has joined #FreeNET
[13:40] <KenMan> "toad doesn't live here anymore, ha ha ha" :)
[13:43] <KenMan> wow, memory fails me again: the original line was "Dave? Dave's not here!" ...
[13:46] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has joined #freenet
[13:51] * guido^pe (~unknown@dsl-082-082-159-162.arcor-ip.net) has joined #freenet
[13:51] * vsalento (~vsalento@cs136227.pp.htv.fi) Quit (Read error: 104 (Connection reset by peer))
[13:52] * vsalento (~vsalento@cs136227.pp.htv.fi) has joined #freenet
[13:52] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[14:02] * FallingBuzzar1 (~srademach@207.152.112.129) has joined #freenet
[14:02] * FallingBuzzard (~srademach@207.152.112.129) Quit (Read error: 104 (Connection reset by peer))
[14:11] * tyesinclair (~tyesincla@170.252.64.1) has joined #freenet
[14:12] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has left #freenet
[14:12] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) Quit ("brb")
[14:12] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has joined #freenet
[14:14] * tyesinclair (~tyesincla@170.252.64.1) has left #freenet
[14:15] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[14:48] <toad_> hehehehehehehehe
[14:49] <d-ArkAngel> what amuses?
[14:51] <toad_> above conversation
[14:56] <hobx_> KenMan: I thought you were quoting from "Frog & Toad"
[14:58] * FallingBuzzard (~srademach@207.152.112.129) has joined #freenet
[15:00] <KenMan> hobx: oh wow! Are you familiar with the literary works of 'Frog & Toad' ??
[15:02] <KenMan> some of the plot lines are absolutely gripping (if you are around 6 years old, anyway)
[15:11] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[15:13] * FallingBuzzar1 (~srademach@207.152.112.129) Quit (Read error: 110 (Connection timed out))
[15:15] <hobx_> "The trial of all crimes, except in cases of impeachment, shall be by jury; and such trial shall be held in the state where the said crimes shall have been committed; but when not committed within any state, the trial shall be at such place or places as the Congress may by law have directed. When sitting for that purpose, they shall be on oath or affirmation! In every place of great resort the monster was the fashion:"
[15:15] <hobx_> At least spam is now becoming educational.
[15:16] <greycat> is that a single piece of spam, or is that from the spam-poetry web page?
[15:16] <hobx_> though I don't think the last sentence belongs
[15:16] <mikeDOTd> i like the spams trying to sell weapons to terrorists
[15:16] <hobx_> it was from a single piece of spam, the anti filter stuff they add
[15:16] <hobx_> (Didn't fool mozilla)
[15:18] <d-ArkAngel> they probably just used the documentation they had been preparing to defend them selves when they get busted for it :-)
[15:30] * sanity_ (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[15:30] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 232 (Connection reset by peer))
[15:47] <jay> [20:17] <d-ArkAngel> does freenet need J2EE to compile?
[15:47] <jay> nope.. it needs freenet-ext.jar
[15:47] * kers (~kers@207.ppp142.rsd.worldonline.se) has joined #freenet
[15:47] <jay> i think you figured that one out already
[15:59] <d-ArkAngel> indeed :-) thanks for the help all the same tho :-)
[16:00] <d-ArkAngel> my problem turned out to be that ant was somehow linked into gjc, which didn't like much of anything
[16:15] <jay> gcj may or may not work
[16:15] <jay> even when done right
[16:19] <jay> best results will be from Sun's JDK or JRE
[16:20] * kers (~kers@207.ppp142.rsd.worldonline.se) Quit (Read error: 110 (Connection timed out))
[16:22] <Ash-Fox> toad_... please make freenet require all users to be RFC 1855 compliant (http://www.apps.ietf.org/rfc/rfc1855.html)
[16:25] <cbreak> in that rfc, the one-to-no-one communication was left out...
[16:31] <jay> hehe
[16:31] <jay> does Newsbyte have to agree to that rfc?
[16:31] <jay> if so, there will be war ;)
[16:32] <jay> im tired of the OT shit on the mailing lists
[16:32] <jay> esp. the last reference to God and christianity
[16:32] <jay> he should be booted off imo
[16:34] <jay> none of it has any place in the world of netiquette
[16:40] <jay> [20:28] * toad_ hugs gott
[16:40] <jay> heh
[16:55] <d-ArkAngel> can't beat a little show of affection. :-)
[16:59] <KenMan> the trouble people are having with ant:gcj is probably fedora-specific. Instead of 'ant jar' you should run 'classic-ant jar' on fedora.
[17:01] <KenMan> Just one more peculiarity on Fedora. I didn't even install gcj. Anyway, it is one more thing to decide - "do I fix it properly now, or wait for Core 2 to become legacized?"
[17:01] <d-ArkAngel> ahh, well that would have been useful to know yesturday. I've removed the ant rpm's alltogether, and installed a clean one
[17:01] <KenMan> good enough :) that is the proper solution!
[17:02] <d-ArkAngel> it's not the first time that a FC customisation has caused me small amounts of pain
[17:02] <KenMan> rpm's really piss me off, mainly because I haven't been using them for the past 7 years.
[17:02] <d-ArkAngel> I'm thinking of doing a LFS setup, if onyl I could get the time to do it :-)
[17:03] <KenMan> i mean, sure they are easy to install, but... YES. I strongly endorse LFS. It can take up to a week to get going with it, if you are time constrained.
[17:03] <KenMan> The recipe is good though. You get what you want, how you want, and nothing BUT what and how YOU want.
[17:04] <d-ArkAngel> I did it ages ago, back then you had to type out all the boot scripts. it sure taught me a lot about linux... well it nearly did, it got to nearly midnight so I just started blindly copy/pasteing them so I could get it to boot :-)
[17:04] <KenMan> heh. Then you learned.
[17:04] <d-ArkAngel> I think the main advantage is that you know what is installed.
[17:05] <d-ArkAngel> and when you type ant it runs ant, and not something that someone else though would be a good idea :-)
[17:05] <KenMan> With LFS, you won't "enhance" the basic packages with unusual default settings, etc.
[17:06] <KenMan> unless YOU know you want to, and then you get exactly what YOU want. Hmmm, seems like there is something to this concept.
[17:06] <d-ArkAngel> is that "enhance" in the same way that windows users suffer from "features"
[17:06] <cbreak> thats why there are nearly no linux viri... no common standards... :)
[17:07] <KenMan> Well, it is clear to me that Fedora is aiming for a less-technical crowd. And they deserve credit for it.
[17:07] <d-ArkAngel> yeah, a 20 Mb virus, just because it has to be able to cope with all the possible file locations :-)
[17:07] <d-ArkAngel> I've gotta say that for none technicals I prefer Mandrake.
[17:08] <KenMan> The Fedora GUI is really quite nice (I set it for GNOME). But the tools are limited to a 'typical, non-technical' subset. Nothing wrong with that, unless you are expecting more.
[17:08] <d-ArkAngel> I've never had problems with none standard things with them, and it's fairlt nice.
[17:08] <d-ArkAngel> Mandrake was the first distro to reach a level that I could live with as a windoze user
[17:09] <KenMan> who created iptables-save and -restore ? I get the feeling it is a standalone package, but it wouldn't surprise me to discover that Fedora created them.
[17:09] <d-ArkAngel> the old redhat biulds I always thought were trying to be bleeding edge too hard.
[17:10] <d-ArkAngel> but fedora sure is impressive.
[17:10] <KenMan> Well, just 2.6 is kind of bleeding edge, depending upon whom you converse with.
[17:10] <d-ArkAngel> I don't know. they're probably just variations on a theme.
[17:10] <d-ArkAngel> true.
[17:10] <d-ArkAngel> there's a BELFS :-)
[17:11] <d-ArkAngel> that's got all the newest things in :-) that made me smile when i saw it :-)
[17:11] <KenMan> yeah. And you can blame someone else when just pulling all the latest things together doesn't work. But I did it for years with tremendous luck and success :)
[17:12] <d-ArkAngel> that had the 2.6 kernel with all the experimental FS enabled and things. (like devfs and the rest)
[17:12] <d-ArkAngel> :-)
[17:12] * d-ArkAngel dosn't mention that his other monitor is currently displaying the windows XP install screens...
[17:13] <KenMan> you don't say ?!
[17:13] <cbreak> strange screensaver you have...
[17:13] <KenMan> hey, is that toad_ around here ?
[17:14] <d-ArkAngel> I'm reformatting for a friend of a friend. I certainly was tempted to throw lnux on and see if they noticed, but they've probably got windows software and wouldn't understan why it didn't work any more :-)
[17:14] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit (".")
[17:16] <d-ArkAngel> to be fair I do dabble with windows on this box, I've got XP, XP 64-bit, and Longhorn preview install on other partitions.... (not that I ever boot them, but I just can't be bothered re-partioning :-)
[17:16] <cbreak> you could donate them to freenet...
[17:17] <d-ArkAngel> the reformatting problem is related to the fact that most of the partitions are covered with Stargate episodes, and I had to sell my DVD writer last week to help pay for some legal fees.
[17:18] <d-ArkAngel> how fast does the storage need to be for freenet?
[17:19] <d-ArkAngel> would it be overly upset at haveing storage over a network connection?
[17:19] <cbreak> faster than your internet connection I think.
[17:19] <d-ArkAngel> with my upload rate, that's not hard :-)
[17:20] <d-ArkAngel> I wonder if I could hack something together to allow a freenet node to use distributed HDD's accross multiple machines...
[17:20] <cbreak> a simple gigabit ethernet should be more than enough :)
[17:20] <d-ArkAngel> :-) I think it'd have to live with 100Mbit.
[17:21] <cbreak> you could mount each partition inside store/xy for different xy
[17:21] <cbreak> hmm...
[17:23] <cbreak> or you could make a network raid... :)
[17:24] <d-ArkAngel> thinking about it my node east enough CPU time as it is...
[17:24] * sanity_ (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[17:24] <d-ArkAngel> trying to compilcate matters wouldn't help matters :-)
[17:25] <cbreak> probably not...
[17:25] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[17:26] * greycat (~wooledg@192.35.79.70) Quit ("Lick Bush in '04")
[17:46] <toad_> hi ppl!
[17:47] <d-ArkAngel> Hullo
[17:51] * toad_ just received the other $$
[17:53] <d-ArkAngel> the processing speed of the simulator seems to bottom out at about 34 seconds per cycle on this box, is that in line with what you were seeing toad? (average hops to data of 2.99 ish)
[17:53] <toad_> d-ArkAngel: no, it's more like 10 times that...
[17:55] <d-ArkAngel> well it's not hit 35 seconds and it's just hovering arround 33.5-34.8 ish up and down each cycle...
[17:57] <d-ArkAngel> in fact that's with the full 400x25 network.
[17:57] <d-ArkAngel> which might explain why it's stopped getting slower.
[17:59] <d-ArkAngel> the succses rate is dire tho, less than 0.1 at 400, tho it seems to be slowly climbing up again, 0.1191 now now
[18:01] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) Quit ("Als ik spreek is het slecht, als ik zwijg is slecht ...")
[18:03] <d-ArkAngel> yeah, so it looks like the network can't cope with new nodes being added every 10,000 requests with these settings.
[18:05] <d-ArkAngel> what machine are you running your node on? your desktop, or the other dual oppy you've got access to?
[18:05] <d-ArkAngel> your simulation even
[18:08] <toad_> did you make changes?
[18:08] <toad_> i mean it slows down radically here...
[18:08] * FallingBuzzard (~srademach@207.152.112.129) has left #freenet
[18:09] <d-ArkAngel> tthe only change I made was the one I mentioned earlier, the duplicate check for n in h in the routing, which does seem to be a significant improvement later on in the construction.
[18:09] <d-ArkAngel> "if(h != null && h.contains(p.n)) continue;" is duplicated in Node.java
[18:09] <d-ArkAngel> that's the only change I've made.
[18:10] <d-ArkAngel> it is a slowdown here too, at the start it's 7-9 seconds on the 25x25 network, and it slows down to 34 ish for 400x25
[18:10] * tyesinclair (~tyesincla@host217-37-104-211.in-addr.btopenworld.com) has joined #freenet
[18:10] <toad_> okay, that's all the $admin dealt with for now...
[18:11] <toad_> d-ArkAngel: hmmm
[18:11] <toad_> well if it's constant at least we know it's not leaking or something...
[18:11] <toad_> run it for a few hours (days?), see what happens to psuccess...
[18:11] <toad_> I think what I should do now is to see if i can optimize the simulation a bit...
[18:12] <d-ArkAngel> well the last 40 or so inserts have all been 34ish seconds. but the psuccess is deffinatly clibming.
[18:12] * tyesinclair (~tyesincla@host217-37-104-211.in-addr.btopenworld.com) has left #freenet
[18:13] <toad_> lets see how far it goes...
[18:13] <d-ArkAngel> I did make some minor changes to the two stat gathering objects, I made global static, so it really is global, and then made local have the stats for the individual block of 10000 requests.
[18:14] <d-ArkAngel> not sure how much effect that would have on the performance, but I wouldn't have thought it was drastic...
[18:15] <toad_> okay, that's good
[18:15] <toad_> commit it please
[18:15] <toad_> you have cvs?
[18:15] <d-ArkAngel> no, anonymouse download only.
[18:16] <toad_> okay, mail me the diff, and your sf nym
[18:16] <d-ArkAngel> (or do you mean the program cvs? :-)
[18:16] <toad_> if you don't have an sf nym, get one
[18:16] <d-ArkAngel> I probably do.. :-)
[18:16] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) Quit (Read error: 60 (Operation timed out))
[18:17] <d-ArkAngel> I've never worked with CVS before, there'll be a decent guide up on sf won't there?
[18:18] <toad_> think so
[18:18] <toad_> start at the freenet CVS page on sf
[18:19] <toad_> that i think links to some docs
[18:19] <d-ArkAngel> I seem to recall someone linking to it when I asked how to get the source by CVS
[18:19] <cbreak> http://www.cvshome.org/docs/manual/
[18:20] <cbreak> and of course "man cvs" :)
[18:20] <d-ArkAngel> :-) no http://www.cvs-for-dummies.com ? :-)
[18:21] <KenMan> toad_: on my last run, the node sent out 118K requests. It got back 51K QR:RNF. Can these be caused by anything *but* "no routes available within queue-time" ?
[18:21] <KenMan> I am wondering how queueing affected % of RNFs. It feels like it affected the distribution of where they happen, but maybe not how much they happen.
[18:22] <toad_> hmmm
[18:22] <toad_> well that's better than it used to be
[18:22] <toad_> right?
[18:22] <toad_> in any case... hmmm
[18:22] <toad_> grep for the string
[18:22] <KenMan> it is ?? Maybe. But not by much if it is... I would have to look through IRC logs to check.
[18:22] <toad_> grep the source for the string...
[18:22] <toad_> KenMan: I'm pretty sure it's significantly better
[18:23] <toad_> KenMan: anyway
[18:23] <toad_> on my node
[18:23] <toad_> the diags on queueing say the overwhelming majority of requests get sent
[18:23] <KenMan> I am not counting DNF msgs, but I would expect there are slightly more of them than RNF. I thought that's about where we were a few weeks ago.
[18:24] <KenMan> Well, it is nice they get sent out. Perhaps on certain nodes, they don't get sent out as well. Perhaps there is a bug with the diag.
[18:24] <KenMan> I mean, what else can cause RNF ? If all messages everywhere get sent, how can RNF happen ? What other causes are there ?
[18:25] <KenMan> I am grepping QRs, so that I only count RNFs. Is that what you were starting to say re:grep ?
[18:26] <toad_> hmmm
[18:26] <KenMan> I log QR,reason for each QR I get in.
[18:26] <toad_> bi
[18:26] <toad_> no
[18:26] <toad_> grep the source for "No route found"
[18:26] <toad_> or whatever the exact message is...
[18:26] <KenMan> oh, i see what you mean :)
[18:26] <toad_> and then trace the usage
[18:26] <KenMan> bah, everything seems to take effort :p
[18:27] <d-ArkAngel> well I've found my SF account.
[18:27] <KenMan> RNF is only sent from one point inside Pending. Going to read it now...
[18:29] <toad_> anything worthwhile takes effort
[18:29] <toad_> that's just life
[18:29] <KenMan> receiveQueueSendFailed sends that message. Now where are the places we make those ?
[18:29] <d-ArkAngel> but hardwork pays off in the future, lazyness pays off now.... ;-)
[18:29] <KenMan> d-ArkAngel: that's the spirit :)
[18:29] <d-ArkAngel> if only I had the time to be lazy :-)
[18:30] <toad_> KenMan: we make QueueSendFailed if and only if we fail to send the message from the queue
[18:30] <toad_> it is however possible that a lot of nodes have overload, looping, CPU usage, or other bugs with queueing...
[18:30] <KenMan> okay. So it is still (netwide at least) a significant problem.
[18:30] <toad_> we have heard a few reports...
[18:30] <toad_> okay, fixed one simulator slower-downer...
[18:30] <KenMan> yes. Need to track some of those down. Perhaps if you lowered your ObBW you would see some of them yourself ?
[18:32] <KenMan> now what I'm not currently doing is logging the source of these RNFs. Otherwise I could tell if certain nodes RNFed much more heavily than others.
[18:32] <toad_> hmm?
[18:32] <KenMan> either RNF is highly concentrated on a few nodes, or else it is pretty well distributed.
[18:33] <KenMan> (evenly)
[18:33] <KenMan> We want to find a node with a high queue-failed-to-send diagnostic value, and see what's going on there.
[18:35] <toad_> how?
[18:36] <toad_> how do we go from an IP address to a node's other stats?
[18:36] <toad_> even its operator?
[18:36] <KenMan> we need 'reporters' in here
[18:36] <KenMan> just tell us which diag to report
[18:37] <KenMan> or, ask for it on the mail list... "if you see 5096 w/QF2S above 20%, please share diag 'XYZ' and 'ABC'"
[18:38] <KenMan> specify a minimal uptime, ask for a fixed response chart... whatever makes it meaningful to you.
[18:39] * toad_ makes simulation use double for everything instead of Key...
[18:41] <d-ArkAngel> toad, I've just mailed you the patch that I've made to the simulator.
[18:41] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[18:41] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[18:42] <d-ArkAngel> I also noticed that I had dropped the HTL to 5 from 10, I'll put it back up to 10 and see how that affects the cycle times.
[18:42] <d-ArkAngel> and psuccess.
[18:43] <toad_> argh
[18:43] <toad_> don't want to have to wrap double's for the DS LRUQueue... hrrm
[18:47] <d-ArkAngel> on which note my headache things I should go get some sleep.
[18:47] <d-ArkAngel> catch you all tomorrow.
[18:48] <KenMan> gnight dAA
[18:48] <d-ArkAngel> is there a way to detach a simulator from a console so I can log out and init 3 this box without killing the 400x25 that's running?
[18:50] <KenMan> does control-Z work in the console ?
[18:50] <d-ArkAngel> yeah, I've done that, and BG'd it
[18:51] <d-ArkAngel> but I'm worried that killing the console might kill the task...
[18:51] <KenMan> that might/should do it. Only one way to find out.
[18:51] <cbreak> try screen
[18:51] <cbreak> a quite nice program, offers multi virtual terminals which are detachable
[18:52] <cbreak> and re attachable
[18:52] <d-ArkAngel> sounds good, I'll have a look at the man page and see what it's got to say for it's self. it'd be a shame to have to restart the simulations again now.
[18:53] <cbreak> just run "screen", use ctrl-a c to create a new shell, ctrl-a n to circle and ctrl-d to detach.
[18:53] <cbreak> reatach with "screen -r"
[18:54] <d-ArkAngel> does that only apply to tasks that I start after running screen?
[18:54] <cbreak> yes.
[18:54] <d-ArkAngel> dang :-)
[18:54] <cbreak> only to those tasks you start inside screen.
[18:54] <cbreak> I doubt you can fg tasks you have bg'ed in an other shell.
[18:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[18:55] <d-ArkAngel> to be fair I don't need them back I just want to be sure that when I log out they'll keep running as BG'd
[18:56] <cbreak> here "sleep" processes continue to run after their shell was exited.
[18:56] <cbreak> don't know for init 3 (whatever that is)
[18:57] <cbreak> (well bg'ed sleeps of course...)
[18:57] <d-ArkAngel> just to kill of the GUI stuff to free up the resources
[18:59] <toad_> > /dev/null 2>&1 &
[18:59] <KenMan> wow, toad_ has written a lot of code to create this simulator!
[19:00] <toad_> KenMan: it's still a bit of a hack in a lot of ways...
[19:00] <d-ArkAngel> well I've got 1 and 2 sent to log files.
[19:00] <d-ArkAngel> so will that survive when I kill the shell?
[19:00] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[19:00] <toad_> hmmm
[19:00] <toad_> LRUQueue's will necessarily create lots of objects, unless they pool, right?
[19:00] <d-ArkAngel> I didn't think to nohup it when I set it off.
[19:00] <toad_> object creation is to be avoided here...
[19:01] <toad_> in the simulator...
[19:01] <toad_> i'm trying to make all keys in the simulator double's
[19:01] <toad_> and cut object creation to an absolute minimum
[19:01] <toad_> I need to either:
[19:01] <toad_> a) Create a Key object that wraps a double OR
[19:01] <toad_> b) Create an LRUQueue variant that uses double's.
[19:01] <toad_> c) Revert current changes
[19:02] <d-ArkAngel> well I don't think an object wrapping a double would be too much of an overhead...
[19:02] <toad_> if I do b), I should probably optimize it significantly using some form of pooling...
[19:02] <toad_> d-ArkAngel: you think? i'm not sure
[19:02] <d-ArkAngel> isn't there a class ready made for that called Double
[19:03] <toad_> sure
[19:03] <d-ArkAngel> java's pretty good with objects for most things...
[19:03] <toad_> well, the current situation:
[19:03] <toad_> d-ArkAngel: no
[19:03] <d-ArkAngel> after all a string is an object no matter how you create that...
[19:03] <toad_> it won't run fast if you allocate objects
[19:03] <toad_> it only looks that way because java uses objects for EVERYTHING
[19:03] <toad_> anyway, currently:
[19:03] <toad_> we have a Key object
[19:03] <toad_> this wraps a BigInteger
[19:04] <d-ArkAngel> but string isn't an attomic type, the only attomic types are int, double and something else....
[19:04] <toad_> which we then turn into a double for the actual calculations
[19:04] <toad_> d-ArkAngel: yes, and if you do string handling constantly, things slow down
[19:04] <toad_> try turning the loglevel up, you'll see what i mean
[19:04] <d-ArkAngel> :-) I'll avoid that thanks :-)
[19:05] <toad_> logLevel=minor, selective debug in logLevelDetail, is a great debugging tool
[19:05] <toad_> but it DOES slow things down
[19:05] <toad_> not to mention producing multigigabyte hourly log files...
[19:06] <toad_> so, a, b, or c?
[19:06] <d-ArkAngel> well which one is the quickest to code?
[19:06] <toad_> c) is quickest
[19:07] <toad_> a) is slightly slowe
[19:07] <toad_> r
[19:07] <toad_> b) is best performance, potentially...
[19:07] <toad_> but most work
[19:07] <toad_> hmmm
[19:07] <d-ArkAngel> well you could try a, because it's in the same general direction as b
[19:08] <d-ArkAngel> if it yeilds a decent amount of perfomance, but not enough then it hints that c is the right direction to be heading...
[19:09] * toad_ does b...
[19:10] <toad_> no i meant
[19:10] <toad_> c is easiest
[19:10] <toad_> b is best
[19:10] <toad_> a is in the middle
[19:10] <toad_> argh
[19:10] <toad_> we need hash.remove
[19:10] <toad_> we need hashtable remove
[19:10] <d-ArkAngel> sorry, I meant hints that b is in the right direction :-)
[19:10] <d-ArkAngel> lol
[19:10] <toad_> so b is impractical
[19:11] <d-ArkAngel> so a it is then?
[19:11] <toad_> yes
[19:12] <d-ArkAngel> [OT] Another triumph for the M$ owns the internet ... http://www.theregister.co.uk/2004/09/07/microsoft_patents_keyboard_navigation/
[19:13] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) has joined #freenet
[19:14] <d-ArkAngel> I guess we'll all just have to keep our mice in good working order :-)
[19:16] * Sugardude (foobar@outpost.zedz.net) has joined #freenet
[19:16] * Sugadude (~Sugadude@207-234-129-95.ptr.primarydns.com) Quit (Nick collision from services.)
[19:16] * Sugardude is now known as Sugadude
[19:23] <i2p_iip> <gott> toad
[19:23] <i2p_iip> <gott> there's something wrong with inserts taking 10 hours to insert 5 megs
[19:23] <i2p_iip> <gott> I was inserting a 5 meg file. I started it around 10 hours ago. It has not finished yet ( 97 % done though ! )
[19:23] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[19:24] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[19:26] <toad_> gott: yes
[19:26] <toad_> M$ patents what ?
[19:27] <toad_> bullshit, lynx existed before 1997
[19:27] <d-ArkAngel> using the arrow keys to pick a link on a page.
[19:27] <toad_> w3 probably used the keyboard in fact
[19:27] <d-ArkAngel> :-)
[19:27] * toad_ looks forward to it being invalidated in court...
[19:27] <d-ArkAngel> I'm just glad europe still haven't been crazy enough to allow patents on software.
[19:28] <d-ArkAngel> (tho the amount of lobby pressure they're getting it looks almost inevitable. :-( )
[19:28] <toad_> |Phelps told a legal conference last year that "You don't just get patents for the sake of getting patents," shortly before he joined Microsoft.
[19:28] <toad_> ROFL
[19:28] <toad_> d-ArkAngel: they will :(
[19:28] <toad_> it's a matter of how long we can hold it off for
[19:28] <toad_> and there is a massive amount to be gained by holding it off
[19:28] <toad_> but if FFII calls another demo, i'm there - regardless of my financial status at the time
[19:29] <mikeDOTd> CANNOT RENAME NODE FILE node-temp TO node <-- showing up in console, but not in logs. is this serious?
[19:29] <toad_> mikeDOTd: hmmm
[19:29] <d-ArkAngel> the problem is that as soon as it comes in, then MS and thier breatheren can start filing suits like crazy against linux and the pattent less.
[19:29] <toad_> probably a bug...
[19:29] <cbreak> here to... does not seem to be serious...
[19:29] <toad_> d-ArkAngel: well patents like THAT one aren't filed to sue with
[19:29] <toad_> they're filed to intimidate wth
[19:29] <toad_> because it's plainly bullshit to anyone who knows anything about the web
[19:29] <d-ArkAngel> think about who can afford the lawyers here?
[19:30] <toad_> right
[19:30] <d-ArkAngel> it's not the open source project owners.
[19:30] <toad_> intimidate
[19:30] <toad_> threaten
[19:30] <toad_> not sue
[19:30] <d-ArkAngel> if they do it now, then pattents will never be let through in EU, but they wait till it becomes legit here, then they'll have a bonanza
[19:30] <toad_> if they did me for that, I'd tell them to bring it on, and get legal aid, or even somebody on a no-win-no-fee basis, very possibly, because it's a clear-cut case afaics
[19:31] <toad_> anyway there's always the EFF...
[19:31] <toad_> d-ArkAngel: some of the OSS project owners can, it depends on who owns the projects :)
[19:31] <d-ArkAngel> the problem is that if there's a landslide of attacks, then EFF and thier firends can't keep it up on all fronts, and it could set back linux -> desktop years.
[19:32] <toad_> d-ArkAngel: if they're all as trivial as this, they could win them all, in reasonably short timeframe
[19:32] <toad_> hmmm
[19:32] <toad_> i've lost your email with the patch
[19:32] <toad_> please send it again
[19:32] <d-ArkAngel> lol
[19:32] <d-ArkAngel> sure.
[19:33] <d-ArkAngel> it's winging it's way.
[19:37] <toad_> thanks
[19:38] <toad_> my fault; i still had it but didn't find it...
[19:38] <toad_> running...
[19:38] <toad_> cycles slow down very fast...
[19:39] <toad_> adding sev eral hundred millis per node added...
[19:40] <d-ArkAngel> try dropping you HTL to 5 also are you interested in the keys still in network output (that code looks quite hevey to me)
[19:41] <d-ArkAngel> anyway, I think I said I was going to bed an hour or so ago, so now I will :-)
[19:42] <d-ArkAngel> night all.
[19:43] * toad_ switches the RNG used to a plain PRNG
[19:43] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) Quit (Remote closed the connection)
[19:43] <toad_> d-ArkAngel: good point
[19:43] <toad_> I'll make it run every 8 runs
[19:43] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) Quit ("leaving")
[19:44] <toad_> dropping htl to 5 would not work with 400 nodes...
[19:44] <toad_> at least it's unlikely imho
[19:46] <i2p_iip> <gott> toad, on a scale of 1 to 10, how pretentious are you ? ( serious question )
[19:48] <toad_> define pretentious
[19:49] <i2p_iip> <gott> do you go to upscale coffeehouses which play 'post-modern music' to pay 4 or 5 dollars for a cup of coffee, where the waitors all think they are better than you ?
[19:50] <toad_> no
[19:50] <toad_> I very rarely visit upmarket eateries/drinkeries
[19:51] <i2p_iip> <gott> I just returned from one.
[19:51] <toad_> bbiab
[19:51] <i2p_iip> <gott> The management and I had a 'misunderstanding'.
[19:51] <toad_> lol
[19:51] <i2p_iip> <gott> Although, I understood quite well : they are a shoal of cunts.
[19:52] <i2p_iip> <gott> fucking cunts. when I join the European Defense Initiative, they are being bombed first when the EU declares war on North America
[19:52] <i2p_iip> <gott> no euro-skeptic here
[19:52] <i2p_iip> <gott> cunts
[19:52] <i2p_iip> <gott> cunts
[19:52] <i2p_iip> <gott> anyway, I'm off.
[19:55] <toad_> gott: eh?
[19:55] <toad_> what's the EDI?
[19:56] <toad_> uh, yeah, sure...
[19:57] * toad_ sticks to cyberspatial delusions...
[19:57] <toad_> cycles get real slow real fast...
[19:57] <toad_> I wonder if this is just a problem with my CPU? d-ArkAngel said his settled at around 30 seconds at 400
[19:58] <toad_> mine have reached 30 seconds and are still rising at 80
[19:58] <toad_> maybe I need an opteron... how convenient that I've been leant shells on two...
[19:58] <toad_> both DP...
[19:58] <toad_> since there's no way i can afford to buy one of my own...
[19:59] <toad_> hmmm
[20:00] <toad_> http://amphibian.dyndns.org/simulations2-new.png
[20:00] <toad_> HRRRRM
[20:01] <toad_> it's definitely falling in the super-long term
[20:01] <toad_> the question is, is there an asymptote at around 0.95, or will it keep falling until it reaches the random-routing line?
[20:01] <toad_> only (absurdly long) time will tell :|
[20:12] <toad_> and that's for the best-case simulations...
[20:12] <toad_> the 100x100 simulations
[20:13] <toad_> sometime in the new year i'll get an opteron though, even if it's a single processor - they seem to be very suited to simulations
[20:13] <toad_> now, next job?
[20:13] <toad_> the ability to serialize the whole simulation to disk and then start it up again would seem to me to be very useful...
[20:19] * lolo-laptop (~lostlogic@68.251.84.226) Quit ("Client exiting")
[20:32] <Ash-Fox> God... I need better coloring skills
[20:39] * Sugardude (~Sugadude@207-234-129-95.ptr.primarydns.com) has joined #freenet
[20:39] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[20:39] * Sugadude (foobar@outpost.zedz.net) Quit (Nick collision from services.)
[20:39] * Sugardude is now known as Sugadude
[20:40] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[20:40] * Gunnar (~gunnar@34.80-203-209.nextgentel.com) Quit (Remote closed the connection)
[20:51] * Ash-Fox (Hal-9000@aap86.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[20:51] * Ash-Fox (Hal-9000@aah119.neoplus.adsl.tpnet.pl) has joined #FreeNET
[20:54] * Sugadude (~Sugadude@207-234-129-95.ptr.primarydns.com) Quit (Remote closed the connection)
[20:54] * Sugadude (tor@cassandra.eecs.harvard.edu) has joined #freenet
[21:09] * kiwi_uk (~kiwi_uk@212.219.225.104) has joined #freenet
[21:10] <kiwi_uk> When I'm browsing freesites, the freenet web control interface will not load until I've stopped some freesites from loading. I assume this is because of a connection limit of some sort... can't find anything in freenet.conf though
[21:10] <kiwi_uk> anyone know anything about this?
[21:17] <Sugadude> kiwi_uk: This is because of a connection limit of your browser.
[21:18] <Sugadude> Most browsers will only have a limited number of connections for a server in order not to flood it.
[21:22] * jayo (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[21:22] * jayo is now known as jay
[21:27] * Gunnar (~gunnar@34.80-203-209.nextgentel.com) has joined #freenet
[21:32] <Ash-Fox> which is why you must hax0rz t3h browser!
[21:33] * Ash-Fox thinks that freenet should come with a customised copy of dillo
[21:45] * guido^pe (~unknown@dsl-082-082-159-162.arcor-ip.net) Quit (Remote closed the connection)
[21:47] * cStyle (~cstyle@host-23-236-220-24.midco.net) has joined #freenet
[21:48] <cStyle> hell
[21:48] <cStyle> hello, anyone here?
[21:49] <toad_> hello
[21:49] <jay> it depends
[21:49] <toad_> not really here but here transitorilly
[21:50] <toad_> what's dillo? the really low footprint one?
[21:50] <Ash-Fox> it's a browser that is under 1mb in size
[21:50] <cStyle> i have a question, i have been using freenet for about a week without a problem, now i upgraded to 5096 and had problems
[21:50] <Ash-Fox> and it works just as well as any other browser, except not CSS support
[21:50] <Ash-Fox> *no
[21:50] <cStyle> i uninstalled and reinstalled java and freenet, but now can not get any clients to connect to me
[21:51] <Ash-Fox> cStyle, fix your firewall/port forwards
[21:51] <cStyle> i checked that
[21:51] <cStyle> i set the port to the same port as the last installed
[21:51] <cStyle> by using the "node address" and "port"
[21:52] <cStyle> should i change it in the config file as well as compared to the GUI?
[21:52] <Ash-Fox> argh
[21:52] <Ash-Fox> I can't find any image application that will open a 924163x24200042 res image without crashing >:/
[21:52] <cStyle> that is a large picture
[21:52] <cStyle> have you tried vueprint?
[21:53] <Ash-Fox> yes
[21:53] * moskau23 (~Miranda@dsl-082-082-232-104.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[21:53] <cStyle> any ideas on what else i should try
[21:53] <Ash-Fox> nope
[21:53] <cStyle> :(
[21:53] <jay> cStyle: re-seed with latest seednodes
[21:53] <Ash-Fox> :/
[21:54] <cStyle> who should i get the file from?
[21:54] <cStyle> the website, or someone esle?
[21:54] <jay> cStyle: ideally from someone here
[21:54] <jay> or the website
[21:54] <cStyle> can soneone dcc it to me?
[21:54] <Ash-Fox> *Sigh* I spend 8 days, leeching a image off a millitary satelite, and I can't even display it :/
[21:55] <cStyle> i downloaded a new copy for the project 2 days ago. i will try again though
[21:55] <jay> cStyle: yer on windows?
[21:55] <cStyle> yes
[21:55] <jay> cStyle: u sure it's even running?
[21:55] <cStyle> 24 bit collor?
[21:55] <cStyle> yes
[21:56] <cStyle> i had ti workign and had a store of 2 GB
[21:56] <cStyle> but after the 96 update, no luck, first the app crashed when i tried to open the config
[21:56] <cStyle> now i can config, but no connect
[21:57] <jay> can u get the gateway page from here: http://localhost:8888
[21:57] <cStyle> 924 163 * 24 200 042 * 3 = 6.70943502 ? 10^13 that is one damn big pictuere
[21:57] <jay> the first app crashed? which app?
[21:57] <Ash-Fox> cStyle 48bit color PNG
[21:57] <cStyle> ok, i was assuming 24
[21:57] <cStyle> how large is it?
[21:58] <Ash-Fox> very
[21:58] <jay> cStyle: if it's running, how many connections do u see from here: http://localhost:8888/servlet/nodeinfo/networking/ocm
[21:59] <Ash-Fox> I'd give you the file size, but it would take a very long time to get the result
[21:59] <cStyle> :) well, i will try some moreonnections [Switch to peers mode]
[21:59] <cStyle> Tue Sep 28 19:59:44 CDT 2004
[21:59] <cStyle> [More details]
[21:59] <cStyle> Connections open (Inbound/Outbound/Limit) 0 (0/0/139)
[21:59] <jay> cStyle: err wait
[21:59] <cStyle> 0 connections
[21:59] <jay> don't flood the channel
[21:59] * Oggy (user@69.5.227.10) has joined #Freenet
[21:59] <jay> ah ok
[21:59] <Ash-Fox> cStyle, it's stored on a tape device
[21:59] <cStyle> oh, i understand
[21:59] * Ogredude (user@69.5.227.10) Quit (Read error: 232 (Connection reset by peer))
[21:59] <cStyle> well, i got to go, thanks for the help
[22:00] <jay> need more time
[22:00] <jay> cStyle: no inbound nor outboud; check freenet.log in the install directory
[22:04] * Redb3ard (~oylerj@c-24-125-12-101.va.client2.attbi.com) Quit (Read error: 113 (No route to host))
[22:05] <jay> Connections open (Inbound/Outbound/Limit) 68 (36/32/200)
[22:06] <jay> toad_: were u able to import my gpg key?
[22:10] * Gunnar (~gunnar@34.80-203-209.nextgentel.com) Quit (Remote closed the connection)
[23:12] * Gunnar (~gunnar@34.80-203-209.nextgentel.com) has joined #freenet
[23:29] * Usurp (syl@trudaine-8-82-230-34-86.fbx.proxad.net) Quit (Read error: 60 (Operation timed out))
[23:54] <cStyle> anyone here?
[23:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[23:54] * sanity (~ian@81-178-94-146.dsl.pipex.com) has joined #freenet
[23:55] <cStyle> how do i fix this: Sep 28, 2004 9:54:08 PM (freenet.node.states.maintenance.Checkpoint, YThread-7, ERROR): unhandled throwable in Checkpoint: Writing NGRoutingTable global estimator: java.lang.NullPointerExceptionjava.lang.NullPointerException
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.
