Timestamps are in GMT/BST.
[0:26] * salahx (salahx@sc1-24.217.174.147.charter-stl.com) Quit ()
[0:44] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[1:13] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) Quit ("Leaving")
[1:23] * proteusguy (~proteusgu@68-114-24-230.cpe.ga.charter.com) has left #freenet
[1:36] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[2:30] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) has joined #freenet
[2:48] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 110 (Connection timed out))
[3:36] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[4:18] * Sugadude (~Sugadude@207.36.86.132) Quit ()
[4:22] * Zorix- (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[4:29] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 104 (Connection reset by peer))
[4:29] * Zorix- (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) Quit (Read error: 104 (Connection reset by peer))
[4:48] * AlexK (~AlexK@p83.129.200.226.tisdip.tiscali.de) has joined #freenet
[4:48] * AlexK (~AlexK@p83.129.200.226.tisdip.tiscali.de) has left #freenet
[4:54] * kers (~kers@193.ppp144.rsd.worldonline.se) has joined #freenet
[5:05] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) has joined #freenet
[5:20] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[5:29] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) has joined #freenet
[6:27] * kers (~kers@193.ppp144.rsd.worldonline.se) Quit ("Leaving")
[6:50] * pupok (~r00t@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[6:50] * pupok (~r00t@81-178-112-130.dsl.pipex.com) has joined #freenet
[7:23] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[7:30] * greycat (~wooledg@192.35.79.70) has joined #freenet
[7:41] * Ash-Fox (Hal-9000@abe180.neoplus.adsl.tpnet.pl) Quit (Read error: 104 (Connection reset by peer))
[7:50] * lolo-laptop (~lostlogic@68.251.84.226) has joined #freenet
[8:18] * Ash-Fox (Hal-9000@abe180.neoplus.adsl.tpnet.pl) has joined #FreeNET
[8:30] * FallingBuzzard (~srademach@207.152.112.129) has joined #freenet
[8:32] * FallingBuzzard (~srademach@207.152.112.129) has left #freenet
[8:40] * kers (~kers@219.ppp135.rsd.worldonline.se) has joined #freenet
[8:51] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit ()
[8:52] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[8:52] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[8:59] <sanity> anyone awake?
[9:00] <hobx_> me
[9:02] <sanity> howdy
[9:02] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[9:05] * Sugadude (~Sugadude@207-234-129-95.ptr.primarydns.com) has joined #freenet
[9:32] * FallingBuzzard (~srademach@207.152.112.129) has joined #freenet
[9:34] * kers (~kers@219.ppp135.rsd.worldonline.se) Quit ("Leaving")
[9:45] * yyop (~yyop@82.152.32.90) has joined #freenet
[9:48] * yyop (~yyop@82.152.32.90) has left #freenet
[10:21] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[10:21] <greycat> Ribbit.
[10:29] * fimp (~fimp@0x50a43685.aaanxx1.adsl-dhcp.tele.dk) has joined #freenet
[10:29] <fimp> hi
[10:30] <fimp> why wont my linux freenet-installation only start once? the next time I want to run freenet, it does seem to start, but I cant access localhost:8888 - I have to reinstall freenet every time. Why?
[10:30] <fimp> *why DO my linux freenet-installation only start once...
[10:30] <greycat> read the logs
[10:31] <greycat> SSK@wlUk5niRuNDUL4lzvdHu6DamhW8PAgM/Greg_Wooledge// is back in.
[10:31] <greycat> (stable network)
[10:32] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[10:32] <fimp> ah, Ill try next time :)
[10:32] <i2p_iip> <Sonax> greycat: Nice to see you back. Missed your flog/site
[10:32] <greycat> thanks...
[10:34] <fimp> are there any public nodes availible?
[10:34] <fimp> to be able to use freenet, when not at home, etc
[10:35] <greycat> ssh tunnels make your freenet node work from far away
[10:37] <fimp> but my computer is not on 24/7
[10:38] <greycat> that can be fixed,
[10:38] <greycat> unless you're in an area where you have electrical power issues, in which case you may need to invest in a UPS too
[10:39] <fimp> i dont want my computer to be on 24/7
[10:40] <fimp> I just thought that some people would have made their nodes public, allowing people to use freenet even when not at home
[10:40] <greycat> the demand for public gateway nodes would cripple anything pretty quickly
[10:40] <fimp> cripple what?
[10:40] <greycat> your computer, your internet connection
[10:41] <greycat> unless you're on a dedicated OC-3 or something, with no per-megabyte bandwidth charges
[10:41] <fimp> people wont use it if it gets too crippled. it would regulate itself i guess
[10:42] <greycat> well then, go ahead and set one up... ;-)
[10:42] <fimp> if i was able, I wouldnt be asking such questions
[10:44] <i2p_iip> <Sonax> fimp: This might not help you much, but there is a "public" node avaliabil from whitin the i2p network. But you would need to get on i2p then when not "at home", so you would not win much there...
[10:45] <i2p_iip> <Sonax> But some months ago there were several public nodes. But i guess they have disapeared... There would also be legal issues for people running such public nodes.
[10:55] <fimp> legal issues? how is that?
[10:57] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) Quit ("Leaving")
[11:01] <i2p_iip> <Sonax> fimp: Imagine i request illegal material through your public node. Then i'm quite shure of what key to look for in your data store when i get the cia/fbi/pet to confinscate your box.
[11:02] * kers (~kers@157.ppp144.rsd.worldonline.se) has joined #freenet
[11:19] <fimp> but I would have nothing to do with it
[11:21] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[11:21] <toad_> hmm, some interesting flog entries today:
[11:21] <toad_> Hmm, and so it begins... i have started censoring FIND. Try and see if you can find out what i have removed. If it is a problem with you, run your own (and if you think i have removed somthing vital, get a hold of me and tell me what you think should be let back in and i will consider.)
[11:22] <toad_> and CofE's longer rant on inserts
[11:23] <fimp> freenet do need some censored indexs
[11:24] <i2p_iip> <Sonax> Toad: I'm not that bad. Really :) I just started removing some sites that don't do editions the "right" way, so they would show up several times. And some CHK sites that got picked up by the spider while the "parent" SSK site was also included. The last part by request by the author.
[11:25] <i2p_iip> <Sonax> fimp: Make you'r own! No time like the present to get it done. I'm shure you will get all the help you want if you ask.
[11:25] * toad_ thinks the former is censorship, the latter is not
[11:25] <toad_> unfortunately censored indexes really need to be run by anonymous people
[11:25] <toad_> people who take their anonymity VERY seriously
[11:28] <toad_> because if you are traceable, you will be traced, and made to censor things that They don't like
[11:28] <toad_> as well as things that you don't like
[11:28] <toad_> so feel free to run one, just be careful
[11:28] <i2p_iip> <Sonax> or looking in the dontsearch.txt file, 103 versions of the same site.
[11:28] <toad_> i.e. use I2P to connect to here and ask how to do it, using a different nick
[11:32] <kers> oh fuck, i haven't noticed that part about the public irc logs in the topic before :o
[11:32] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[11:32] <toad_> kers: is that a problem?
[11:34] * toad_ changes topic to 'Upgrade to 5095 (5089 is mandatory, more recent builds have queueing, major queueing bugfixes, fixes to seednodes generation; upgrade ASAP!) | Unstable: Upgrade to 60231 | Channel logs: http://newton.matcmp.ncc.edu/~lockej/freenet/chanlog/ | #freenet-politics and #freenet-chat are available for offtopic discussions'
[11:35] <kers> toad_: not really, i just wasn't aware they were available on the web
[11:43] * kers (~kers@157.ppp144.rsd.worldonline.se) Quit ("Leaving")
[11:46] * fimp (~fimp@0x50a43685.aaanxx1.adsl-dhcp.tele.dk) Quit ("Leaving")
[11:57] * Pandora (~BoxPandor@64.147.163.245) has joined #freenet
[11:58] * Pandora (~BoxPandor@64.147.163.245) Quit (Client Quit)
[12:11] * nextgens (~nextgens@d80-170-176-64.cust.tele2.fr) has joined #freenet
[12:13] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[12:16] * Ash-Fox (Hal-9000@abe180.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[12:16] * Ash-Fox (Hal-9000@aay171.neoplus.adsl.tpnet.pl) has joined #FreeNET
[12:18] <nextgens> hi all
[12:29] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has joined #freenet
[12:32] * unknown_ (~unknown@dsl-082-082-157-234.arcor-ip.net) has joined #freenet
[12:32] * unknown_ (~unknown@dsl-082-082-157-234.arcor-ip.net) Quit (Client Quit)
[12:34] * guido^pe (~unknown@dsl-213-023-233-040.arcor-ip.net) Quit (Read error: 104 (Connection reset by peer))
[12:35] * guido^pe (~unknown@dsl-082-082-157-234.arcor-ip.net) has joined #freenet
[12:41] * nextgens (~nextgens@d80-170-176-64.cust.tele2.fr) Quit ("Download Gaim: http://gaim.sourceforge.net/")
[12:41] * nextgens (~nextgens@d80-170-139-43.cust.tele2.fr) has joined #freenet
[12:54] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) Quit (Remote closed the connection)
[12:56] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[12:57] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) has joined #freenet
[12:58] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[13:04] * robilad (~topic@mpiat2313.ag2.mpi-sb.mpg.de) Quit (Remote closed the connection)
[13:19] * kers (~kers@246.ppp135.rsd.worldonline.se) has joined #freenet
[13:24] <jay> ..
[13:32] <jay> wake up all you non-conformists
[13:37] * hobx__ (~chatzilla@ankh.math.chalmers.se) has joined #freenet
[13:37] <hobx__> Programing problem:
[13:38] <hobx__> From an int, create a float whose binary decimal expansion is the decimal expansion of the int in reverse.
[13:39] <greycat> what is a "binary decimal expansion"?
[13:43] <hobx__> a dumb term that is actually used
[13:43] <hobx__> for writing 1/4 as .01 etc
[13:44] <hobx__> there is no quicker way then reversing the bits I think
[14:00] <hobx__> Not perfect, but it works:
[14:00] <hobx__> int s = Integer.parseInt(args[0]);
[14:00] <hobx__> double r = 0;
[14:00] <hobx__> for (int i = 0 ; i < 32 ; i++) {
[14:00] <hobx__> if ((((long) s) & (1l << i)) != 0)
[14:00] <hobx__> r += Math.pow(0.5,i + 1);
[14:00] <hobx__> }
[14:01] <hobx__> Can lose the casting:
[14:02] <hobx__> double r = 0;
[14:02] <hobx__> for (int i = 0 ; i < 32 ; i++) {
[14:02] <hobx__> if ((s & (1 << i)) != 0)
[14:02] <hobx__> r += Math.pow(0.5,i + 1);
[14:02] <hobx__> }
[14:20] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[14:21] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[14:32] <toad_> hi ppl
[14:34] * toad_ has a dilemma: whether to go to the trade justice demo on sunday. two problems: 1) 1st day of labour party conference. there will therefore be a large pro-hunting demo, an enormous anti-war demo, and a smaller trade justice demo. the problem is, will this just turn into a major riot - pro-hunting people against everyone else? 2) prior commitments, which can be ignored if needed.
[14:35] <jay> any demo nowadays can be provoked into violence it seems
[14:35] <nextgens> toad_:hi
[14:35] <nextgens> toad_: insert is working better here than downloads :-? is it 'normal'?
[14:35] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[14:35] <toad_> nextgens: seems unlikely, everyone else tells me the opposite
[14:36] <jay> the ones in new york were peaceful until they were provoked, just like past demos in NY
[14:36] <toad_> jay: sure it CAN...
[14:36] <toad_> there are two different scenarios here
[14:36] <nextgens> toad_: do you want a Fuquid screenshoot?
[14:36] <nextgens> Fquid
[14:36] <toad_> one is the militant branch of the anarchist/anti-war contingent go attack some multi-national-coporation outposts
[14:36] <toad_> that's what always happens at demos
[14:36] <toad_> but it's a minor problem
[14:37] <jay> toad_: im just saying, going to demos today means a good chance of being arrested
[14:37] <toad_> it usually happens in the evening after the main demo has passed off peacefully
[14:37] <jay> http://www.2600.com/rnc2004/
[14:37] <toad_> it's worth the risk. the major concern is this: 100,000 anti-war demonstrators, maybe 10k trade justice demonstrators, and 10k pro-hunting demonstrators, within bottle throwing range
[14:38] <jay> heh
[14:38] <toad_> if some bright spark amongst the pro-hunting demonstrators, most of whom are working in agriculture and are therefore "hard men"...
[14:38] <toad_> or some of the anarchists on our side...
[14:39] <toad_> decide that a punch up would be fun, then it could get ugly fast
[14:39] <toad_> I wouldn't want to, of course - there's no point, we've won the hunting argument
[14:39] <toad_> but there are always nutters...
[14:39] <nextgens> *doing a screenshoot
[14:40] <jay> nutters? is that brit for 'crazies'?
[14:41] <toad_> you get the picture? I don't really mind spending a night in one of HM's cells if it's for a good cause... I'm a little concerned about the effect of tear gas on asthma though.. but those aren't the big concern
[14:41] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[14:41] <toad_> jay: yeah
[14:41] <jay> yeah the tear gas is a factor i would think
[14:41] <jay> what do they use anyway?
[14:41] <toad_> CS i think
[14:41] <toad_> well no
[14:41] <toad_> CS they use close up
[14:41] <jay> keep in mind im clueless in these areas
[14:41] <toad_> don't know what they use for long range
[14:42] <jay> what's CS?
[14:42] <toad_> but it's always referred to as "tear gas"
[14:42] <jay> counter strike?
[14:42] <jay> hehe
[14:42] <toad_> jay: chemical pepper spray
[14:42] <toad_> can be pretty nasty, but so can a truncheon
[14:42] <jay> they aren't doing that in nyc it seems
[14:43] <jay> they're just arresting people blindly
[14:43] <toad_> Can't write to file "freenet.jpg"
[14:43] <toad_> hmmm
[14:43] <toad_> well the police here are pretty good since genoa...
[14:43] <toad_> :)
[14:43] <jay> nyc police are pretty well armed and modernized
[14:43] <jay> they're practically military
[14:44] <toad_> if they have any sense they'll make sure the hunting demo is kept as far away from the rest as possible.. although everyone seems to want to use the beach...
[14:44] <jay> and we have military in nyc also
[14:44] <toad_> figures
[14:44] <jay> they all talk to each other and are friendly
[14:44] <toad_> :)
[14:44] <jay> and we're like animals in the zoo to them
[14:45] <jay> they got some kind of tazer i believe
[14:45] <jay> and rubber bullets
[14:45] <jay> i dont feel really tough around them ;)
[14:45] <toad_> :)
[14:46] <jay> or even bean bag bullets
[14:46] <jay> like in the movie jackass
[14:50] <toad_> bean bag bullets?
[14:50] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[14:50] <jay> yup
[14:50] <jay> almost lethal
[14:50] * toad_ mails the local transport coordinator for his/her opinion
[14:51] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[14:52] <toad_> hi sanity
[14:52] <toad_> any chance of staying long enough to say something? :)
[14:53] <toad_> The project's current Paypal balance is $586.74
[14:53] <hobx__> that wasn't a join. That was a broken dsl modem
[14:53] <toad_> yay
[14:54] <hobx__> eh fuck it
[14:54] <hobx__> I teach in the afternoon tomorrow, I'll prepare through lunch
[14:55] * nextgens is now known as nextgens|busy
[14:56] <hobx__> yeah, 15 to 17 even.
[14:56] <hobx__> NP.
[14:59] * jabawok_w (~jabawok_w@203-59-216-170.dyn.iinet.net.au) Quit (Read error: 113 (No route to host))
[15:02] * hobx_ (~chatzilla@ankh.math.chalmers.se) Quit (Ping timeout: 14400 seconds)
[15:05] * nextgens|busy is now known as nextgens
[15:05] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[15:06] * nextgens (~nextgens@d80-170-139-43.cust.tele2.fr) Quit ("bye")
[15:06] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[15:32] <greycat> RNFs are depressing.
[15:34] <toad_> greycat: very.. when are you getting these? how many open conns? what approximate local traffic level?
[15:35] <greycat> 1) now. 2) 33. 3) Smoothed local mean traffic (queries per hour): 12173.210
[15:38] <toad_> yes i mean are you hammering it with frost?
[15:38] <toad_> and are you getting them on inserts, or on requests?
[15:39] <greycat> I don't use frost. the RNFs are on both retrievals and inserts. I'm trying to insert my freesite (just one file changed, the rest collide), with a few occasional freesite requests in a browser at the same time.
[15:39] <greycat> prior to that, I was doing much heavier freesite requesting.
[15:40] <greycat> I had to set maxNodeConnections=40 to avoid having freenet turn my internet connection into mush
[15:41] <toad_> hmmm
[15:41] <toad_> curious
[15:42] <toad_> 33 conns ought to be enough
[15:42] <toad_> maybe it's the inserts...
[15:45] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit ("/")
[15:45] * jabawok_w (~jabawok_w@203-59-126-22.dyn.iinet.net.au) has joined #freenet
[16:03] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[16:03] * Hory (~Miranda@81.196.25.110) Quit (Remote closed the connection)
[16:20] * greycat (~wooledg@192.35.79.70) Quit ("Lick Bush in '04")
[16:20] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit (Read error: 54 (Connection reset by peer))
[16:21] * sanity (~ian@81-178-112-130.dsl.pipex.com) has joined #freenet
[16:34] <Ash-Fox> http://anandtech.com/linux/showdoc.aspx?i=2213 proof that intel is better than amd
[16:35] * pupok (~r00t@81-178-112-130.dsl.pipex.com) Quit (Read error: 104 (Connection reset by peer))
[16:38] <toad_> uhh
[16:38] <toad_> i have a hard time believing a 3.4GHz P4 even with 2MB of cache beats an FX-53 in long mode...
[16:39] <toad_> the mysql benchmarks are whacky
[16:40] <toad_> either there's a measurement error or there's some serious issues with the chipsets...
[16:40] <toad_> ohhh
[16:40] <toad_> doh
[16:40] * toad_ reading it wrong - less is better ;)
[16:42] <toad_> [ot] also mental ray ought to be significantly faster in 64 bit mode - k8's have two SSE2 units, you can only use one in 32-bit mode...
[16:42] * kers (~kers@246.ppp135.rsd.worldonline.se) Quit (Read error: 113 (No route to host))
[16:42] <toad_> [ot] povray gets a huge boost for exactly that reason...
[16:43] <toad_> [ot] Ash-Fox: those benchmarks clearly show AMD chips beat intel on most benchmarks - usually cheaper AMD chips too
[16:44] <toad_> [ot] especially on 64-bit
[16:44] <toad_> [ot] you're reading it wrong - check the "more is better" or "less is better"
[16:44] * pupok (~r00t@81-178-112-130.dsl.pipex.com) has joined #freenet
[16:46] <Ash-Fox> yay
[16:47] <toad_> [OT] this unfortunately is why AMD K8 chips are so fscking expensive at the moment :<
[16:47] <toad_> (which is why i don't have one!)
[16:47] * Ash-Fox only buys pentium, not going to have a proccessor that melts on him
[16:47] <Ash-Fox> had it happen once, not letting it happen again
[16:48] <toad_> Ash-Fox: yeah, intel had the edge up to the ~ 2000+'s because of that... AMD hadn't been rigorous enough about motherboard approvals
[16:49] <toad_> or they may even have messed up the reference design
[16:49] <toad_> anyway it got sorted out after ~ 2000+, via kt400 generation roughly - all modern boards have working thermal cutouts
[16:49] <toad_> i know, because I've caused a great many cutouts over the years :)
[16:50] <Ash-Fox> I know that at least if my server is going to be locked in a small isolated room for a few years without reboots, I can trust a intel system not to burn up
[16:50] * FallingBuzzard (~srademach@207.152.112.129) has left #freenet
[16:51] <toad_> [increasingly OT] [OT] Ash-Fox: a modern AMD system uses less power than an intel one and has equivalent thermal protection. the only difference is the intel one will throttle the clock down to 25% and continue running, whereas the AMD one will just shut down
[16:51] <toad_> [ot] but this is getting really off topic
[16:51] <Ash-Fox> I don't think anyone minds but you :P
[16:52] <toad_> An unusual problem occured while running the OpenSSL benchmark. Even though we are using Intel validated heatsink/fan combos we recieved a continuous stream of errors from the operating system regarding thermal temperatures. An example syslog can be seen below:
[16:52] <toad_> Message from syslogd@linux at Mon Sep 18 01:57:27 2004 ...
[16:52] <toad_> linux kernel: CPU#0: Temperature above threshold
[16:52] <toad_> This does not bode well for the processor. Our processor test bed is completely caseless, and if we have issuse with our 3.6GHz processor out of a normal case, we can't imagine what issues might exist in a full enclosure.
[16:52] <toad_> LOL, thermal problems in that benchmarking article, on the P4 ;)
[16:52] <toad_> http://anandtech.com/linux/showdoc.aspx?i=2213&p=8
[16:52] <toad_> at the bottom
[16:54] <Ash-Fox> heh
[16:56] <toad_> interesting that he mentions some compatibility issues with x86-64 at the end, he dual boots... irritating that that's not all sorted out by now
[16:57] <toad_> bbl
[17:06] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) Quit ("Iedereen vindt mij TOF!!!!!!!")
[17:47] <Archeron> anyone here running freebsd 5?
[17:48] <guido^pe> Archeron: I am. I never got Freenet running on it...
[17:48] <Archeron> no?
[17:49] <guido^pe> no, did you?
[17:49] <Archeron> I am having trouble with freenet causing kernel panics.
[17:49] <Archeron> it has run fine for a long time...
[17:50] <Archeron> but since the 2'nd of Sept, the thread code has been borked.
[17:50] <guido^pe> Wasn't that bad here. Over here, it would just consume CPU time without end and never show any actual life signs
[17:50] <Archeron> well... I am not sure... either threading or the SMP scheduling code.
[17:50] <guido^pe> (life signs: Port 8888 becoming available, the data store dir appearing...)
[17:51] <guido^pe> What jvm are you using?
[17:51] <Archeron> http://www.freebsd.org/cgi/query-pr.cgi?pr=71827
[17:52] * Zorix (Brandon@fl-69-69-7-231.dyn.sprint-hsd.net) has joined #freenet
[17:54] <guido^pe> That's the same Jvm I tried. I'm not using the beta version (of FreeBSD), though
[17:56] <guido^pe> Archeron: well, good luck to you, I'll go to bed now. cu
[17:56] * guido^pe (~unknown@dsl-082-082-157-234.arcor-ip.net) Quit (Remote closed the connection)
[18:42] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) has joined #freenet
[18:54] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) Quit (Remote closed the connection)
[19:19] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) has joined #freenet
[19:52] * Sugadude (~Sugadude@207-234-129-95.ptr.primarydns.com) Quit (Remote closed the connection)
[19:58] * d-ArkAngel (~robert@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) Quit ("using sirc version 2.211+KSIRC/1.3.10")
[20:18] <KenMan> what no chatters ? is freenet all fixed ?
[20:19] * KenMan is converting over to Fedora
[20:28] <toad_> hi KenMan
[20:29] * toad_ been mostly doing freenet-unrelated comp stuff today (backups, moving files around, etc)
[20:29] <toad_> KenMan: there are a number of problems though
[20:29] <toad_> 1. inserts REALLY suck.
[20:29] <toad_> 2. probably cause of 1: timeouts didn't take queueing into account.
[20:30] <toad_> 3. reports of infinite loops in queueing, not reproducible
[20:30] <toad_> 4. NaNs sent by 5095
[20:30] <toad_> for MRI
[20:30] <toad_> despite code to detect and avoid this
[20:31] <toad_> 5. transfer failed not handled right in fproxy in 5095; patch in unstable, also partially for others
[20:31] <KenMan> ugh !
[20:31] <KenMan> someone hacking the net ? or just faulty code ? I never like to predict on this matter.
[20:31] <toad_> 6. lots to do with the simulation, but it's rather slow still
[20:32] <toad_> KenMan: hmm, which one?
[20:32] <KenMan> #4 of course
[20:32] <toad_> seems to happen on many nodes
[20:32] <toad_> probably just a bug
[20:32] <toad_> simulator has produced all sorts of interesting stuff
[20:33] <KenMan> there were two channels in which to communicate an MRI, i guess there still are ?
[20:33] <toad_> ummm
[20:33] <toad_> which two?
[20:33] <toad_> there is MRIPacketMessage
[20:33] <KenMan> MRIPeerPacket and inside something else
[20:33] <KenMan> yeah, that train of thought :)
[20:33] <toad_> we got rid of the inside-an-FNP-message
[20:33] <toad_> in 5089
[20:33] <toad_> which was a reset
[20:34] <KenMan> okay. just checking. Maybe I'll try to track that one down then...
[20:34] <KenMan> I've been playing Fedora for a day or two.
[20:34] <toad_> i implemented FastSlidingBucketsKeyspaceEstimator which uses doubles instead of BigIntegers for keys in routing... 3-4x speedup for somewhat reduced but probably quite sufficient accuracy...
[20:34] <KenMan> yeah, i read about that. Good work!
[20:35] <toad_> 7. CofE is pissed off at 1. !
[20:35] <toad_> :)
[20:35] <KenMan> well, maybe because he is the only one who inserts... :?
[20:35] <toad_> various other things less prominent or less often reported
[20:35] <toad_> i'm not sure i can merge everything yet anyway...
[20:35] <toad_> KenMan: LOL
[20:36] <toad_> but i probably ought to merge something
[20:36] <KenMan> i'll fire up a 5095 and see what I can observe.
[20:36] <toad_> i have locally merged everything, and generated a changelog, but i need to test it and fix remaining sometimes serious bugs e.g. that infinite loop
[20:37] * jabawok_w (~jabawok_w@203-59-126-22.dyn.iinet.net.au) Quit (Read error: 110 (Connection timed out))
[20:37] <KenMan> you mean before you push it all into stable ?? or unstable ?
[20:37] <toad_> also i'm not sure i like 2.6's scheduling behaviour; the simulator thread seems to cause the node to slow down significantly despite being more niced
[20:37] <toad_> KenMan: stable
[20:37] <toad_> KenMan: the work since 5095, that is
[20:38] <KenMan> i thought those schedulers (AIO deadline CFQ etc) were PROCESS schedulers, not block IO schedulers !! hehe
[20:38] <toad_> heh
[20:38] <toad_> i think they have several process schedulers too
[20:39] <KenMan> sometimes you can learn something when you pay attention, but generally, I like to invent understandings based on nothing.
[20:39] <toad_> if so i probably need to try some of the others...
[20:39] <toad_> heh
[20:39] <KenMan> oh, why am i telling YOU this, you already know it :)
[20:39] <toad_> if i'd known how long it would take for my backup verification script to run.. i'd have invested more time optimising it...
[20:40] <KenMan> hashing ? (de)compression-ing ?
[20:40] <toad_> all of the above
[20:41] <toad_> i have daily backups for the last 3 months
[20:41] <toad_> these consist of a set of manifest files
[20:41] <toad_> including MD5s for each file
[20:41] <toad_> and I have a "datastore", which is indexed by MD5, with the two-hex-digit directories we know from freenet
[20:41] <toad_> in which the bzip2'd files live
[20:41] * KenMan is impressed, and lazy
[20:42] <KenMan> I just wait for my drives to crash, so I can get some exercise by swearing loudly. I don't get enough exercise.
[20:42] <toad_> hehe
[20:42] <KenMan> I've got 80Gs to be recovered when the $'s start flowing back into the old wallet...
[20:42] <toad_> this isn't my stuff - I have other less elaborate arrangements for my stuff
[20:43] <toad_> this is the semi-version-control system i built for dad's stuff
[20:43] <KenMan> ah, intriguing. You care more about his data than yours... or is it your 'reputation' ? ;)
[20:43] <toad_> well my _important_data is on the sourceforge CVS server
[20:44] <toad_> of course they don't back that up
[20:44] <toad_> but we do
[20:44] <toad_> hopefully many other people do too
[20:44] <toad_> my other local data is mostly stuff that doesn't matter that much
[20:44] <toad_> e.g. my mailboxes
[20:45] <KenMan> SF don't back it up ? or at least use heavy RAID redundancy ??
[20:45] <toad_> they back up the metadata
[20:45] <toad_> but not the actual cvs trees
[20:45] <KenMan> ugh, that's depressing to learn.
[20:45] <toad_> just backing up the metadata is a feat of engineering
[20:45] <toad_> anyway that's what their policy says
[20:45] <toad_> it's possible that they do backup the cvs trees
[20:46] <toad_> but just don't want to disappoint us if it fails
[20:46] <KenMan> pragmatic attitude
[20:47] <toad_> okay all done...
[20:47] <KenMan> how do I play with the simulator stuff you built ??
[20:47] <toad_> KenMan: src/freenet/node/simulator/
[20:47] <toad_> java the Main
[20:47] <toad_> you can pass a load of arguments which are in Main
[20:48] <toad_> then capture the output
[20:48] <toad_> and grep it
[20:48] <toad_> and graph it
[20:48] <toad_> i can send you scripts
[20:48] <KenMan> your main goal is/was to evaluate the estimators ?
[20:48] <toad_> i think i committed them actually
[20:48] <toad_> well and to evaluate NGR per se
[20:48] <toad_> and to evaluate different bootstrap strategies and so on
[20:49] <toad_> now, i think there is no power to my DVD writer... either that or it's bust...
[20:49] <toad_> or something wierd's happening
[20:49] <toad_> so i may need to reboot..
[20:50] <KenMan> unmount. Unbend a paperclip... hdparm -i /dev/hdX ... then swear...
[20:50] <toad_> yup, have to switch off... brb, ~ 10 mins
[20:50] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit ("using sirc version 2.211+KSIRC/1.3.11")
[21:02] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[21:02] <toad_> hi
[21:02] * toad_ back
[21:02] <toad_> hey KenMan
[21:02] <toad_> :)
[21:09] <KenMan> hi. I am still getting used to using GNOME for now... no more keyed-desktop switching, for now.
[21:10] <KenMan> lots to learn... in the meanwhile, lots to hate.
[21:11] <KenMan> I only want two desktop spaces. Previously I had one big wide (two screens' worth) pan-able desktop.
[21:12] <KenMan> Alt-left and Alt-right has been my mode of operation for 7 years, so it's hard to mentally rewire.
[21:13] <KenMan> the default iptables for Fedora seems kind of silly - everything is accepted. Else, everything is rejected. Yeah, that makes sense to me. (scratching my head)
[21:14] <toad_> hmmm
[21:14] * toad_ has never really got into gnome
[21:14] <toad_> last time i tried to use it on debian, all i remember is it insisted on having the menu at the top for all apps
[21:14] <toad_> and its find utility REALLY sucked
[21:15] <KenMan> it seems a little heavy, but... I'll give it a week or so before I decide whether to switch back to 'fvwm98' , or code up fvwmXP ...
[21:15] * toad_ wonders why you don't use kde...
[21:16] <KenMan> i've tried it once, but not long enough to know the major diffs between KDE & GNOME
[21:17] <KenMan> i don't see eclipse in the menus, or in the path. I thought it came with Fedora...
[21:17] * KenMan hasn't eclipsed for several months
[21:17] <toad_> curious
[21:18] <toad_> okay, that seems to have worked... just got to wait for it to backup a fresh (i.e. not previously stored) 291MB...
[21:19] <toad_> whereas there's massive overlap in the main 322MB
[21:19] <toad_> (3 months worth of backups of the 322MB takes up 200MB)
[21:19] <toad_> well more than 3 months actually
[21:19] <toad_> 3 months of daily snapshots
[21:19] <toad_> plus a few much earlier ones
[21:20] <toad_> hmmm this could take a while...
[21:20] <toad_> what to do in the meantime?
[21:20] * KenMan is long overdue for a decent backup...
[21:20] <toad_> well this is pretty extreme...
[21:20] <toad_> of course it's irregular
[21:20] <toad_> because i'm lazy
[21:21] <KenMan> go read www.linuxtoday.com , or http://www.gigaom.com/2004/09/the_voice_over_i.php
[21:21] <toad_> but it's pretty extreme to write your own hash datastore rotating backup system i suppose
[21:22] <KenMan> it keeps your coding synapses active
[21:22] <toad_> |The VoIP insurrection over the last decade marks a milestone in communication history no less dramatic than the arrival of the telephone in 1876.
[21:22] <toad_> uhmm...
[21:22] * toad_ thinks he'll read old freenet-tech mails instead
[21:23] <KenMan> heh
[21:23] <KenMan> yeah, it's kind of long. But it has an interesting perspective, so far.
[21:23] <toad_> this new data seems to consist mostly of full-page-scan JPEGs...
[21:23] <toad_> so it takes aaages for the backup script to not compress them
[21:24] <KenMan> oh, i guess you saw Sun's commitment to "open source" Solaris 10 ??
[21:24] <toad_> KenMan: that's pretty cool
[21:24] <KenMan> if and when it happens, sure.
[21:24] <toad_> ZFS sounded interesting, although I'm personally waiting for a stable merged reiser4.1 (in linux 2.7.50 or so)
[21:25] <toad_> reiser4.1 will have compression that actually INCREASES performance
[21:25] <toad_> a concept largely laughed at in computing for the last 20 years
[21:25] <toad_> but open sourcing solaris is very important strategically - next thing'll be the JVM
[21:25] <toad_> KenMan: any info on IBM's supposed open sourcing of dragonvoice?
[21:26] <toad_> or whatever it's called, IBM's recognition engine?
[21:26] * toad_ read in new sci but couldn't find anything close on /.
[21:26] <KenMan> I saw an AT&T online demo , and vaguely mighta heard about IBM's contribution, but no info.
[21:27] <KenMan> oh, i read too fast. Recognition, not text-to-speech... !!
[21:27] <toad_> KenMan: yeah, if it's true, it's a substantial gain
[21:27] <toad_> it's possible it's just a newsci reporter misinterpreting a /. headline though
[21:28] <KenMan> cool. I knew someone working with that IBM stuff several years ago, it did not impress me at the time.
[21:28] <toad_> a text-to-speech that doesn't sound crap would be nice too of course...
[21:28] <KenMan> My last update (5 years ago) was that if you loaded the interpreter with a 'context' it would better guess the correct (but limited) vocabulary.
[21:28] <toad_> KenMan: doesn't matter if it's not perfect. M$: "We will never see open source speech recognition, it's too sexy"
[21:29] <toad_> there's some big-biz apps - all the telephone booking stuff etc - that uses limited-vocabulary recognition
[21:29] <KenMan> MS burns my britches. They are so big, they burn a lot of britches...
[21:29] <toad_> i think that's what they were going to OSS - their limited-vocab engine
[21:30] <toad_> not the one-speaker any-text voice-control voice-dictation thingy
[21:30] <toad_> but they said it was worth $10M
[21:30] <KenMan> it is neat shit, but there's someone else with better voice technologies: "Them"
[21:30] <toad_> hehe, of course ;)
[21:30] <KenMan> i mean, beyond "mature technology" .
[21:31] <toad_> huh?
[21:31] <KenMan> well, when you invest heavily for many years, i suppose it pays off.
[21:31] <toad_> one cheat they undoubtedly use: you don't need to nail down the text, just recognize it phonetically, and then SEARCH PHONETICALLY
[21:31] <toad_> s/they/They
[21:32] <KenMan> yes, but they have semantics nailed 100%... as in , accurate interpretation.
[21:32] <toad_> you can do the same thing with OCR
[21:32] <toad_> uh huh. how do you know this?
[21:32] <KenMan> uh, i am just guessing.
[21:32] <toad_> ;)
[21:32] * toad_ doesn't think the NSA have an AI
[21:32] * KenMan think the NSA *is* an AI
[21:32] <toad_> they may be able to break AES-128 on the other hand
[21:33] <toad_> KenMan: well that's different ;)
[21:33] <toad_> KenMan: a government agency, a corporation, IS an AI - usually a very stupid one
[21:33] <toad_> but it's not a MACHINE intelligence :)
[21:33] * KenMan has tremendous respect for certain US agencies accomplishments and abilities
[21:34] <KenMan> so how is your backup ? all done ?
[21:34] <toad_> I wonder if they have a practical universal compromize for SHA-1?
[21:34] <toad_> KenMan: compressing huge full-page TIFFs at the moment
[21:35] <KenMan> SHA1, probably not. Practical means "cost efficient" and convenient. But capable, probably.
[21:35] <toad_> KenMan: why not?
[21:35] <toad_> KenMan: there are some birthday attacks (plaintext+hash->modified plaintext for same hash) attacks for SHA-0
[21:36] <KenMan> it depends how far down the 'fear of uncertainty and doubt' curve you lie.
[21:36] <toad_> fear of?
[21:36] <toad_> not fear and?
[21:36] <toad_> or fear comma?
[21:36] <KenMan> well, who was responsible for CREATING SHA 0 and SHA 1 ??
[21:37] <toad_> KenMan: that's my implication
[21:37] <KenMan> were they doing their best to serve their customer, or just performing a self-serving act ??
[21:37] <toad_> KenMan: civilian crypto is probably 10 years+ behind the NSA
[21:37] <toad_> but I'm not convinced that AT THE TIME they'd have put out something compromized as designed
[21:37] <toad_> that would arguably be against their mandate
[21:37] * jayo (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[21:38] <toad_> but just because it was clean 10 years ago... :)
[21:38] <toad_> hi jayo
[21:38] <KenMan> that is what research is all about !!
[21:38] <jayo> yo toad_
[21:38] * jayo is now known as jay
[21:38] <KenMan> hi jay
[21:39] <jay> KenMan: hey man.. haven't seen you in a couple of weeks
[21:39] <KenMan> i was away
[21:40] <toad_> Large-file bandwidth is actually pretty decent. I've downloaded a 650MB movie
[21:40] <toad_> in just a few hours using a fresh transient node on a cable modem.
[21:40] <toad_> woah
[21:40] <toad_> that's from an old mail on tech
[21:40] <KenMan> hopefully i'll be mostly away permanently - as in "employed" ...
[21:40] <toad_> I knew large files worked well.. but on a fresh transient?!!?!
[21:40] <toad_> KenMan: well please pop in from time to time e.g. weekly
[21:40] <toad_> so we can discuss latest things Freenet
[21:42] <toad_> like today ! ;)
[21:47] <jay> KenMan: work for anything government related.. apparantly they got lots of cash here in the states
[21:47] * KenMan on phone
[21:47] <jay> and their in debt.. go figure
[21:47] <jay> they're
[21:47] <jay> im supposed to be away.. bbiab
[21:48] <toad_> LOL jay
[21:48] <toad_> yay
[21:48] <toad_> it finished
[21:48] <toad_> the backup - and the verifications too
[21:49] <toad_> now to make the real backup (write to DVD)
[22:08] <toad_> this is pretty entertaining...
[22:09] <toad_> old thread with newbie with just enough clue to troll with (accidentally)
[22:11] <toad_> also the faith in NGR on the part of most participants is curious
[22:13] <KenMan> TCO argument ? http://www.iranian.ws/iran_news/publish/article_3822.shtml
[22:16] <toad_> > Is there just one? I thought it's meant like DC, that anyone can start a
[22:16] <toad_> > little network.
[22:16] <toad_> There is only one Freenet. It has several hundred thousand nodes.
[22:16] <toad_> LOL
[22:17] <toad_> tom K has come a long way :)
[22:17] <KenMan> long way from ?
[22:17] <toad_> n/m
[22:18] <KenMan> faith in NGR ? we follow the leader. I remember you deciding to abandon CPA in a moment's notice. I was *there* man!!
[22:18] <KenMan> ;)
[22:19] <KenMan> in fact, it might even have been my subtle (subconscious) suggestion.
[22:19] <toad_> KenMan: heh
[22:19] <toad_> it was a pretty long argument actually
[22:19] <toad_> > There is only one Freenet. It has several hundred thousand nodes.
[22:19] <toad_> Well what is the point of having it so large when you can't find all the
[22:19] <toad_> data in it if you wanted to?
[22:19] <toad_> LOL!
[22:21] <KenMan> now, I've never pretended to understand NGR fully, but, it seems to be aiming for CPA+
[22:22] <KenMan> as in, CPA + speed boost . That is my best guess as to a 'stated goal.'
[22:22] <toad_> not exactly
[22:22] <toad_> it does things significantly differently
[22:23] <KenMan> but the goal is pretty much the same, when it comes to routing, no ?
[22:23] <KenMan> I mean, I'm not trying to lead you into an argument...
[22:24] <KenMan> but maybe we have not focused on the desired functionality of a routing algorithm in sufficient detail.
[22:24] <KenMan> yet.
[22:24] <toad_> well that's what the simulations are for...
[22:25] <KenMan> the thing we seem to have come to learn, so far, is that a certain amount of information (requests) must be gathered before something meaningful can be derived from it.
[22:25] <toad_> yeah
[22:25] <toad_> but this varies depending on how we set things up
[22:25] <toad_> bootstrapping, the details of the estimators, etc
[22:25] <toad_> have you seen the graphs yet?
[22:25] <KenMan> after 10 requests of a route, we are still poking in the dark. After 100, we still can't see. After 1000, maybe the illumination begins to form.
[22:26] <KenMan> I looked at those estimator graphs you made (glad to see a GnuPlot) but I didn't really 'get it'
[22:26] <KenMan> they wanted to converge at a set level after X requests...
[22:26] <KenMan> maybe I did 'get it' :)
[22:27] <KenMan> you were surprised by the large X required to have convergence.
[22:28] <KenMan> that's what I took away from it. I don't understand each estimator though.
[22:29] <KenMan> NGR , or CPA, or whatever, is really just playing a game of chance. Trying to apply stats to the whole deal.
[22:29] <toad_> KenMan: chance?
[22:30] <toad_> the graphs clearly show random routing sucks a lot worse than NGR does
[22:30] <KenMan> I know why you say a node cannot choose it's own spec. In turn, that is why one cannot just state its spec. But what if we could work around the one problem ??
[22:30] <KenMan> With guaranteed redundancy or something ??
[22:30] <toad_> hmmm
[22:31] <toad_> I don't see why we can't make freenet work with heuristic specializations
[22:31] <toad_> and I see no end of problems in verifying a stated specialization
[22:31] <KenMan> i'm on your side :) whatever works...
[22:31] <toad_> unless you guarantee that a node with spec X will have all data near X
[22:31] <toad_> in which case you run into issues with LRU
[22:32] <toad_> I'm not saying something like Kademlia can never work
[22:32] <toad_> I'm saying Freenet can
[22:32] <KenMan> it must perform with a certain degree of effectiveness, that is all.
[22:32] <toad_> and perhaps it can work better
[22:32] <toad_> KenMan: huh?
[22:32] <KenMan> i mean, we cannot make a single absolute statement about another node. That is the 'game' part of it all.
[22:33] <toad_> yup
[22:33] <toad_> that's just life
[22:33] <toad_> I cannot say that KenMan isn't actually m0davis in disguise, for instance
[22:33] <KenMan> heh, how about Iakin ? is he still around, or busy working ??
[22:34] <KenMan> I've seen him 4 weeks since...
[22:35] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit (".")
[22:35] <KenMan> m0davis is dead, remember? Now that guy called Someone, I would tend to think he was soros, based on his email address. But I think soros is long gone.
[22:35] <KenMan> tessier is due to swing by about a week before thanksgiving.
[22:36] <KenMan> just for a check-up.
[22:36] <KenMan> and lolo is too busy with real life these days/months...
[22:36] <KenMan> who am i forgetting ? at least jay frequents this channel reliably.
[22:37] <toad_> there's a mathematical result that a DHT will scale with log IFF the connections between nodes have a distribution of 1/d
[22:37] <KenMan> thelema is off doing charitable service on the african continent...
[22:37] <toad_> recent simulations (not mine, ian's) suggest that LRU produces exactly this distribution
[22:38] <KenMan> conns have a dist ? you mean, they all have exactly d connections ? or am i misreading you ?
[22:38] <toad_> no
[22:39] <toad_> if the probability of connecting to a node is proportional to d, the distance between our spec and their spec
[22:39] <toad_> for all connections between nodes
[22:39] <toad_> then we get logarithmic hops
[22:39] <KenMan> oh... OH. that is important. but announcements seem to be humping that concept :(
[22:40] <toad_> simulations say LRU conforms to this
[22:40] <toad_> this means we need to ditch the bidi sort order and instigate something based on LRU
[22:40] <toad_> but first I'm going to make some sims that actually include connections
[22:40] <KenMan> specifically to what do you refer as 'LRU' ?? simply the 'least recently used' connection ?
[22:40] <toad_> KenMan: why would announcements mess that up?
[22:40] <toad_> KenMan: yup
[22:41] <toad_> least recently used, in terms of requests
[22:41] <toad_> this is actually a whole lot simpler with muxing
[22:41] <toad_> because before, we had to do exceptions for live conns as they were transferring data
[22:41] <KenMan> i thought announcements were initiated with a random spec value, one that does not relate to anything but random()...
[22:41] <toad_> but it's simply, when we want to dump a conn or node, dump the one we least recently sent a request to
[22:42] <toad_> KenMan: yes, and?
[22:42] <KenMan> i was listening to thelema explain it, and I was pretty sure he said that.
[22:43] <KenMan> we need nodes to share their best guess. Or something. But having ten new peers pull me in this direction for a little while doesn't seem to help, when 60 others are pulling somewhere else.
[22:43] <toad_> huh?
[22:43] <toad_> ahhh
[22:43] <KenMan> the longer a peer plays the "guess my spec" game, the closer they are likely to come. Agree?
[22:43] <toad_> obviously we have problems during rapid expansion
[22:44] <toad_> KenMan: huh?
[22:44] <toad_> a new node simply doesn't have a spec
[22:44] <toad_> that's why we need announcement
[22:44] <toad_> the algorithm WILL produce one out of thin air eventually
[22:44] <toad_> simulations prove that at least on fully connected networks
[22:44] <KenMan> okay, i am out of my element here :)
[22:44] <toad_> but announcements should accelerate the process
[22:45] <KenMan> and when does a node decide it will no longer announce ?
[22:45] <KenMan> or "be announced to" or however that works ??
[22:45] <toad_> they're only supposed to do it once
[22:45] <toad_> in practice they can get away with it while they're still newbie
[22:45] <toad_> if they're malicious they can connect using a different identity
[22:45] <KenMan> once when they are young, or once when they connect to a new peer ??
[22:45] <toad_> once when they're young, in theory
[22:45] <KenMan> okay
[22:46] <toad_> that's the intention anyway
[22:46] <toad_> hopefully not too far from the reality if you make high HTL announcements work
[22:47] <KenMan> so, if i am a harvester, can I subvert the system by providing estimator data for a 'target' node ??
[22:47] <toad_> hmm?
[22:47] <toad_> how?
[22:47] <KenMan> or is that data almost worthless ? it will change rapidly ?
[22:48] <toad_> the StoreData estimators?
[22:48] <toad_> that's supposed to change reasonably quickly
[22:48] <toad_> we throw it out if the specialization is insane
[22:48] <KenMan> so would timestamping a seed be useful ? or it doesn't matter ?
[22:48] <toad_> how?
[22:48] <KenMan> good point.
[22:49] <KenMan> I am still getting connection attempts after several weeks of no node.
[22:49] <toad_> hehe
[22:50] <KenMan> Where are these references persisting ? Can't we address this ? not just to be nice to me, but to help those poor blind souls ??
[22:50] <toad_> i'm not sure
[22:51] <toad_> they should be replaced..
[22:51] <toad_> given the current sort algo
[22:51] <KenMan> oh, so those nodes who can't manage to get enough good connections keep some bad ones around, with the philosophy that bad ones are better than none ? maybe ?
[22:51] <toad_> possibly
[22:52] <KenMan> I mean, my old ref isn't being moved around the network, right ?? RIGHT ??
[22:52] <KenMan> only living on in stale RT nodes... right ?
[22:53] <KenMan> hell, shouldn't you be getting tired yet ? I didn't realize how late it got...
[22:53] <toad_> yeah, i don't think it's being moved around the network
[22:53] * toad_ is just babysitting the backup... once it starts writing the DVD I'll go to bed
[22:53] <KenMan> how do I open a private channel to toad with XChat ?
[22:53] <toad_> no idea
[22:54] <toad_> i just use /msg, ksirc will open a window
[22:55] <KenMan> it doesn't open a window here. So no privvy conversation... ;)
[22:56] <toad_> you could just /msg me
[22:56] <toad_> /msg toad hi
[22:56] <KenMan> I just did that. Didn't it work ??
[22:56] <toad_> /msg toad_ hi
[22:56] <KenMan> oh...
[22:56] <toad_> get my name right :)
[23:00] <KenMan> another night of very clear stars...
[23:02] <KenMan> time to gaze upon them. I see the Milky Way very nicely. It was very clear at higher latitudes as well.
[23:03] <toad_> where are you?
[23:05] <KenMan> I am near Baltimore, at home.
[23:05] <KenMan> I think light pollution has the biggest impact, not latitude.
[23:06] <KenMan> now that you bring it up, it may be time for me to visit the State of Intoxication again.
[23:07] <toad_> there's a ...?
[23:07] <toad_> or do you just mean to get drunk?
[23:07] <KenMan> to have a few strong beers.
[23:07] <toad_> it's State of Inebriation though isn't it?
[23:07] <KenMan> well, it starts out that way, sure.
[23:07] * toad_ thought maybe you were in one of the drier states
[23:08] * toad_ looks up inebriation...
[23:08] <toad_> 2 meanings, one is alcoholism (as an addiction), the other is just temporary drunkenness
[23:09] <toad_> s/states/States
[23:10] <KenMan> Inebriation, Intoxication, Insomnia, brief sleep, then a pleasant morning, followed by a painfully SLOW afternoon. That's my usual path when drinking.
[23:10] <toad_> not seriously intoxicated then if pleasant morning
[23:11] <KenMan> perhaps as a teenager, i passed out from drinking once or twice. But not now. I might throw up violently if I overdid it nowadays, but never pass out. I have what I call 'reverse hangovers.' They catch up with me late the next day. It is WEIRD.
[23:12] <toad_> heh
[23:12] <KenMan> I always get half my normal sleep, and wake rested. Feel great. Then go downhill many hours later.
[23:13] <KenMan> Oh well, it's not like I drink every day, or even every week. So I suppose some people would accuse me of not being dedicated.
[23:14] <KenMan> I bought a half-gallon of fresh made orange juice today. It smelled so darned good I didn't even look at the price. But it is GOOD!!!
[23:16] <KenMan> so i won't ruin it by mixing it with liquor.
[23:18] * interrupt (~chatzilla@64.122.23.217) has joined #freenet
[23:18] * interrupt (~chatzilla@64.122.23.217) Quit (Client Quit)
[23:25] * pupok (~r00t@81-178-112-130.dsl.pipex.com) has left #freenet
[23:31] <KenMan> i imagine, that long before the invention of telescopes and astronomers, some human with good vision and some cognition recognized the Milky Way for what it is - a huge blob of stars.
[23:32] <KenMan> I mean, there are bright stars and dim ones. So that cloud (on a perfectly clear night) could only be constructed of these things that produce light.
[23:33] <KenMan> Or some god. Or some sort of sky/space creature. But someone must have gotten it right, and well before the dawn of astronomy.
[23:33] * sanity (~ian@81-178-112-130.dsl.pipex.com) Quit ()
[23:34] <KenMan> Otherwise my faith in the mental capacity of humans is greatly eroded.
[23:39] <KenMan> besides, back when global pollution consisted exclusively of feces, the night sky was much clearer.
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.