Timestamps are in GMT/BST.
[1:23] * tavino (~tavino@espians.com) has joined #freenet
[1:24] * tavino is now known as tav
[3:59] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) has joined #freenet
[4:01] * pittaman (~mistery_b@d51A410CA.kabel.telenet.be) Quit (Client Quit)
[4:43] * TLF (~francisco@197.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[5:20] * ShaunMacPherson (~ShaunMac@Sudbury-HSE-ppp3977356.sympatico.ca) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * goatee (~goatee@ip216-239-84-102.vif.net) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * pwk___ (~phweak@dsl-62-3-71-168.zen.co.uk) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * Digger3 (~digger3@kf-dn-tg01-174.dial.kabelfoon.nl) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * MrNaughty (MrNaughty@d199-126-253-66.abhsia.telus.net) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * pupok (~r00t@81-178-94-146.dsl.pipex.com) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * vsalento (~vsalento@cs136227.pp.htv.fi) Quit (burroughs.freenode.net irc.freenode.net)
[5:20] * Overand (common@69.0.81.177.adsl.snet.net) Quit (burroughs.freenode.net irc.freenode.net)
[5:21] * Overand (common@69.0.81.177.adsl.snet.net) has joined #freenet
[5:22] * Digger3 (~digger3@kf-dn-tg01-174.dial.kabelfoon.nl) has joined #freenet
[5:22] * pwk_ (~phweak@dsl-62-3-71-168.zen.co.uk) has joined #freenet
[5:25] * pupok (~r00t@81-178-94-146.dsl.pipex.com) has joined #freenet
[5:26] * goatee (~goatee@ip216-239-84-102.vif.net) has joined #freenet
[5:27] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) has joined #freenet
[5:33] * vsalento (~vsalento@213.243.136.227) has joined #freenet
[5:36] * ShaunMacPherson (~ShaunMac@Sudbury-HSE-ppp3977356.sympatico.ca) has joined #Freenet
[6:16] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[6:38] * Tafs (~tafs@39.80-202-219.nextgentel.com) has joined #freenet
[7:11] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[7:15] * MrNaughty (MrNaughty@d199-126-253-66.abhsia.telus.net) has joined #freenet
[8:12] * d-ArkAngel (~Rob@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) has joined #freenet
[8:12] <d-ArkAngel> morning all
[8:13] <d-ArkAngel> toad : server is back up
[8:13] <d-ArkAngel> had to power it off, seems that it doesn't have suspend support in the BIOS :-(
[8:15] <d-ArkAngel> also I took the opportunity to make it a little quieter, not sure if the increase in heat will have negative impact on stability, but untill it crashes we can't really tell :-)
[8:22] * kers (~kers@7.ppp144.rsd.worldonline.se) has joined #freenet
[8:22] * guido^pe (~unknown@dsl-082-082-157-241.arcor-ip.net) has joined #freenet
[8:24] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) has joined #freenet
[8:46] * TLF (~francisco@197.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[8:53] * hobx_ (~chatzilla@ankh.math.chalmers.se) Quit (Remote closed the connection)
[9:09] * plixed (~salgo@p50823D15.dip.t-dialin.net) has joined #freenet
[9:11] * plixed (~salgo@p50823D15.dip.t-dialin.net) Quit (Client Quit)
[9:19] * plixed (~plixed@p50823D15.dip.t-dialin.net) has joined #freenet
[9:34] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[9:38] * TLF (~francisco@197.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[9:57] * hobx_ (~chatzilla@ankh.math.chalmers.se) has joined #freenet
[10:21] * dli_ (~tst@anna-bhat.uchicago.edu) has joined #freenet
[10:22] <toad_> hi dli_
[10:22] <dli_> toad_: hi
[10:22] <dli_> toad_: my friends all told me that freegate stopped working
[10:23] <toad_> dli_: woah
[10:23] <toad_> that's bad news
[10:23] <dli_> toad_: as expected
[10:23] <toad_> just the website? or the whole service? it's decentralized, right?
[10:24] <dli_> toad_: whole service, not even https proxies
[10:24] <toad_> probably it was a little too centralized...
[10:24] <toad_> they can't be blocking HTTPS, can they? you need it for ebiz?
[10:24] <toad_> brb, need to get some fuel
[10:24] <dli_> toad_: if net police join, then, it fails
[10:27] <toad_> rehi
[10:27] <toad_> dli_: huh?
[10:27] <dli_> toad_: I want to help my friends to get freenet installation files
[10:27] <toad_> dli_: sure
[10:28] <toad_> dli_: there's a link called Spread Freenet
[10:28] <toad_> that produces a ZIP you can send them
[10:28] <dli_> toad_: net police join, and block all IP in freegate
[10:28] <toad_> this is assuming you already have a working node?
[10:28] <toad_> dli_: hmm, scary
[10:28] <toad_> that strongly suggests the seednodes are blocked...
[10:28] <dli_> toad_: I put the zip on my apache server
[10:29] <toad_> and?
[10:29] <dli_> then, my IP seems to be blocked in China
[10:29] <toad_> dli_: you didn't email them?
[10:29] <toad_> was it called freenet.zip?
[10:30] <toad_> is it that simple?
[10:30] <toad_> or did you rename it?
[10:30] <dli_> toad_: I called it goj.zip
[10:30] <toad_> if so, then it probably didn't contain the word freenet in plaintext, only compressed
[10:30] <toad_> which means their scanner checks zips :<
[10:30] <dli_> toad_: passwd protected
[10:30] <toad_> dli_: okay, now you're scaring me
[10:30] <dli_> toad_: you may check, http://riz-sage.uchicago.edu/usr/goj.zip
[10:31] <toad_> i can't imagine them brute forcing password protected zips just to check whether there's illicit content in them
[10:31] <toad_> which leaves two possibilities:
[10:31] <dli_> toad_: just amazing
[10:31] <toad_> 1. they identified you personally as a threat, and blocked you, because they watch this channel
[10:32] <toad_> 2. they block all password protected zipfiles
[10:32] <dli_> toad_: I remember I ran freenet 1 year ago on my box, my IP got blocked in China, the IP, all services
[10:32] <cbreak> still the same IP?
[10:32] <dli_> cbreak: yeah, the same ip
[10:33] <toad_> yeah, 3. they blocked your IP because you run a freenet node, and they block freenet nodes either from seednodes (!!) or by harvesting (!!!!!!)
[10:33] <toad_> or by detecting session bytes (omg!)
[10:34] <toad_> dli_: that's all you did to provoke them?
[10:34] <toad_> dli_: you had people try to connect to your node from china?
[10:35] <toad_> dli_: i discovered the problem with your zip
[10:35] <toad_> it's got the word "freenet" in it
[10:35] <toad_> in fact it's listable
[10:35] <toad_> root@amphibian:~# unzip -l ~browser/goj.zip
[10:35] <toad_> Archive: /home/browser/goj.zip
[10:35] <toad_> Length Date Time Name
[10:35] <toad_> -------- ---- ---- ----
[10:35] <toad_> 0 09-26-04 13:56 tmp/
[10:35] <toad_> 15003167 09-24-04 18:07 tmp/seednodes.ref
[10:35] <toad_> 10293774 09-24-04 18:00 tmp/freenet-java-webinstall.exe
[10:35] <toad_> 9192500 09-24-04 18:07 tmp/seednodes.ref.bak
[10:35] <toad_> 12 09-26-04 12:14 tmp/gong.txt
[10:35] <toad_> -------- -------
[10:35] <toad_> 34489453 5 files
[10:35] <toad_> so just because it's encrypted doesn't mean it's unreadable - at least the directory is readable
[10:36] <toad_> also, that's not enough files to install freenet
[10:36] <toad_> you need a bunch more - especially freenet.jar, freenet-ext.jar, and a few EXEs
[10:37] * toad_ recommends making an encrypted zip file containing the stuff you want, renaming it to album.mp3, then making another encrypted zip file to put it in...
[10:38] * cbreak recomends using gpg for encryption
[10:38] <toad_> well yeah, zip is pretty crappy
[10:39] <dli_> I still don't understand how much computting power they have, to crack all passwd protected zips?
[10:39] <toad_> dli_: they don't have to
[10:39] <dli_> it's not because we discuss here, I did come here when my box got blocked
[10:39] <dli_> it's not filename of course, I made the goj name by a shift to each letter of "Free-Net-Installation", fni -> goj
[10:39] <toad_> dli_: it's the filename
[10:39] <toad_> did you get my paste above?
[10:39] <toad_> it's the filenames INSIDE the zip
[10:39] <toad_> which are plaintext
[10:40] <toad_> because of a design flaw in the zip format
[10:40] <cbreak> dli_: It looks like ZIP Encryption only encrypts data (well... IF it encrypts something), not the catalog.
[10:40] <dli_> cbreak: that makes sense
[10:40] <toad_> <dli_> it's not because we discuss here, I did come here when my box got blocked
[10:40] <toad_> okay
[10:41] <toad_> so what did you run on the box?
[10:41] <dli_> cbreak: I should have used double zipping, then
[10:41] <toad_> just frenet?
[10:41] <toad_> did you run anything else e.g. an HTTP proxy?
[10:41] <dli_> toad_: on the http server, there's no freenet
[10:41] <toad_> also did you cease to be blocked
[10:41] <toad_> ah ok
[10:41] <toad_> two different IPs
[10:41] <toad_> one is the HTTP server
[10:41] <dli_> toad_: yeah, after I stopped freenet, it was back
[10:41] <toad_> which got blocked because it served a file with an illegal keyword in
[10:42] <toad_> that's not surprising really
[10:42] <toad_> the other ran a node - did it run anything else?
[10:42] <toad_> dli_: how much after?
[10:42] <dli_> toad_: yeah, the one with freenet installation on the http server
[10:42] <dli_> toad_: I don't know for sure, but like two months
[10:42] <toad_> dli_: huh?
[10:43] <toad_> dli_: were you running anything else?
[10:43] <toad_> a proxy... or the distribution servlet (the spread freenet link) ?
[10:43] <dli_> toad_: I noticed it was blocked only when my friends asked me to send him some research papers, I sent in email, blocked (no error, no return)
[10:43] <toad_> or did you have 8888 globally readable?
[10:44] <dli_> toad_: no, I have bastille firewall, only 22 readable
[10:44] <toad_> okay so it was definitely blocked due to that, and not due to email, proxy, whatever?
[10:45] <toad_> statistically speaking, it seems likely that your node would get into the seednodes at some point during that time...
[10:45] <dli_> toad_: I didn't check carefully, but after I stopped freenet, it readible in china later
[10:45] <dli_> toad_: but not this time, it's not running freenet for long
[10:46] <toad_> dli_: you're running a node, and it hasn't been blocked yet?
[10:46] <toad_> how long has it been running for?
[10:46] <dli_> toad_: I will try to ask my friends to download the installation file by scp
[10:47] <dli_> toad_: I run it on my laptop, my laptop doesn't provide service except ssh
[10:47] <toad_> dli_: how long has your node been running?
[10:47] <toad_> and has it connected to any chinese addresses?
[10:47] <dli_> toad_: not running continuously, I only test it here
[10:47] <toad_> ah ok
[10:47] <toad_> that's why it hasn't been blocked
[10:48] <toad_> even if they harvest it probably won't show up
[10:48] <dli_> toad_: no, should not be, I don't visited websites in China often
[10:48] <toad_> yeah i mean the node
[10:48] <toad_> i suppose you don't know?
[10:48] <dli_> toad_: no, I don't remember surfing websites in China for long
[10:49] <toad_> uhh
[10:49] <toad_> i mean the freenet node
[10:49] <toad_> but if it's not constantly up, probably not
[10:49] <d-ArkAngel> toad, my server is up if you need so sim anything
[10:49] <toad_> what i'm getting at is:
[10:49] <toad_> there are a few different possibilities
[10:49] <toad_> one is that they just download the seednodes and block them
[10:49] <toad_> another is that they detect the FNP protocol bytes, and then block stuff that uses them
[10:49] <dli_> toad_: I want to ask my friends to use my seednodes
[10:50] <dli_> toad_: if they block FNP, then, there's no hope
[10:50] <dli_> toad_: I suppose it's unlikely :(
[10:50] <toad_> dli_: if they block the seednodes then there's some serious ]work to be done
[10:50] <toad_> if they block FNP then it'll have to wait for sessionv2 :|
[10:51] <cbreak> FNP (Freenet Node Port) is different for each node, and FNP (Freenet Node Protocoll) is encrypted, mostly...
[10:51] <toad_> cbreak: it's the "mostly" that's the problem
[10:51] <toad_> like most protocols it has a few fixed bytes
[10:51] <toad_> because oskar is an idiot ;)
[10:52] <cbreak> :)
[10:52] * cbreak has a Network Systems Exam on Monday...
[10:53] <dli_> toad_: then can definitely detect freenet is running on certain nodes
[10:53] <toad_> dli_: there were some tests done by dialing up on a chinese ISP via international telephone call...
[10:53] <toad_> dli_: hmm?
[10:53] <toad_> can they, or do they just read the seednodes list?
[10:54] * TLF (~francisco@197.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[10:54] <dli_> toad_: let me update the installation files, and instruct a friend to use scp to download
[10:54] <toad_> please
[10:54] <dli_> toad_: I suppose no need to hide the files in scp
[10:54] <toad_> hmm?
[10:54] <toad_> no
[10:54] <dli_> toad_: okay
[10:55] <toad_> dli_: I wonder if we should do some dial-up tests?
[10:55] <toad_> have they put any sort of countermeasures in place to prevent that?
[10:55] <dli_> toad_: later
[10:56] <dli_> toad_: it's easy for them to run a node there, and block all nodes it can talk to
[10:56] <toad_> dli_: that's called "harvesting"
[10:56] <toad_> unfortunately there is very little we can do about it
[10:57] <dli_> toad_: oh, I didn't relealize it's that simple
[10:57] <toad_> :<
[10:58] <dli_> toad_: then, the only hope is that freenet become popular, so popular that they just couldn't afford blocking even a fraction of running nodes
[10:58] <toad_> unfortunately recent results from some of ian/oskar's more theoretical simulation work suggest there will never be anything we can do about it
[10:58] <toad_> I used to think we could fix it
[10:58] <toad_> but now I'm not so sure
[10:59] <toad_> dli_: it's a pretty dubious hope, unfortunately.. with highly dynamic connections, they'll just wait till you connect to a node they know about
[10:59] <toad_> and then knock you out
[10:59] <toad_> still there MAY be things we can do
[11:01] <d-ArkAngel> it will as always be a battle of wits, of us updateing things faster than they can keep up. we can make it hard for them to follow, but anyone who can doanload the client should be fairly able to work out ways to detect nodes, assuming they've got infinate patients and resources
[11:02] <dli_> d-ArkAngel, no, we are losing, they got more people and resource
[11:02] <d-ArkAngel> yup, that is to be expected.
[11:02] <toad_> it depends how much they want to spend
[11:02] * toad_ eek http://www.cnn.com/2004/US/10/01/copying.dollars.ap/index.html
[11:03] <dli_> d-ArkAngel, easy, remember the reccent google story? they just force google to do so, they spend like $0, google does it for them
[11:04] <dli_> toad_: calling one friend, hopefully, there would be a test
[11:04] <d-ArkAngel> google is a company, we're not. they can't "convince" us to do it for them ;-)
[11:06] <toad_> heh
[11:06] <dli_> d-ArkAngel, yeah, just one example, how much resource they got
[11:06] <toad_> do it for us! or we'll buy you up!
[11:06] <toad_> the bottom line with google is that they'd simply have blocked google
[11:06] <toad_> in fact they did for a while
[11:06] <toad_> if it didn't cooperate with their censorship requirements
[11:07] <toad_> google can't afford that loss because it's advertising revenues from china are rapidly rising
[11:07] <d-ArkAngel> thing is, if they really wanted to they could just block everything.
[11:07] <toad_> no, they couldn't
[11:07] <toad_> they couldn't afford to
[11:08] <d-ArkAngel> I mean then allow sites.
[11:08] <toad_> possible, but VERY expensive
[11:09] <d-ArkAngel> more expensive than fighting us? (in the long term)
[11:09] <toad_> yes, probably
[11:09] <toad_> depending on what rabbits we pull out of our collective hats
[11:09] <d-ArkAngel> and we're probably not the only group they have to contend with.
[11:10] <toad_> right now they can just block the seednodes. if we make that harder, they can harvest and block. if they care enough they can block the session bytes. if we fix that they're back to harvesting - and identifying nodes, then blocking anyone who connects to them
[11:10] <toad_> d-ArkAngel: no, of course
[11:10] <toad_> but they just blocked FreeGate
[11:10] <toad_> (somehow)
[11:10] <toad_> we don't know how, since it was closed source, allegedly for security reasons ;)
[11:11] <d-ArkAngel> any idea what IPv6 adds to the party?
[11:11] <toad_> it'd be funny if it wasn't so sad
[11:12] <d-ArkAngel> well they're just takeing a more active role in censorship than other govenments
[11:13] <d-ArkAngel> imageine what RIAA would be like if it had the powers that they have!
[11:13] <toad_> this is all predicated on freenet being a threat anyway - and if freenet can't scale, it won't be much of a threat
[11:13] <toad_> d-ArkAngel: watch and wait!
[11:13] <d-ArkAngel> :-)
[11:13] <dli_> instructing my friends to download psftp.exe
[11:14] <toad_> dli_: good idea
[11:14] <toad_> you know it's command line?
[11:14] <toad_> how techie are they?
[11:14] <dli_> toad_: I think it's like ftp
[11:15] <toad_> yes
[11:15] <toad_> but it's like command-line ftp
[11:15] <toad_> not like ws_ftp
[11:15] <dli_> toad_: not at all, windows guy
[11:16] <toad_> dli_: what does the zip include?
[11:16] <toad_> the one on that server didn't have everything needed
[11:16] <toad_> can you send me it
[11:16] <toad_> i'll check it over?
[11:16] <dli_> toad_: freenet-java-webinstall.exe
[11:16] <toad_> dli_: AND?
[11:16] <dli_> seednodes.ref
[11:16] <toad_> and?
[11:17] <toad_> you need at an absolute minimum for windows, nodeconfig.exe, freenet.exe, the other exe, the installer exe, freenet.jar, freenet-ext.jar, and seednodes
[11:17] <toad_> this will be packed up for you by the Spread Freenet link
[11:17] <toad_> unless there's a bug
[11:17] <toad_> it will include locally generated seednodes
[11:18] <dli_> toad_: give me a link to get those files?
[11:18] <toad_> dli_: http://127.0.0.1:8888/, click on Spread Freenet, click New Distribution Page, click link, download zip
[11:20] * toad_ graphing new simulations, aiming for 95% at 2.5% fetchfrac..
[11:20] <toad_> dead on half way between htl=11 and htl=12 ok
[11:20] <toad_> need to run them for a bit longer though
[11:21] <dli_> toad_: but my node is linux
[11:21] <toad_> dli_: doesn't matter
[11:21] <toad_> it should automatically fetch those files when you start it
[11:21] <toad_> for both platforms
[11:21] <toad_> the zip works on either
[11:21] <dli_> toad_: missing file :( freenet.exe
[11:21] <toad_> don't whatever you do send them freenet-webinstall.exe
[11:22] <toad_> dli_: get it from http://freenetproject.org/snapshots/
[11:22] <toad_> the webinstaller connects directly to freenetproject.org
[11:22] <dli_> toad_: save in the freenet dir?
[11:22] <toad_> save in the dir with the other files
[11:22] <toad_> probably called distrib
[11:22] <toad_> inside the freenet dir
[11:22] <toad_> then try again
[11:25] * guido^pe (~unknown@dsl-082-082-157-241.arcor-ip.net) Quit (Remote closed the connection)
[11:26] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-100x25.png
[11:26] <toad_> hmm
[11:26] <toad_> down to 0.95, they go mostly down then up
[11:26] <toad_> below they seem to go up then down
[11:26] <toad_> interesting...
[11:27] <cbreak> thought about using a running, decaying average?
[11:27] * toad_ uploads new jar to scabserver
[11:27] <toad_> cbreak: hmm?
[11:27] <toad_> ah
[11:27] <toad_> yeah, might be an idea
[11:29] <dli_> toad_: then, I download the zip file from freenet
[11:29] <dli_> downloading the freenet.zip file, slow :)
[11:29] <cbreak> AvgSuccess = (1-a)AvgSuccess + a*NewSuccess (Well... stolen from TCP :)
[11:30] <toad_> dli_: it shouldn't be... ah
[11:30] <toad_> i suppose it's limited as if it was being served to the outside world
[11:30] <dli_> my connection got some problems
[11:30] <toad_> because usually it is...
[11:30] <dli_> toad_: network cable problem, looks like
[11:30] <toad_> dli_: ahhh
[11:31] <dli_> toad_: just 4 MB?
[11:31] <toad_> sounds reasonable
[11:32] <toad_> although check the size of the seednodes file...
[11:32] <cbreak> the unstable one is 3.9 MB
[11:32] <toad_> yeah i mean the generated one
[11:32] <toad_> in the zip
[11:32] <toad_> dli_: what's the size of the seednodes file in the zip?
[11:34] <dli_> 24113 10-02-04 09:32 freenet-distribution/seednodes.ref
[11:34] <cbreak> Interested in an java.io.EOFException in freenet.presentation.FCPRawMessage.<init> ?
[11:34] <toad_> dli_: that looks rather small
[11:35] <toad_> how many nodes are in it?
[11:35] <toad_> extract it, count the number of physical.tcp= lines
[11:35] <toad_> cbreak: it's probably like all the other ones
[11:35] <toad_> but dcc me it
[11:35] <toad_> what were you running at the time?
[11:36] <dli_> toad_: you want me to count the file?
[11:36] <cbreak> client.cli.Main splitfile download
[11:36] <toad_> dli_: cat seednodes.ref | grep version= | wc -l
[11:36] * lolo-laptop (~lostlogic@node-4024215a.mdw.onnet.us.uu.net) Quit (K-lined)
[11:37] <toad_> k-lined? i doubt it...
[11:37] <dli_> toad_ 21
[11:37] <toad_> dli_: hmmm
[11:37] <toad_> mail me it
[11:37] <toad_> is that the compressed size
[11:37] <toad_> or the raw size?
[11:38] <dli_> toad_: raw size
[11:38] <toad_> dcc not working
[11:38] <cbreak> hmm...
[11:39] <toad_> it just seems unlikely that 21 nodes would fit in 21KB
[11:39] <toad_> unless they were all sans estimators, and i don't see why that would be
[11:42] <toad_> 50x25 simulations run FAST on dual opterons :)
[11:43] <cbreak> you have multi threading in the simulation?
[11:43] <toad_> dli_: hmm
[11:43] <toad_> looks credible
[11:44] <toad_> recommend you set up a fresh local node to test that set of seednodes
[11:44] <toad_> cbreak: no
[11:44] <toad_> just running 2 simulations
[11:44] <toad_> cbreak: d-ArkAngel was looking into making it parallel though
[11:44] <toad_> would be useful for the bigger ones - 400+ nodes
[11:46] <cbreak> if you simulate two requests at a time, I wonder if that could cause the simulation to be non deterministic.
[11:46] <toad_> cbreak: yes, it would
[11:46] <toad_> as would the considerable amount of randomness we already use
[11:46] <cbreak> hmm... yes, you are right :)
[11:47] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-50x25.png
[11:49] * toad_ looking for 95%, so will simulate 8, 7 as well
[11:50] <toad_> want to get approximately the same performance at each HTL so I can show whether it scales
[11:53] <toad_> i mean at each network size
[11:56] <toad_> brb
[11:56] <dli_> toad_ I could not find a compatible Java Runtime Environment installed on this machine
[11:57] * toad_ back
[11:57] <toad_> dli_: eh?
[11:57] <toad_> who said that?
[11:58] <dli_> I thought the .zip contains everything
[11:58] <dli_> toad_: my friend couldn't install it :(
[11:58] <toad_> you trying it on a windows machine?
[11:58] <toad_> ah ok
[11:58] <dli_> toad_: yeah, mostly windows in China
[11:58] <toad_> so either a) he downloads the JRE (java.sun.com isn't blocked, is it?) or b) you send him the JRE (technically in violation of copyright)
[11:58] <cbreak> the zip does obviously not contain a Java Runtime-
[11:59] <toad_> or c) you send him the wininstaller-with-jre
[11:59] <toad_> yeah, we're not legally allowed to distribute the JRE except as a bundle
[11:59] <dli_> toad_: that's much better :)
[11:59] <dli_> toad_: I see
[11:59] <dli_> toad_: let me find the jre download page
[12:00] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-50x25.png
[12:00] <toad_> woah
[12:00] <toad_> htl 7 is a major jump below hto 8
[12:00] <toad_> htl 8
[12:01] <dli_> toad_: which one is the smallest suite to download for java? saw 40 MB :(
[12:02] <toad_> dli_: java runtime environment
[12:02] <toad_> not java development kit
[12:02] <toad_> JRE shouldn't be more than maybe 10MB
[12:03] <dli_> toad_: great
[12:03] <toad_> good or bad?
[12:03] <toad_> you could just ask him to download it, i doubt it's blocked - it's an essential component of a lot of corporate web services stuff
[12:07] <dli_> toad_: I couldn't find it :(
[12:07] <toad_> dli_: eh?
[12:07] <dli_> toad_: I got J2SE 1.4.2_04, but it's 90 MB
[12:07] <toad_> hmmm
[12:07] <toad_> click downloads
[12:07] <toad_> then j2se sdk
[12:08] <toad_> accept the license agreement
[12:08] <toad_> Windows Offline Installation, Multi-language (j2sdk-1_4_2_05-windows-i586-p.exe, 51.09 MB) Download Now!
[12:08] <toad_> ewwww
[12:08] <toad_> i'm sure i've seen it as 10MB...
[12:08] <toad_> what's his bandwidth?
[12:09] <dli_> toad_: dsl, 1.5Mbs, I think, Java(TM) 2 Runtime Environment, Standard Edition 1.4.2_05
[12:09] <toad_> all sorts of stuff needs java
[12:09] <toad_> just guide him to that page and get him to download it
[12:10] <dli_> toad_: it's better to switch to a free java, so a single installation file is possible :(
[12:10] <toad_> dli_: it is
[12:10] <toad_> we have one somewhere
[12:10] <toad_> but it doesn't include the other files iirc
[12:10] <toad_> there are a million reasons why free java is better
[12:11] <toad_> http://freenetproject.org/snapshots/freenet-java-webinstall.exe
[12:11] <toad_> get that
[12:11] <toad_> bundle it with the other files in the zip
[12:11] <toad_> then it's only 14MB or so for the whole package
[12:11] <dli_> toad_: okay:)
[12:12] <dli_> toad_: I think that's the file I included in my zip
[12:12] <toad_> it's 10MB
[12:12] <toad_> you included freenet-webinstall.exe, which is 50kB, right?
[12:12] <toad_> well
[12:12] <toad_> the point is you need ALL the files
[12:13] <toad_> and yes, this should be easier
[12:13] <dli_> toad_: I mean I previously just downloaded that java-webinstall.exe
[12:14] <cbreak> freenet-webinstall.exe != freenet-java-webinstall.exe I think
[12:14] <toad_> dli_: yes but your earlier zip didn't include freenet.jar :)
[12:14] <toad_> hmmm
[12:14] <dli_> toad_: okay
[12:14] * Ash-Fox (Hal-9000@aau121.neoplus.adsl.tpnet.pl) Quit (Nick collision from services.)
[12:15] * Ash-Fox (Hal-9000@aas13.neoplus.adsl.tpnet.pl) has joined #FreeNET
[12:18] * toad_ posts to the list re including that file in the zip by default
[12:18] <toad_> a good idea imho
[12:18] <cbreak> the whole java?
[12:18] <toad_> need to check whether there are legal issues
[12:19] <toad_> cbreak: yes
[12:19] <toad_> it would be included in the Distribution Servlet ZIP
[12:19] <toad_> which would be 14MB instead of 4MB
[12:19] <cbreak> My personal opinion oposes including to much platform dependant code in a multi platform distribution.
[12:19] <toad_> which means practically all nodes, windows or linux, will have a copy in distrib/
[12:20] <toad_> cbreak: my personal opinion is that we should make it easy for freenet to spread to people who can't get to the main web site
[12:20] <toad_> and make it easy for it to spread peer to peer amongst friends, forming a less distorted topology than if everyone connects to the seednodes
[12:20] <toad_> ideally we'd make an EXE of all the file
[12:20] <toad_> s
[12:20] <cbreak> Is java that rare on Windows PCs?
[12:20] <toad_> but that's pretty hard and would involve way too much code
[12:20] <toad_> cbreak: yes
[12:21] <cbreak> Well, wouldn't be that harmfull for me :)
[12:21] <toad_> it'd just be an extra 10MB disk space usage by default
[12:21] <toad_> that's about all the impact it would have
[12:22] <toad_> assuming there are no legal issues
[12:22] <toad_> with redistributing the java+freenet EXE
[12:22] <cbreak> ok. Would the distribution servlet include the file if I downloaded it now?
[12:22] <toad_> also we might want to update it
[12:22] <toad_> cbreak: hmm?
[12:22] <toad_> cbreak: right now the dist servlet includes everything for linux or windows, EXCEPT the JRE
[12:22] <toad_> hopefully the version of java isn't too terrible...
[12:30] <toad_> brb
[12:32] <cbreak> I am reading the Sun Licence. It appears they allow redistribution, under certain circumstances. (See B. in Agreement)
[12:34] <cbreak> The Ofline Installer for 1.5.0 is 14.29 MB...
[12:34] <toad_> cbreak: redistribution of what?
[12:34] <toad_> cbreak: so why's the offline installer for 1.4.2 50MB?
[12:34] <cbreak> Java(TM) 2 Runtime Environment Standard Edition 5.0
[12:35] <cbreak> Don't know.
[12:35] <cbreak> The online installer is only 221 KB :)
[12:35] <toad_> heh
[12:35] <toad_> hmmm http://amphibian.dyndns.org/simulations2-fetchfrac0.025-200x25.png
[12:36] <toad_> htl 16 is about 93%
[12:36] <cbreak> http://java.sun.com/j2se/1.5.0/jre-1_5_0-license.txt
[12:37] <cbreak> and there is also text concerning Redistribution in http://java.sun.com/j2se/1.5.0/jre/README
[12:37] <toad_> okay
[12:37] <toad_> that's the bundling clause
[12:37] <toad_> that's good
[12:38] <toad_> it means we CAN redistribute the bundle EXE
[12:40] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-200x25.png
[12:40] <toad_> hmmm
[12:41] <toad_> htl=17 should settle around 94.5%
[12:41] <d-ArkAngel> indeed
[12:41] <toad_> 8 12.5 17
[12:41] <toad_> good
[12:42] <toad_> 8 for 50x25, 12.5 for 100x25, 17 for 200x25
[12:42] <toad_> that's good, a nice logarithmic progression.. we hope...
[12:43] <d-ArkAngel> yeah it's certainly looking like HTL brings logarythmic improvments.
[12:43] <toad_> d-ArkAngel: 3 points is hardly enough to make that deduction though...
[12:43] <d-ArkAngel> if that green line was only in a slightly better possition
[12:43] <d-ArkAngel> I mean on that graph you have :-)
[12:43] <toad_> green on 200x25?
[12:43] <toad_> ignore green
[12:43] <toad_> brown is where it's at
[12:44] <d-ArkAngel> lol
[12:44] <toad_> 8 + 4.5 = 12.5. 12.5+4.5 = 17 :)
[12:44] <toad_> d-ArkAngel: lol? you think i mean politically? ;)
[12:45] <toad_> updated http://amphibian.dyndns.org/simulations2-fetchfrac0.025-50x25.png
[12:45] <toad_> htl=8 is about 94%
[12:45] <toad_> nothing else is close
[12:45] <d-ArkAngel> indeed. if that green line was a little higher it'd be a much nicer set of lines
[12:46] <toad_> d-ArkAngel: it might be, but then it wouldn't be log
[12:48] <d-ArkAngel> but the HTL should be decreasing returns, and the geturn from 15->16 is much less than 16->17
[12:50] <toad_> hmmm http://amphibian.dyndns.org/simulations2-fetchfrac0.025-200x25.png
[12:50] <toad_> looks like brown IS too low - it'll be closer to 94% than 95%
[12:50] <toad_> but still, brown is closer than blue
[12:50] <toad_> to the target 95%
[12:50] <toad_> brb
[12:51] <d-ArkAngel> I wonder if we could sim 16.5 HTL :-)
[12:52] <cbreak> drop with probability 0.5 at htl 16
[12:56] <toad_> hmm?
[12:56] <toad_> hmm, the commies have got dli_
[12:57] <toad_> :)
[12:59] <d-ArkAngel> how would know such a thing? Are you really a spy in out midts?
[13:00] <d-ArkAngel> :-)
[13:00] <cbreak> just use /whois to find someones ip, use your favorite backdoor to install a troyan, and use that to access the local webcam to watch from a save distance what's happening.
[13:00] <d-ArkAngel> lol
[13:02] <toad_> lol
[13:02] * toad_ doesn't have a webcam
[13:02] <toad_> like most sane people :)
[13:03] * toad_ is slightly annoyed we don't have film from the Genoa bust though
[13:03] * d-ArkAngel is glad not to be sane ;-) www.scabserver.com/webcam.php
[13:03] <toad_> the indymedia center on site got raided
[13:03] <toad_> but we just had a gap in the radio
[13:04] <d-ArkAngel> it would be kinda fun to see a dozen people in black sunglasses bust in :-)
[13:04] <toad_> d-ArkAngel: you should have a freecam on a 10 minute DBR
[13:04] <toad_> when we get inserts working
[13:04] <d-ArkAngel> :-) I will do.
[13:05] <d-ArkAngel> but before I start playing with things I'll contribute first :-)
[13:06] <cbreak> http://douglas.min.net/mirrors/altern.org/bo2kfun/best.html
[13:07] <toad_> now
[13:07] <toad_> i expect 400 will use htl of 21.5
[13:07] <toad_> well i project
[13:08] <toad_> i expect it will use substantially more :)
[13:08] <toad_> but i project it will use 21.5
[13:09] * kers (~kers@7.ppp144.rsd.worldonline.se) Quit (Read error: 113 (No route to host))
[13:11] * toad_ starts 2 threads with 400 nodes each on bigserver
[13:12] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-100x25.png
[13:12] <toad_> hmmm
[13:13] <toad_> ooh
[13:13] <toad_> it looks like the 400 node simulator is working...
[13:13] <toad_> 21 htl: Overall success ratio: 0.9139294710327456
[13:13] <toad_> 22 htl: Overall success ratio: 0.9616
[13:13] <toad_> global being 0.9139294710327456 and 0.9655837563451777 resp
[13:13] <cbreak> so its more or less log
[13:14] <toad_> at least up to 400
[13:14] <cbreak> so next 800, 1600 and so forth :)
[13:14] <toad_> well probably up to 400
[13:14] <toad_> and this is with the same number of conns too!
[13:14] <toad_> unfortunately the log base is not linear with the number of conns
[13:14] <cbreak> more connections make routing better?
[13:14] <toad_> at least it doesn't look that way
[13:15] <toad_> i think so but they also make it slower
[13:15] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-400x25.png
[13:15] <toad_> hmmm
[13:15] <toad_> nothing to see here folks, move along.. :)
[13:15] <toad_> graph started too early
[13:15] <toad_> the next one will be good
[13:16] <toad_> if we can prove that it is in fact log, then that would be a major breakthrough
[13:16] <toad_> the next step would be to play with the # conns a bit
[13:16] <cbreak> prove in the physician or mathematician sense? :)
[13:16] <toad_> then add some more realism - specifically, load
[13:17] <toad_> cbreak: heh
[13:17] <toad_> prove in the empirical sense of "it looks log, my experiments so far don't refute it's log-ness"
[13:17] <cbreak> prove by deduction, one of my favorites :)
[13:18] * Hory (~Miranda@81.196.25.110) Quit (Read error: 54 (Connection reset by peer))
[13:18] <toad_> hmmm
[13:18] <toad_> maybe it isn't log...
[13:18] <toad_> 22 htl
[13:18] <toad_> Overall success ratio: 0.8758563074352548 (global)
[13:18] <toad_> 21 htl
[13:18] <toad_> Overall success ratio: 0.86640838650866 (global)
[13:19] <dli_> toad_: no connection, open connections (0/0/200)
[13:20] <toad_> dli_: hmmm
[13:20] <toad_> interesting
[13:20] <toad_> did you test it locally?
[13:20] <toad_> with the same seednodes?
[13:20] <toad_> and what's the uptime?
[13:20] <dli_> toad_: wait
[13:20] <dli_> toad_: I have connections locally
[13:20] <toad_> dli_: that's with a brand-new node with that seednodes?
[13:21] <dli_> toad_: just with the freenet.zip I made here
[13:21] <toad_> dli_: i mean the local node?
[13:21] <dli_> toad_: okay, let me test it here
[13:22] <toad_> ok
[13:22] <toad_> keep it running for a few hours on the other end
[13:22] <toad_> if possibe
[13:22] <toad_> le
[13:22] <toad_> just in case it's just that it has relatively few nodes and they're all busy or something
[13:23] <toad_> IIRC, the old simulations only got log scaling on HTL, if the # conns was also allowed to scale
[13:23] <dli_> toad_: I got (0/0/200) locally
[13:24] <toad_> dli_: hmmm
[13:24] <toad_> might just be bad seednodes
[13:24] <dli_> toad_: okay:)
[13:24] <toad_> dli_: how many conn attempts?
[13:24] <toad_> how big is the RT?
[13:25] <dli_> toad_: the seednode doesn't work here :(
[13:25] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-400x25.png
[13:25] <toad_> hmmmm
[13:25] <toad_> dli_: go to the Routing Table page
[13:25] <toad_> dli_: under advanced
[13:26] <toad_> how many nodes in RT, how many attempted conns, etc?
[13:26] <toad_> paste the top few lines
[13:26] <toad_> i could give you a
[13:26] <toad_> n/m
[13:26] <dli_> toad_: just make a working seednodes :(
[13:27] <cbreak> If you have a working node, you could get one from there. If you haven't, get them from freenetproject.org/snapshots
[13:28] <dli_> cbreak: it was working here :(
[13:28] <toad_> dli_: how?
[13:28] <toad_> dli_: you could get them to try the central seednodes
[13:28] <toad_> if that doesn't work, then it's probably blocked
[13:29] <toad_> I could make one locally but a) my node isn't running atm
[13:29] <toad_> because of simulations
[13:29] <toad_> and b) it's probably way too close to the official seednodes
[13:29] <toad_> dli_: yeah, try the central seednodes
[13:29] <dli_> toad_: okay :)
[13:29] <toad_> the individual nodes in the central seednodes could well be blocked
[13:29] <toad_> recommend you start a local node with the central seednodes as well
[13:30] <toad_> instead of the one with your local seednodes
[13:30] <cbreak> I could give you unstable seeds, but you should rather use stable...
[13:30] <toad_> dli_: or download from the Routing Table page
[13:31] <toad_> that's probably different to the one included in the distro ZIP
[13:31] <toad_> due to some sort of rare bug (because nobody uses the distro servlet ;( )
[13:31] <toad_> i will fix soon
[13:31] <toad_> anyway, check whether it is different
[13:31] <toad_> but it's possible that your seednodes suck because your node sucks :)
[13:31] <dli_> toad_: okay:)
[13:32] <toad_> the central seednodes SHOULD work - there are tons of them at least - BUT they may be blocked
[13:32] <toad_> finding out whether they are blocked would be nice
[13:32] <dli_> toad_: where to get the central seednodes
[13:33] <toad_> dli_: freenetproject.org/snapshots/seednodes.ref.zip
[13:33] <cbreak> freenetproject.org/snapshots/
[13:33] <toad_> i think there's a .zip...
[13:33] <toad_> there isn't
[13:33] <toad_> hmmm
[13:33] <dli_> toad_: it's a bz2, okay for me
[13:33] <toad_> you'll need to re-zip it for transport
[13:34] * guido^pe (~unknown@dsl-082-082-157-241.arcor-ip.net) has joined #freenet
[13:35] <dli_> toad_: okay:)
[13:36] * moskawa23 (~Miranda@dsl-082-082-236-161.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
[13:39] * toad_ simulating 400x25@28 and 400x50@22
[13:42] <dli_> toad_: seednodes.ref is under \Program Files\Freenet right?
[13:43] <toad_> dli_: when installed on windows? probably
[13:43] <toad_> if you just want to replace it, then shut down the node first
[13:43] <toad_> THEN copy over it
[13:43] <dli_> toad_: how to shut it down? in windows :(
[13:44] <toad_> on the rabbit icon i think
[13:44] <toad_> you can tell it to quit
[13:44] <dli_> toad_: okay:)
[13:44] <toad_> did that work?
[13:46] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-400x25.png
[13:46] <dli_> still downloading
[13:46] <toad_> hrrrrm
[13:46] <toad_> that's not too encouraging
[13:46] <toad_> raising the rt size to 50 doesn't make very much difference
[13:46] <toad_> no more than the diff between 21 and 22 htl...
[13:46] <toad_> apparently
[13:46] <dli_> toad_: got to make it simple, most people in China are just windows dummies
[13:47] <toad_> raising htl to 28 doesn't help much either...
[13:47] <toad_> dli_: like most people everywhere, sadly. There may be bugs in the dist servlet seednodes generator.
[13:47] <toad_> it would definitely be a good idea to include a JRE
[13:47] <toad_> so these are ways we can improve from today's problems
[13:48] <dli_> toad_: yeah, maybe, a way to update seednodes (hard) :(
[13:48] <toad_> Overall success ratio: 0.9238716148445336
[13:48] <toad_> that's with htl 28...
[13:48] <toad_> dli_: huh?
[13:49] <dli_> toad_: now, there are three steps :( very unlikely for a general user to fight through
[13:51] <cbreak> does the sim scale better with larger stores, but same store_frac?
[13:53] <toad_> why would it?
[13:53] <toad_> dli_: well the seednodes.ref produced by the dist servlet SHOULD be sufficient
[13:54] <toad_> if it's not, either your node wasn't up for long, or wasn't very good, or there's a bug
[13:56] <dli_> toad_: my node was up, when I made the zip
[13:56] <dli_> toad_: where to specify the NAT gateway IP?
[13:57] <toad_> dli_: sure, but how many connections did it have? and how long had it been up? if you start the node and expect it to produce good seednodes 30 seconds after startup, then you're in for a nasty surprise; it won't
[13:57] <toad_> dli_: huh?
[13:57] <cbreak> on %ipAddress= in freenet.ini
[13:57] <toad_> dli_: why do you need to?
[13:57] <cbreak> just remove the % and enter the external IP of your node
[13:58] <toad_> there's no need to specify the ip address
[13:58] <toad_> what IS critical is forwarding the port
[13:58] <dli_> toad_: great
[13:58] <toad_> if they're on a university LAN where you can't forward ports, then their node will probably look a lot like newsbyte's
[13:58] <toad_> it might work - with a lot of luck
[14:02] * Tafs (~tafs@39.80-202-219.nextgentel.com) Quit (Read error: 104 (Connection reset by peer))
[14:02] <dli_> toad_: have difficulty in stopping freenet :(
[14:02] <toad_> dli_: strange
[14:03] <dli_> I don't understand why :(
[14:03] <toad_> what about the rabbit icon?
[14:03] <dli_> right click doesn't work :(
[14:03] <toad_> left click?
[14:03] <cbreak> you could try to kill the java process in the process manager (I think I saw something like that on Win)
[14:03] <toad_> yeah, ctrl+alt+del and find the process called java.exe
[14:03] <toad_> but the rabbit should work
[14:05] * guido^pe (~unknown@dsl-082-082-157-241.arcor-ip.net) Quit (Read error: 113 (No route to host))
[14:06] <dli_> strange enough for me
[14:06] <dli_> anyway, too late in China now
[14:06] <dli_> my friend will try tomorrow
[14:06] <toad_> okay
[14:06] <toad_> i won't be here tomorrow
[14:06] <toad_> i will be here on monday
[14:06] <toad_> please tell me if it works
[14:07] * guido^pe (~unknown@dsl-213-023-244-238.arcor-ip.net) has joined #freenet
[14:07] <toad_> hmmm
[14:08] <dli_> toad_: sure
[14:08] <toad_> 400x50 is actually WORSE than 400x25 on the same HTL
[14:08] <toad_> according to this graph...
[14:08] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-400x25.png
[14:08] <toad_> and 28 is still not enough
[14:08] <dli_> toad_: I didn't realize we can use taskman in windows to kill it
[14:09] <dli_> toad_: some like "stop-freenet" in windows doesn't work
[14:09] <toad_> dli_: the rabbit doesn't work then?
[14:10] <dli_> toad_: my friend said no menu from it at all :(
[14:10] <toad_> hmmm
[14:11] <dli_> toad_: can I replace the freenet-webinstall.exe with freenet-java-webinstall.exe in the freenet.zip here?
[14:11] <toad_> i don't see why not
[14:12] <dli_> toad_: okay, it would save the java trouble, I will ask more friends to try
[14:18] * toad_ hmmm
[14:19] <toad_> running another 400x25... this time @35
[14:19] <toad_> 35 ought to be sufficient...
[14:32] <toad_> bbl
[14:39] <Ash-Fox> Blah... I need to stop watching old Doctor Who... I had a nightmare last night of being chased by floating tincans with plungers
[14:39] <Ash-Fox> Shouting "exterminate"
[14:40] <cbreak> hmm... better than my nightmares...
[14:40] <Ash-Fox> this would be one of my friend's nightmares: http://www.deviantart.com/deviation/11067812/
[14:41] <Ash-Fox> Don't know why we're always having dreams about being chased
[14:43] <cbreak> there are tonns of books about dream interpretation... :)
[14:43] <Ash-Fox> I don't want my dream interpreted
[15:21] <toad_> :)
[15:37] * TLF (~francisco@209.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[15:44] * toad_ back
[15:44] <toad_> hi ppl
[15:47] <toad_> http://amphibian.dyndns.org/simulations2-fetchfrac0.025-400x25.png
[15:48] <toad_> 94% at HTL 35, with 400x50
[15:49] <toad_> so it's 8, 12, 17, 35
[15:49] <toad_> hrrm
[15:50] <toad_> for 50, 100, 200, 400 nodes resp.
[15:50] <toad_> and it may be even more than 35 actually
[15:50] <toad_> to get 95%, or to do it without x50
[15:51] <toad_> [EXPLETIVE DELETED]
[15:51] <cbreak> hmm... and 300x25 :)
[15:52] <toad_> should probably try 400x100...
[15:52] <toad_> and a few others
[15:52] <toad_> cbreak: huh?
[15:55] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[15:55] <cbreak> somewhere between 200 and 400 nodes the required HTL stops to be more or less linear. More samples may be helpfull for an interpolation.
[15:55] <toad_> hmm?
[15:55] <toad_> it's not linear
[15:56] <toad_> it's log up to somewhere between 200 and 400
[15:56] <toad_> more or less log
[15:56] <toad_> and then it changes gear
[15:56] <cbreak> well... on a log paper it's linear :)
[15:56] <toad_> and becomes something else
[15:56] <toad_> perhaps a different straight line on log paper?
[15:58] <Ash-Fox> You are the doctor, you are the enemy of the Daleks.. You will be exterminated. Exterminate! Exterminate! Exterminate! Exterminate!
[16:00] <toad_> okay
[16:01] <toad_> do the daleks have any knowledge of anonymous p2p networks? can we trade?
[16:02] * TLF (~francisco@209.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[16:03] <cbreak> hmm... If I taylor interpolate with that data, I get a HTL of 263 for 800 Nodes... And since we all know how well Taylor works...
[16:03] <toad_> hmm?
[16:04] <toad_> what's taylor fitting?
[16:04] <toad_> polynomial?
[16:04] <toad_> any chance of fitting a polylog?
[16:04] <cbreak> yes. Finds a polynomial of degree n-1 for n data points
[16:05] <toad_> i suppose you fit a polylog by taking logs first :)
[16:05] <toad_> anyway what does that give you?
[16:06] <cbreak> taylor interpolation is fairly meanlngless most of the time...
[16:06] <toad_> we need more points, of course
[16:07] <toad_> 200 is closer to 18 than 17
[16:07] <cbreak> but it gives 1/1050000 t^3 - 1/1875 t^2 + 43/300 t + 43/21
[16:08] <toad_> which makes it 8,12,18,35
[16:09] <cbreak> with 18 on 200x25, the polynom gives 193 for size 800
[16:09] <toad_> cbreak: and as a polylog?
[16:11] <toad_> diffs of ~ 0.4, 0.4, 0.7
[16:13] <cbreak> If I set use 0,1,2,3 as x axis, 4 has value 72
[16:13] <toad_> hmm?
[16:13] <cbreak> where 2^t*50 is the network size
[16:13] <toad_> ooh
[16:13] <toad_> clever
[16:13] * TLF (~francisco@131.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[16:14] <toad_> divide it down then log2 it
[16:15] <toad_> bbiab
[16:25] * Ash-Fox (Hal-9000@aas13.neoplus.adsl.tpnet.pl) Quit (Client Quit)
[16:32] * Ash-Fox (Hal-9000@aas13.neoplus.adsl.tpnet.pl) has joined #FreeNET
[16:34] * toad_ starting 3 800-node sims on the BigServer
[16:34] <toad_> along with 2 400-node sims already running
[16:35] <toad_> hopefully this won't seriously degrade service for its real users :|
[16:35] <cbreak> well, let me note I don't think that taylor is the correct interpolation... maybe some kind of linear least squares fitting would be more fitting...
[16:35] <cbreak> you could nice it to 19 :)
[16:51] <toad_> bbl
[16:51] <toad_> see you on monday probably
[16:51] <toad_> might be back today but i doubt it
[16:51] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[17:11] * Hory (~Miranda@81.196.25.110) Quit (Read error: 104 (Connection reset by peer))
[17:13] * TLF (~francisco@131.Red-81-40-113.pooles.rima-tde.net) Quit ("Adi?s a todos y a todas, y que les vaya bien. || http://www.vitaliano.esp.cc")
[17:26] * guido^pe (~unknown@dsl-213-023-244-238.arcor-ip.net) Quit (Remote closed the connection)
[17:43] * moskau23 (~Miranda@dsl-082-082-239-084.arcor-ip.net) has joined #freenet
[17:52] * blueph`le (bluephile@69-160-205-193.clvdoh.adelphia.net) Quit (Client Quit)
[18:14] * guido^pe (~unknown@dsl-213-023-244-238.arcor-ip.net) has joined #freenet
[18:58] * cbreak (~cbreak@dhcp-32-094.via-eth.ch) Quit ("leaving")
[20:04] * TLF (~francisco@131.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[20:04] * goatee (~goatee@ip216-239-84-102.vif.net) Quit (Read error: 104 (Connection reset by peer))
[20:04] * TLF (~francisco@131.Red-81-40-113.pooles.rima-tde.net) Quit (Client Quit)
[20:29] * dli_ (~tst@anna-bhat.uchicago.edu) Quit ("Leaving")
[20:40] * yonkel-cluster (~yonkeltro@pcp04664407pcs.wilog501.pa.comcast.net) has joined #freenet
[20:41] <yonkel-cluster> ok i am ready to setup a node...it can run all the time.....i just need some help with the seting up my nat/firewall setup
[20:41] <yonkel-cluster> the prob is that there are 2 nat's seperating the box in question from the internet
[20:42] <yonkel-cluster> so do i set the ipAddress=x.x.x.x to the first nat ip or the second?
[20:58] <yonkel-cluster> um hello?
[20:59] <d-ArkAngel> which ever one will be on the outside
[21:10] * yonkeltron (~yonkeltro@pcp04664407pcs.wilog501.pa.comcast.net) has joined #freenet
[21:32] * i2p_iip_ (~changate@host.teitel.net) Quit (Remote closed the connection)
[21:32] * tav is now known as tav|offline
[21:33] * i2p_iip (~changate@host.teitel.net) has joined #freenet
[21:35] * tav|offline is now known as tav
[21:45] * yonkel-cluster (~yonkeltro@pcp04664407pcs.wilog501.pa.comcast.net) Quit (Client Quit)
[21:54] * KenMan (~dogzilla@pcp403292pcs.mntcrm01.md.comcast.net) Quit (Read error: 60 (Operation timed out))
[22:15] * aga (~aga@max33.permtelecon.ru) has joined #freenet
[22:19] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[22:21] * aga (~aga@max33.permtelecon.ru) Quit ("Leaving")
[22:22] * aga (~aga@max33.permtelecon.ru) has joined #freenet
[22:31] * guido^pe (~unknown@dsl-213-023-244-238.arcor-ip.net) Quit (Remote closed the connection)
[22:33] * d-ArkAngel (~Rob@cpc2-midd2-5-0-cust16.midd.cable.ntl.com) Quit (Read error: 113 (No route to host))
[22:46] * aga (~aga@max33.permtelecon.ru) Quit ("Leaving")
[22:53] * yonkel-cluster (~yonkeltro@pcp04664407pcs.wilog501.pa.comcast.net) has joined #freenet
[22:56] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit ("Client exiting")
[23:00] * goatee (~goatee@ip216-239-84-102.vif.net) has joined #freenet
[23:38] * greycat (rfc1413@wooledge.org) has joined #freenet
[23:41] <greycat> signal needs to increment the NIM slot numbers on his page.
[23:42] <yonkel-cluster> could someone tell me why my freenet.conf file has only 2 lines, niether of which are going to help me configure my node with my NATs (yes plural)?
[23:43] <greycat> remove the file and then run "java -jar freenet.jar --config" to regenerate a good one
[23:43] <greycat> then edit it by hand
[23:43] <yonkel-cluster> k
[23:43] <yonkel-cluster> should i stop freenet first?
[23:43] <greycat> yes
[23:44] <yonkel-cluster> it seems to have stopped after listenPort [12255]
[23:45] <greycat> you did hit enter, right...?
[23:46] * yonkel-cluster grumbles and turns red
[23:46] <yonkel-cluster> of course
[23:46] * yonkel-cluster grins
[23:47] <yonkel-cluster> greycat: if i wanted a 10gb store....what would i enter?
[23:47] <greycat> just edit the file by hand afterward
[23:48] <yonkel-cluster> k
[23:48] <greycat> change the storeSize line to storeSize=10G
[23:48] <greycat> (I think the --config tool has a bug that prevents you typing 10G there...)
[23:50] <yonkel-cluster> greycat: i have 2 NATs...........
[23:50] <yonkel-cluster> what do i change my ipAddress line to say?
[23:51] <greycat> whichever IP address will let the outside world reach you
[23:51] <yonkel-cluster> my ip address or the address of one of the routers?
[23:51] <yonkel-cluster> ok
[23:54] <yonkel-cluster> greycat: storeSize=10GB is an acceptable value?
[23:54] <greycat> as long as your disk is at least that big
[23:57] <yonkel-cluster> ummm i got an error
[23:57] <yonkel-cluster> onfig error: storeSize=7GB - Value could not be parsed - format error perhaps? - expected Long (whole number, up to 9,223,372,036,854,775,808, kKmMgGtTpPeE accepted - example 2.1m = 2,100,000) - detail: java.lang.NumberFormatException: For input string: "7GB"
[23:57] <yonkel-cluster> java.lang.NumberFormatException: For input string: "7GB"
[23:57] <yonkel-cluster> at java.lang.NumberFormatException.forInputString(NumberFormatException.java:48)
[23:57] <yonkel-cluster> at java.lang.FloatingDecimal.readJavaFormatString(FloatingDecimal.java:1213)
[23:57] <yonkel-cluster> at java.lang.Double.parseDouble(Double.java:202)
[23:57] <yonkel-cluster> at freenet.config.Params.parseLong(Params.java:362)
[23:57] <yonkel-cluster> at freenet.config.Params.getLong(Params.java:314)
[23:57] <yonkel-cluster> at freenet.node.Node.init(Node.java:3454)
[23:57] <yonkel-cluster> at freenet.node.Main.main(Main.java:478)
[23:57] <yonkel-cluster> greycat: i changed it to 7GB
[23:58] <greycat> 7G not 7GB
[23:58] <yonkel-cluster> o
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.
