Timestamps are in GMT/BST.
[0:05] * jman88 (i=jman@) has joined #freenet
[0:09] * ButcherCH (n=caesar@) has joined #freenet
[0:09] * ButcherCH is now known as Caesarus
[0:24] * jman88 (i=jman@) Quit (Read error: 110 (Connection timed out))
[0:41] * NullAcht15 (n=NullAcht@) Quit (Remote closed the connection)
[0:45] * PhrostByte is now known as phrosty
[0:50] * phrosty (n=phrosty@) Quit ("PostQuitMessage(0);")
[0:54] * phrosty (i=phrosty@) has joined #freenet
[1:03] * this (n=chatzill@) has joined #freenet
[1:05] * this (n=chatzill@) has left #freenet
[2:07] * jman88 (i=jman@) has joined #freenet
[2:10] * ralith_ (n=ralith@) has joined #freenet
[2:13] * Caesarus (n=caesar@) Quit ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
[2:20] * Iceman_B (n=Ice@) Quit ("You cannot see nor judge others. The only thing you see and judge is you. You. Can. Not. SEE. Others. So...stfu...ktnxbye.")
[2:25] * Ralith (n=ralith@) Quit (Read error: 110 (Connection timed out))
[2:26] <CIA-14> zothar * r11276 /trunk/apps/pyFreenet/refbot.py: refbot: Bots now gather each other's node identities; closer to bot2bot ref trading.
[2:27] * jman88 (i=jman@) Quit (Success)
[2:48] <CIA-14> zothar * r11277 /trunk/apps/pyFreenet/refbot.py: refbot: Introduce commented config code for bot2bot cooperative announcements.
[2:48] <CIA-14> zothar * r11278 /trunk/apps/pyFreenet/refbot.py: refbot: Fix description of cooperative bot announcements option (still commented).
[2:49] * ralith_ is now known as Ralith
[3:13] * oryx8801 (n=oryx_880@) has joined #freenet
[3:23] * oryx8801 (n=oryx_880@) has left #freenet
[3:34] <Ralith> finally got a workaround for that damned display bug
[3:40] * droptothteop (n=droptoth@) Quit (Read error: 110 (Connection timed out))
[3:46] * kandahar (n=kandahar@) has joined #freenet
[3:47] <kandahar> Having a problem with my noderef (with a dynamic IP) anyone have any advice?
[3:49] <kandahar> this is the problem: my node ref has two IPS - the first is the one I was on for three days, the second is one my ISP assigned temporarily while I was having a network problem
[3:50] <kandahar> the actual IP I've been on for the last couple of hours however is not in my node ref :( freenet is up and running, but I'm connected to less peers than I would prefer to be to
[3:54] <Ralith> Why does frost wait so long to send?
[3:59] * sanity (n=ian@) has joined #freenet
[3:59] * ChanServ sets mode +o sanity
[3:59] * sanity (n=ian@) Quit (Client Quit)
[4:00] * sanity (n=ian@) has joined #freenet
[4:00] * ChanServ sets mode +o sanity
[4:08] * jman88 (i=jman@) has joined #freenet
[4:18] * CIA-14 (i=cia@) Quit ()
[4:19] <_ph00> freenet is not detecting my external nat address any more
[4:19] <kandahar> freenet's fucked up mine too
[4:19] <kandahar> I restarted my node
[4:19] <kandahar> it has now picked up my new one, but the port at the end is wrong (it's not my listen port, it's a different one)
[4:20] <_ph00> that's normal
[4:20] <_ph00> nat changes port numbers
[4:20] <_ph00> but
[4:21] <_ph00> if you have your listen port properly forwarded, it should detect the roigght port number
[4:21] <kandahar> so do I have to change my router to forward the new port now?
[4:21] * CIA-14 (i=cia@) has joined #freenet
[4:21] <kandahar> okay, i'll see how it goes
[4:21] <_ph00> wha new port? the done detcted as port number with the external nat?
[4:21] <_ph00> no, you don't need to forward that
[4:22] <kandahar> okay.
[4:23] <_ph00> only check that the port foprwarding on your listen port is OK
[4:23] <kandahar> yep that's the easy bit :) it always is.
[4:25] <kandahar> if my node ref changes, does freenet automatically 'pass' it to peers who had the previous one (even if they're not connected)
[4:27] <_ph00> only if the connect, I suppose
[4:27] <_ph00> they
[4:28] <kandahar> indeed, I was meaning if they were connected through a couple of other peers
[4:28] <Zothar> if you can get a connection to at least one peer, your node can use that connection to insert your updated ARK, which your other peers should eventually be able to retrieve and reconnect to you
[4:28] <kandahar> the problem being, if two nodes were both on a dynamic IP address, and both IPs changed, they would never be able to contact each other again, unless something like that was happening
[4:28] <kandahar> aahh good :) that answers my question
[4:28] <_ph00> aaahhh... OK then.
[4:29] <kandahar> the ARK is my IP encrypted with something else in my ref?
[4:33] * jman88 (i=jman@) Quit (Connection timed out)
[4:35] <Zothar> the ARK _is_ your ref; only the people who have seen your ref know the key to your ARK
[4:36] <Zothar> (I think there might be a few ref fields left out of the ARK because they're not needed, but I don't remember for sure and am too lazy to check the code)
[4:36] <kandahar> cool, and if the version of the ref they have is out-of-date the ARK will overwrite it presumably
[4:38] <zorton> Zothar: congrats, your bot worked quite well and grabbed 10 node refs
[4:38] <Zothar> zorton: good to hear; you migth have noticed some of my refbot commit messages a few hours ago; inter-bot trading is getting closer :)
[4:39] <Zothar> kandahar: yes
[4:39] <zorton> Zothar: cool, i'll run the update at some point and rerun the bot
[4:40] <Zothar> zorton: it'll probably not actually do inter-bot ref trading until this weekend or later still, but at least I'm closer now
[4:40] <Ralith> Why does frost wait so long to send any given message?
[4:40] <Ralith> it's not like it takes ages to upload
[4:40] <Ralith> but it takes ages to START uploading
[4:41] <zorton> Zorthar: would it worth running the bot again from time to time?
[4:41] <_ph00> it start uploading when that board start updating, probably
[4:42] <kandahar> ph00: didn't know that about frost
[4:43] <_ph00> boards don't update themselves all the time, they do that in turns
[4:43] <kandahar> yeah, I thought posting was processed immediately though. might be wrong
[4:44] <_ph00> so if you send a message to a board that updated a few minutes earlier, frost will wait untill next update for that mesaage
[4:44] <_ph00> you can also force update with "refresh board"
[4:44] <Ralith> _ph00, ah
[4:44] <Ralith> thanks
[4:44] <_ph00> np
[4:45] <Ralith> _ph00, also, where does freenet store files being downloaded via it's queue while they're in progress?
[4:46] <kandahar> temp directory, encrypted I think.
[4:46] <_ph00> temp-store or something like that, but you can't see them anyway, they're encrypted until the download is complete
[4:46] <Ralith> ah
[4:47] <_ph00> kandahar; we can't connect. I know that you can??t connect to me because I can't port forward, so your port forwarding is not working properly
[4:47] <kandahar> depends if you use thaw, and what settings you use as well I think
[4:48] <kandahar> yeah, it's a bit strange. i'm going to try forwarding an additional port... shouldn't have to but i'll see what it does
[4:48] <Ralith> _ph00, and one other thing: is anyone ever going to address the swing+xgl issue?
[4:49] <_ph00> thaw or fproxy don't make a difference. thaw is just an interface to the node. once the download or the upload has started you can shut thaw down and it will keep going
[4:49] <kandahar> unless you don't add the download to the global queue, that is
[4:49] <_ph00> Ralith; sorry, I don't even know what that is
[4:49] <Ralith> _ph00, swing+xgl=blank window
[4:49] <_ph00> thaw adds downloads and uploads to the global queue
[4:50] <Ralith> or swing+aiglx
[4:50] <Ralith> well, really, swing+(compiz||beryl)
[4:50] <_ph00> I don't know what you're talking about, really
[4:50] <kandahar> yes, you can set 'global queue' to false though in which case it doesn't
[4:50] <Zothar> zorton: re running the bot again: yeah, if not all of your peers stick around on Freenet, you may need to supplement your peers from time to time; you can configure how many refs it adds at a time
[4:51] <_ph00> well, OK. In that case you'll have to keep thaw open
[4:51] <Ralith> isn't it better to replenish your peers from frost?
[4:51] <Ralith> with encrypted transfer and throwaway ID?
[4:52] <_ph00> yeah. I guess so. But you will connect to strangers anyways
[4:52] <kandahar> I'm still confused about my physical.udp line in my node ref - it gives out two IPs: the first one is not my correct IP, but has my correct FNP port at the end. The second one *is* my correct IP, but the port is wrong.
[4:52] <_ph00> so, that's *slightly* better
[4:52] <zorton> Zothar: good to know, things seem to be running decently right now
[4:53] <kandahar> the advantage of doing that is that no one can 'spy' on topology and figure out who's connecting to who. But I can't really conceive of people doing that at this point.
[4:53] <_ph00> kandahar; *and* you can't connect to peers who are behind a NAT with no port forwarding (me)
[4:53] <zorton> i'm assuming the robustness of my freenet connection will be based upon how many "old timer" nodes I'm connected to
[4:53] <Ralith> kandahar, it's good to have good habits :P
[4:53] <kandahar> true
[4:53] <Zothar> zorton: yeah, until you become someone else's "old timer"
[4:54] <Zothar> kandahar: multiple IPs in your ref is not a problem as long as the one that people can connect to you with is there
[4:54] <kandahar> yep. I got that now.
[4:55] <kandahar> so the limitation of being behind a NAT is that I can't connect to anyone who's behind a NAT who isn't able to forward ports, even if I can forward mine
[4:56] <zorton> it's funny the JSTUN plugin keeps unloading and freenet complained it couldn't find my IP earlier
[4:56] <zorton> the connection has been up steady though
[4:57] <_ph00> jstun is loaded here, yet the node can't detect the external IP
[4:57] <Zothar> kandahar: you should be able to connect to those behind NAT if you can forward your port unless I'm forgetting some detail
[4:58] <Zothar> been awhile since I've let it run this long:
[4:58] <Zothar> # networkSizeEstimateSession: 779 nodes
[4:58] <Zothar> # networkSizeEstimate24h: 545 nodes
[4:58] <Zothar> # networkSizeEstimate48h: 622 nodes
[4:58] <Zothar> # avrConnPeersPerNode: 6.762522 peers
[4:58] <Zothar> # nodeUptime: 4d4h
[4:59] <kandahar> I had about the same number of nodes in my 48hour session before as well, a few hours ago
[4:59] <kandahar> does it work that out from the number of ARKs I've pulled in?
[5:00] <Zothar> kandahar: no, that estimate is based on location swap requests
[5:00] <_ph00> kandahar; the limitation of being behind a NAT is that you can't connect to pees who are behind NATs if *neither* of you can't port forward. if one of the two can port forward, that should be enough. Now, I'm behind a NAT and I can't port forward because of my lousy ISP, and I can't connect to you, so I guess your port forwarding is not working properly
[5:01] <zorton> anyone know a good way to test if the port forwarding is working?
[5:02] <zorton> in my case I can most certainly forward but the UDP stuff has been a bit wonky...but I don't have a good way of testing
[5:02] <kandahar> it should be working though ph00. the NAT is my own router, my PC is on a static internal IP, and it's forwarding both TCP and UDP packets. I never had a problem connecting to anyone before earlier today, when I got disconnected, and freenet got confused by my external IP changing
[5:03] * hlag (n=snowball@) has joined #freenet
[5:03] <_ph00> zorton; I would use bittorrent to test port forwarding
[5:04] <_ph00> kandahar; yes, it should be working but it's not. I wonder what the problem is...
[5:04] <zorton> _ph00: the udp forwarding?
[5:05] * Zothar (n=Zothar@) Quit ("Chatzilla 0.9.74 [Firefox 1.5.0.8/2006102516]")
[5:05] <_ph00> no, the tcp, I was assuming that if one worsk the other should work too
[5:06] <zorton> yeah
[5:06] <zorton> tcp forwards work great, it's the UDP stuff that's been problematic
[5:06] <kandahar> what's weird is that before I had connection problems, freenet knew it was behind a NAT
[5:06] <kandahar> it gave an error, and said that I would have problems because I might be connected behind a symmetric NAT. as it turned out, I had problems connecting to nobody
[5:07] <kandahar> now, recently, I've restarted my node (because my reference didn't include my external IP, so I presumed it was borked) and freenet doesn't give me an error for being behind a NAT anymore. But it turns out I've got problems connecting to one or two people :)
[5:07] <_ph00> another weird thing is that I am behind a NAT but I never saw the NAT warning
[5:08] <kandahar> I reckon that's the problem
[5:08] * CIA-14 (i=cia@) Quit ()
[5:08] <kandahar> btw, azeurus is happy with being behind the NAT. so forwarding is definitely working for me
[5:09] <_ph00> that's the tcp forwarding
[5:09] <kandahar> True,
[5:09] <_ph00> anywaysw
[5:09] <_ph00> I 'd guess that it's a bug
[5:09] <_ph00> because
[5:09] <kandahar> Isn't freenet using UDP because it's supposed to work better with NATs than TCP? :D
[5:10] <mk_> what was that attack on freenet a while ago? something to do with positioning hostile nodes around the target...?
[5:10] <_ph00> freenet was detecting both addresses until yesterday (or maybe the day before)
[5:10] <_ph00> now, it only detects the internal address
[5:13] <BosHaus> yeah, mine wouldn't detect the external, now its griping that its behind a nat
[5:13] <BosHaus> even though UDP port forwarding works great for I2P
[5:15] <_ph00> doesn't detect external, not even with port forwarding ?
[5:15] <_ph00> well
[5:15] <_ph00> that's definitely a bug
[5:15] <BosHaus> it detected my external now
[5:15] <_ph00> it did?
[5:15] <_ph00> ok
[5:15] <BosHaus> it shows it 2x in the darknet config though.. which is odd
[5:16] <BosHaus> noderef I mean
[5:16] <BosHaus> whatever its called :)
[5:16] <_ph00> I don't get it
[5:16] <_ph00> where
[5:16] <_ph00> 2x ?
[5:16] <BosHaus> physical.udp=ip:port;ip:port
[5:16] * BosHaus was kicked from #freenet by FreenetLogBot
[5:16] * BosHaus (i=nobody@) has joined #freenet
[5:16] <BosHaus> oops :P
[5:17] <_ph00> you mean it shows both addresses?
[5:17] <BosHaus> the exact same thing 2x
[5:17] <BosHaus> I'd paste it but its running in a VM
[5:17] <BosHaus> since I use a mac..
[5:17] <_ph00> it runs on a vm anyways
[5:17] <BosHaus> oh
[5:17] <BosHaus> well, anyway, it seems to have detected it
[5:18] <BosHaus> finally
[5:18] <BosHaus> but I wonder why it says I"m behind a nat now
[5:18] <_ph00> and a mac is not so different from a linux os, which is used by many freenet users
[5:18] <_ph00> but that's not the point
[5:18] <BosHaus> I'm just too lazy to set it up :)
[5:18] <_ph00> you have your port forwardeed
[5:18] <_ph00> and it keeps complaining about the NAT?
[5:19] <BosHaus> yep.. DNAT net loc:192.168.0.236 udp 13498 in shorewall
[5:19] <BosHaus> which should be doing it
[5:19] <_ph00> I think we can only wait for the devs to look at this
[5:19] <BosHaus> "Port restricted cone nat detected"
[5:20] * CIA-14 (i=cia@) has joined #freenet
[5:20] <_ph00> teh waird is that I am definitely behind a nat, and I can't port forward because it'n not my nat but the ISP's, and that ISP is an asshole, so I can't port forward...
[5:21] <_ph00> yet, freenet never complained about NATs
[5:21] * MineHaunter (n=van85@) Quit (Read error: 104 (Connection reset by peer))
[5:21] <zorton> _ph00: how does your ISP prevent you from port forwarding?
[5:22] <_ph00> they own the router, and I don'y have access to it
[5:22] <zorton> and the router spits DHCP towards your network with a private IP?
[5:23] <_ph00> I'm on a WAN run by the ISP
[5:23] <zorton> and your IP isn't publicaly routable? or is a public IP but firewalled to heck?
[5:23] <_ph00> the router does the switch, I see both 'local' addresses and 'real' adresses when I netmap
[5:24] <_ph00> there are different ranges
[5:24] * sleon (n=sleon@) has joined #freenet
[5:24] * ChanServ sets mode +o sleon
[5:24] * mk_ (n=mk@) has left #freenet
[5:24] <zorton> sounds like you need to switch ISP's
[5:25] <_ph00> I'm on a 39 range, locally, I also am on a 85.18 range, which is not accessible from the internet, and if I want to be able to listen for inbound connections I have to pay extra (4 euros a day!) amd I get another IP on a 62.101 range, which is the same rage as their webpages etc
[5:26] <zorton> sux
[5:26] <_ph00> zorton; sadly enough, that's the only ISP in theis damn country that gives me some real speed, expecially on uploads
[5:26] <_ph00> but
[5:27] <_ph00> I'm very limited by that asshole ISP
[5:27] <_ph00> but I like the speed and the stability of the connection
[5:27] <zorton> phew, there must not be much of a outbound pipe from your area
[5:27] <_ph00> btw, switching ISP is not so easy, that would take months and cost money
[5:27] <_ph00> and no, I don't live in some losy Asian or African country, I'm actually inside the EU
[5:28] <Ralith> O.o
[5:28] <zorton> don't worry i'm familar with the problem, I'm up in Alaska and for a long time access was limited to dialup while the rest of the country started going broadband
[5:28] <_ph00> there is only one company that runs fiber, all the other have adsl
[5:29] <Ralith> zorton, sattelite! :P
[5:29] <zorton> latency gets you
[5:30] <zorton> i've installed Sat inet links and typicaly you see latency around 600ms at the very least
[5:30] <Ralith> yeah, but at least you get bandwidth
[5:30] <zorton> you're having to talk someting like 25k miles one way towards the transponder at the speed of light
[5:30] <zorton> true, but not typicaly much better than ADSL offerings up here
[5:31] <zorton> there was a country doing NNTP downlinks with terrestial uploads though
[5:31] <zorton> a friend of mine and I where going to setup something for that but never got to it
[5:31] <_ph00> I thought about satellites, I was thinking, if I get myself a damn fast satellite connection and share it with 10 or 12 people, that would cost as much as the lousy ISP, maybe less, and give me a better service... but the problem are uploads. you cant send to the satellite, uploads go on the telephone line (that sounds crazy, but that's what I heard)
[5:31] <_ph00> (not sure about that, tho)
[5:31] * sleon_ (n=sleon@) Quit (Read error: 110 (Connection timed out))
[5:32] <Ralith> _ph00, the thing is
[5:32] <_ph00> and then, there's the lag
[5:32] <Ralith> you pay for the data channel
[5:32] <Ralith> so download==price
[5:32] <Ralith> download/upload==2*price
[5:32] <zorton> _ph00: it depends on your latitiude, checkout wikipedia's page on sat internet
[5:32] <zorton> some services are handling uploads via sat now instead of terrestrial
[5:32] <_ph00> latitude is OK, I'm in suthern EU
[5:32] <_ph00> southern
[5:32] <Ralith> for non-serving usage modem is fine, anyway
[5:35] <_ph00> anyways
[5:35] <zorton> _ph00: http://en.wikipedia.org/wiki/Satellite_internet
[5:35] <_ph00> I'm gonna keep the lousy assholes for now
[5:36] * zorton shrugs
[5:37] * kandahar (n=kandahar@) Quit (Read error: 110 (Connection timed out))
[5:40] * kandahar (n=kandahar@) has joined #freenet
[5:51] * heph (n=heph@) has joined #freenet
[6:21] * mk_ (n=mk@) has joined #freenet
[6:21] <mk_> how will premix routing be implemented without making the darknet harvestable?
[6:22] <kandahar> onion encryption or something i guess
[6:22] <kandahar> don't really know how that works
[6:23] <mk_> onion encryption is pretty standard, it seems that you use it to basically mean 'premix routing'
[6:23] <mk_> the problem I'm getting at is that you have to choose a set of routers
[6:24] <mk_> to be part of that premix chain. but choosing your 'friends' is a bad idea
[6:24] <mk_> but what other option is there?
[6:25] <kandahar> umm... anyone on the network
[6:25] <kandahar> shouldn't need to be someone directly connected to you
[6:25] <kandahar> only someone who routes to you
[6:26] <mk_> yes but the only people you know on the network are your friends. if you can choose anyone on the entire network, that would defeat the point of having a darknet, wouldn't it?
[6:27] <kandahar> you don't *choose* the premixers
[6:27] <kandahar> i imagine it would be a random process, similar to inserting/retrieving
[6:27] <mk_> so who would choose them?
[6:28] <kandahar> er, I don't know. I'm not the person to ask this
[6:28] <mk_> your friend that you are trying to defend against in the first place?
[6:29] <mk_> perhaps, but until someone else answers, you're quite free to share whatever knowledge you have
[6:32] <kandahar> if you use encryption on the routing, than a note will only know a) where the request comes from and b) where to send it to, so it won't be possible to ever learn any peers other than your immediate ones
[6:32] <kandahar> mixmaster style
[6:33] <kandahar> this is all future development stuff, though, and I'm sure they will be working out the best way to do it
[6:33] * scsifuzzy (n=taaten@) has joined #freenet
[6:33] <scsifuzzy> hey guys
[6:34] <scsifuzzy> i'm new
[6:34] <kandahar> welcome to the club
[6:34] <scsifuzzy> thanks :)
[6:35] <mk_> right, but you understand that you have to give the "where to send it to" instruction, yes?
[6:35] <mk_> if you're my friend, I can give you the encrypted instruction to connect to... who? I don't know any of the rest of the network
[6:36] <kandahar> sure.
[6:36] <scsifuzzy> I need a ref or something to connect to freenet, but the site says that I need someone I trust, can anyone explain?
[6:36] <scsifuzzy> i'm so lost
[6:36] <kandahar> you don't know the real location of it, however you know an abstract kind of it
[6:36] <mk_> location of what?
[6:36] <kandahar> the other node.
[6:37] <mk_> scsifuzzy, see the channel topic for where to go to get refs
[6:37] <scsifuzzy> oooooh
[6:37] <scsifuzzy> ok
[6:37] <scsifuzzy> thanks
[6:37] * scsifuzzy (n=taaten@) has left #freenet
[6:37] <mk_> scsifuzzy, you won't find anyone you can trust unless you know a friend who uses freenet, so you'll ...
[6:37] <kandahar> you're not harvesting it because you'll never know where it is, only the nodes that directly connect to it do, but you know which node you have to send it to to ultimately get there
[6:38] <mk_> to it? to an onion router?
[6:39] <mk_> you're saying that they should be identified by virtual addresses, like in, say, mute?
[6:39] <kandahar> any node.
[6:39] <kandahar> they already are, pretty much
[6:39] <kandahar> I only know the IP addresses of the peers directly connected to me, yet I can route to any peer on the network.
[6:40] <mk_> that's... mind-bogglingly inefficient, or wasteful, or... something.
[6:40] <kandahar> it's called sacrificing efficiency for privacy
[6:41] <mk_> using the freenet layer (and not the standard internet layer) to do the premix would ... well, kill the available resources
[6:42] <mk_> you can only sacrifice so much efficiency before usability is severly affected, and when usability is affected, so is privacy
[6:43] * kandahar_ (n=kandahar@) has joined #freenet
[6:43] * kandahar (n=kandahar@) Quit (Read error: 104 (Connection reset by peer))
[6:48] <kandahar_> does 'location' have any particular significance?
[6:49] <mk_> in what context?
[6:49] <kandahar_> darknet peer information
[6:49] <BosHaus> GPS coordinates!
[6:49] <BosHaus> ;)
[6:52] * hlag (n=snowball@) Quit ()
[7:04] * CIA-14 (i=cia@) Quit ()
[7:10] <fridim> the last stable version seems to crash down every day in average.
[7:11] <kandahar_> Mine stayed up for two days.
[7:11] <kandahar_> I had to reset due to connection problems
[7:14] * kandahar_ is now known as kandahar_afk
[7:32] * kandahar (n=kandahar@) has joined #freenet
[7:36] * CIA-14 (i=cia@) has joined #freenet
[7:38] * kandahar_afk (n=kandahar@) Quit (Read error: 113 (No route to host))
[7:41] <_ph00> I'm getting lower up- and download speed since 1006
[7:42] <_ph00> the BW values in stats page look like normal
[7:42] <_ph00> but files up/downloads take even longer than they used to
[7:42] <_ph00> some are actually stuck
[7:42] <_ph00> or so slow that they look stuck
[7:43] * funnycat (n=root@) has joined #freenet
[7:44] <kandahar> both my output rates and input rates are consistently below what I specified (and what my bandwidth can handle). downloads vary, as you expect
[7:44] <kandahar> don't know what it was like before, i'm new with 1006 - maybe its just coincidence because a number of people are joining 0.7
[7:45] <_ph00> BW rates being consitently lower than you allow is normal
[7:45] <_ph00> I allow 1000K in both direction and I get ~10
[7:45] <kandahar> unlike previous versions of freenet? good :)
[7:45] <_ph00> always been so
[7:45] <kandahar> heh, interesting.
[7:46] <_ph00> you can set your max allowd BW as high as your total BW, the node will only use a small fraction of that
[7:47] * funnycat (n=root@) has left #freenet
[7:50] * _ph00 (n=z@) Quit ("Leaving")
[7:53] * CIA-14 (i=cia@) Quit ()
[7:54] * phrosty (i=phrosty@) Quit ("PostQuitMessage(0);")
[7:57] * CIA-14 (i=cia@) has joined #freenet
[8:07] * CIA-14 (i=cia@) Quit ()
[8:15] * railk (n=railk@) has joined #freenet
[8:20] * phrosty (i=phrosty@) has joined #freenet
[8:27] <fridim> "you can set your max allowd BW as high as your total BW, the node will only use a small fraction of that" <- is that wanted or is it due to other's allowed max bandwith ?
[8:29] * CIA-14 (i=cia@) has joined #freenet
[8:35] * CIA-14 (i=cia@) Quit ()
[8:42] <zorton> or is it due to the speed of freenet in general
[8:43] <mk_> I don't know, but it might be one of two settings
[8:44] <mk_> one will be the average (which may be higher, or lower, at times)
[8:44] <mk_> and this one might be the hard limit on how high it can be
[8:50] * ShipHead (i=ShipHead@) Quit (Remote closed the connection)
[8:53] * CIA-14 (i=cia@) has joined #freenet
[8:59] * whiterabbit (n=Whiterab@) has joined #freenet
[9:01] * railk (n=railk@) Quit ("Cya, wouldn't want ta be ya!")
[9:10] * sanity (n=ian@) Quit ()
[9:16] * IMCensored1_ (n=KMIntern@) has joined #freenet
[9:17] * _ph00 (n=z@) has joined #freenet
[9:22] * CIA-14 (i=cia@) Quit ()
[9:32] * IMCensored1 (n=KMIntern@) Quit (Read error: 110 (Connection timed out))
[9:32] * IMCensored1_ is now known as IMCensored1
[9:38] * mk_ (n=mk@) has left #freenet
[9:58] * kandahar (n=kandahar@) has left #freenet
[10:17] * CIA-14 (i=cia@) has joined #freenet
[10:20] * phrosty (i=phrosty@) Quit ("PostQuitMessage(0);")
[10:57] * Tritman (i=Tritman@) has joined #freenet
[11:08] * timmy2chk (n=violent@) has joined #freenet
[11:33] <fridim> the node crashed again :/
[11:34] * bluescreen (n=bluescre@) has joined #freenet
[11:35] * bluescreen (n=bluescre@) has left #freenet
[11:39] <fridim> i'am checking the log.
[11:55] * railk (n=railk@) has joined #freenet
[12:03] * colione (n=colione@) has joined #freenet
[12:20] * toad_ (n=toad@) has joined #freenet
[12:20] * ChanServ sets mode +o toad_
[12:54] * slinky (n=slinky@) Quit (Read error: 104 (Connection reset by peer))
[12:55] * slinky (n=slinky@) has joined #freenet
[13:12] * nextgens (n=nextgens@) has joined #freenet
[13:13] <nextgens> hi
[13:13] * nextgens is fixing mantis*
[13:13] * whiterabbit (n=Whiterab@) Quit (Read error: 110 (Connection timed out))
[13:16] * sanity (n=ian@) has joined #freenet
[13:16] * ChanServ sets mode +o sanity
[13:29] <sleon> woot
[13:29] <sleon> i recenlty managed to download 500 mb video with freenet!
[13:29] <sleon> did not even notice it
[13:29] <sleon> thats cool
[13:30] * greycat (i=rfc1413@) has joined #freenet
[13:33] <nextgens> sanity> toad_> ok, mantis is fixed : dovecot seems to be working too ... anything else ?
[13:34] <sanity> nextgens: thanks, apologies for the screw up - it was careless of me
[13:34] <nextgens> emu will need a reboot anyway
[13:34] <nextgens> Will do tonight or this afternoon
[13:36] <sanity> that monitoring script thing is cool
[13:40] <sopues> its been reporting many times about that particular mirror though
[13:40] <sopues> I wonder if the mirror has slower update time than the script or something
[13:47] * m01sbc (n=m01sbc@) has joined #freenet
[13:49] <paveq> nextgens: ping
[13:49] <paveq> nextgens: did you got my reply?
[13:52] <nextgens> paveq> yes, I got your mail, thanks
[13:54] <paveq> hmm?
[13:55] <paveq> I didnt
[13:55] <paveq> I didn's send any mail
[13:55] <paveq> aarg
[13:55] <paveq> typos :P
[14:00] * railk (n=railk@) Quit ("Cya, wouldn't want ta be ya!")
[14:01] * m01sbc (n=m01sbc@) Quit (Remote closed the connection)
[14:04] * IMCensored1_ (n=KMIntern@) has joined #freenet
[14:11] * Urs_ShPo (n=gaim@) has joined #freenet
[14:12] * Urs_ShPo (n=gaim@) Quit (Client Quit)
[14:13] * Romain (n=Romain@) has joined #freenet
[14:13] <Romain> hey yall
[14:18] <toad_> hiya folks
[14:18] <toad_> nextgens: thanks for fixing mantis
[14:20] * Zothar_Work (n=chatzill@) has joined #freenet
[14:23] * Urs_ShPo (n=gaim@) has joined #freenet
[14:24] * Urs_ShPo (n=gaim@) Quit (Remote closed the connection)
[14:24] * Urs_ShPo (n=gaim@) has joined #freenet
[14:24] <toad_> nextgens: what was it?
[14:25] * Urs_ShPo (n=gaim@) Quit (Client Quit)
[14:26] * Urs_ShPo (n=gaim@) has joined #freenet
[14:26] * Urs_ShPo (n=gaim@) Quit (Remote closed the connection)
[14:27] * Urs_ShPo (n=gaim@) has joined #freenet
[14:30] <nextgens> toad_> what was what ?
[14:31] * Urs_ShPo (n=gaim@) Quit (Client Quit)
[14:31] * Urs_ShPo (n=gaim@) has joined #freenet
[14:31] * IMCensored1 (n=KMIntern@) Quit (Read error: 110 (Connection timed out))
[14:31] * IMCensored1_ is now known as IMCensored1
[14:32] <nextgens> suPHP has a debian package for apache2 now ... Maybe we can migrate everything to apache2 now
[14:32] <nextgens> will do when I get around it
[14:32] <sanity> nextgens: that would be great
[14:33] * railk (n=railk@) has joined #freenet
[14:38] * Romain (n=Romain@) Quit ()
[14:45] <toad_> nextgens / sanity do you think we should provide auto-updating for plugins? a lot of people (including Entry Point's author) still have the old Librarian plugin, which has security issues... and there is feature work to be done on it too now that people are actually producing indexes
[14:45] <toad_> sanity: also, i'd be interested in your views on RSS feeds
[14:50] <nextgens> toad_> I don't think it's urgent
[14:50] <nextgens> toad_> and before we do, we need a rock solid plugin api
[14:51] <nextgens> with a defined versionning scheme
[14:51] <toad_> you think _that_ is urgent? why?
[14:51] <nextgens> I wrote I don't think it is
[14:51] <nextgens> as most plugins aren't maintained ...
[14:52] <toad_> here's the dilemma
[14:52] <nextgens> and we have been planning to create a new API since ages
[14:52] <toad_> the Librarian plugin could be a significant benefit to freenet
[14:52] <toad_> Entry Point wants to maintain an index, and has produced one
[14:52] <nextgens> hence creating an updater for our already legacy API is kinda dumb imo
[14:52] <toad_> at the same time, a lot of people have old versions of Librarian pre-installed which are insecure
[14:53] <toad_> and any changes we make will probably need the plugin to be updated
[14:53] <nextgens> Bombe> any progress on the "newGen." plugin api ?
[14:53] <toad_> now, second issue
[14:53] <toad_> RSS is wildly popular outside of Freenet
[14:53] <toad_> mozillaman (freenetwatch) wants to provide an RSS feed on his site
[14:53] <toad_> this will at present be insecure, and won't be auto-detected by the browser; must be added manually
[14:54] <toad_> in both cases we have content initiatives from the community which I want to support
[14:54] <toad_> which require some effort on this side
[14:55] <nextgens> may I restart emu *now* ?
[14:55] <toad_> if you're going to be here to deal with it if it fails to start, then yes
[14:56] <nextgens> ok, I'll do it now then
[14:57] * nextgens looks for the credentials to logon the serial line
[14:57] * phrosty (i=phrosty@) has joined #freenet
[14:59] <toad_> what would be a good MIME type for a freenet index?
[14:59] <nextgens> ok, will do
[14:59] <nextgens> brb
[14:59] * toad_ doesn't think x-application/librarian is good
[14:59] * nextgens (n=nextgens@) Quit ("rebooting")
[14:59] <toad_> application/x-librarian isn't good either, it could mean anything
[14:59] <toad_> application/x-freenet-index maybe
[15:12] [freenode-connect VERSION]
[15:12] * FreenetLogBot (n=PircBot@) has joined #freenet
[15:12] * Topic is 'http://freenetproject.org/download.html (990 - mandatory), please read that page before asking for help here. If you want to exchange references, join #freenet-refs | http://wiki.freenetproject.org/FrequentlyAskedQuestions | logs: http://emu.freenetproject.org/irc/ | Tor blocked due to spam (get an op to voice you) | #freenet-fr #freenet-se #freenet-es #freenet-'
[15:12] * Set by toad_ on Fri Dec 01 22:56:34 UTC 2006
[15:12] * nextgens sets mode +o FreenetLogBot
[15:12] * nextgens wonders why we have postgresql on emu
[15:13] <nextgens> there is no debian dependancy on it
[15:15] <nextgens> sanity> have you installed it ? are you using it ?
[15:18] * railk (n=railk@) has joined #freenet
[15:22] <nextgens> RAAAHHH
[15:22] <nextgens> emacs is the default editor :<
[15:23] <railk> does FreenetLogBot frequently repeat itself? :P
[15:24] <sleon> hehe :)
[15:24] <railk> well this is a great way to make the chan completely unreadable :P
[15:24] <Zothar_Work> somebody take down that site so FreenetLogBot can get "sorted" :)
[15:25] <toad_> :)
[15:25] <toad_> :(
[15:25] <railk> someone just +b FreenetLogBot!*@* temporarily...
[15:26] * toad_ sets mode -v FreenetLogBot
[15:26] <nextgens> update-alternatives --config editor
[15:26] * toad_ sets mode -o FreenetLogBot
[15:26] <nextgens> nah
[15:26] * toad_ sets mode +v FreenetLogBot
[15:26] <nextgens> I'll fix it
[15:26] * FreenetLogBot was kicked from #freenet by toad_
[15:26] * FreenetLogBot (n=PircBot@) has joined #freenet
[15:26] * Topic is 'http://freenetproject.org/download.html (990 - mandatory), please read that page before asking for help here. If you want to exchange references, join #freenet-refs | http://wiki.freenetproject.org/FrequentlyAskedQuestions | logs: http://emu.freenetproject.org/irc/ | Tor blocked due to spam (get an op to voice you) | #freenet-fr #freenet-se #freenet-es #freenet-'
[15:26] * Set by toad_ on Fri Dec 01 22:56:34 UTC 2006
[15:26] <nextgens> done
[15:27] <nextgens> toad_> keep quiet :)
[15:27] <toad_> ok...
[15:27] <railk> lol
[15:28] <toad_> nextgens: could you reinject the messages in /root/messages-to-inject ?
[15:28] * nextgens sets mode +o FreenetLogBot
[15:29] <nextgens> what do you mean inject ?
[15:29] <nextgens> they have shunt headers
[15:29] <nextgens> they have been produced by gmane
[15:30] <nextgens> toad_> you can shut down qrunner, copy them back to the shunt folder and restart it
[15:31] * sanity_ (n=ian@) has joined #freenet
[15:32] <nextgens> toad_> is working on binary blobs still on your todo ?
[15:34] <paveq> :D
[15:35] <paveq> so, nextgens, you were talking something about monitoring script?
[15:35] <paveq> and possibility for paveq@freenetproject..?
[15:35] <nextgens> paveq> I've created the mail alias and configured the monitoring script
[15:35] <nextgens> I've used your gmail box
[15:35] <paveq> oh nice
[15:36] <paveq> where did you get my gmail? :)
[15:36] <paveq> probably will show up in google or smthing
[15:36] <paveq> oh
[15:36] <paveq> http://blitzkrieg.homelinux.org/
[15:36] <nextgens> grepping for paveq on the spamassassin whitelist iirc
[15:36] <toad_> nextgens: lots of things are on my TODO :|
[15:37] <toad_> nextgens: they were in the shunt folder, but they weren't in the modqueue
[15:37] * K-roy|DjKaos| (i=K-roy@) has joined #freenet
[15:37] <toad_> nextgens: they didn't show up on the site
[15:37] <paveq> nextgens: what the monitor script does?
[15:37] <toad_> nextgens: i mean on the web moderation interface
[15:37] <toad_> nextgens: also, they are not in the relevant format anyway
[15:37] <paveq> 1730 @ FreenetLogBot >> The monitoring script has detected a network glitch on ftp://ftpmirror.sectoor.de/freenet : it has been disabled
[15:37] <paveq> Hmm.
[15:37] <toad_> nextgens: they were exported with the tool to export messages
[15:37] <toad_> nextgens: so i can't just copy them back
[15:38] <nextgens> hmm
[15:38] <toad_> best thing is probably to reinject them using bin/inject
[15:38] <toad_> some guy on frost shows compuserve and AOL using a 1400 MTU :(
[15:38] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[15:39] <nextgens> they can't use p2p software anyway so it doesn't matter :p
[15:39] <toad_> AOL has provided real IP connectivity for many years now
[15:39] <nextgens> toad_> ok, will inject them then
[15:39] <toad_> compuserve too i would hope
[15:40] <toad_> so do we need to reduce the max packet size to 1400-28?
[15:40] <paveq> doesn't the rfc say 1500 MTU is ok to use?
[15:40] <paveq> if aol is breaking standard...
[15:40] <toad_> not exactly
[15:40] <paveq> thats their fault
[15:40] <toad_> well, right now we assume 1472
[15:41] <toad_> because some PPPoE (DSL) and some (most?) PPPoA is 1472
[15:41] <paveq> yes
[15:41] <nextgens> toad_> they have been injected
[15:41] <toad_> 1476 is GRE and 1492 is PPPoE
[15:41] <paveq> why not actually detect it?
[15:41] <toad_> because it's virtually impossible to detect it from pure java
[15:41] <toad_> we'd have to include yet another bit of native code
[15:41] <paveq> hmmh
[15:41] <toad_> and we'd need a default anyway for when running on systems where the native code doesn't work
[15:42] * toad_ doesn't see why PPPoATM would be 1472... ATM uses 53-byte packets so it could be any size...
[15:43] <paveq> ipv6 over ipv4 over atm :P
[15:44] <paveq> was it like 20% actually data, 80% overhead
[15:44] <nextgens> toad_> they are in the queues .. you have to remove the shunt headers before moderating them
[15:45] * fridim (n=fridim@) Quit ("Quitte")
[15:46] <CIA-14> toad * r11281 /trunk/freenet/src/freenet/io/comm/UdpSocketManager.java: Assume a 1400 byte MTU. Apparently AOL has one. :(
[15:47] <toad_> paveq: the ATM headers are really small
[15:48] * fridim (n=fridim@) has joined #freenet
[15:48] * Ralith (n=ralith@) Quit (Read error: 110 (Connection timed out))
[15:49] <paveq> toad_: least in *nix you could try looking MTU in /proc
[15:49] <paveq> dunno, if its worth it
[15:49] <toad_> paveq: only if you're directly connected
[15:50] <toad_> most people are behind at least one NAT
[15:50] <toad_> probably two
[15:50] <paveq> hmm
[15:50] <paveq> true
[15:50] <nextgens> toad_> that's a bad idea imo
[15:50] <nextgens> it's really small
[15:50] <toad_> nextgens: what's a bad idea? supporting AOL users? why?
[15:51] <nextgens> and ease the fingerprinting of the protocol
[15:51] <nextgens> make it configurable if you wish ... but not default!
[15:51] <toad_> you said don't make it configurable
[15:52] <nextgens> because most users won't configure it
[15:52] <nextgens> and those who will might do silly things
[15:52] <nextgens> but I'm not happy with a default at 1400
[15:53] <paveq> how much this affects performance-wise?
[15:53] <toad_> well we need to support AOL users
[15:54] <toad_> so I'm going to set it to 1400 for now and make it configurable later
[15:54] <nextgens> :(
[15:54] <paveq> what if it tries to send 1500 package over smaller link, does router split it or what
[15:55] <toad_> the router splits it
[15:55] <paveq> assume mtu to router is 1500
[15:55] <toad_> well, that's the best case scenario
[15:55] <Zothar_Work> well, did I not say 1000 at one point? I thought I was joking... :)
[15:55] <toad_> the massive repetition we see when too big packets are sent, resulting in over-use of bandwidth, low payload %, etc, suggests that often fragmentation fails completely
[15:55] <toad_> and the message is just dropped
[15:56] <toad_> i dunno how that happens though
[15:56] <Zothar_Work> paveq: problem is that it's the end to end MTU rather than just the first hop that we'd like to be at or lower than
[15:56] <paveq> yep
[15:56] <nextgens> Zothar_Work> I've fixed viewsvn
[15:56] <Zothar_Work> nextgens: cool, thanks
[15:57] <paveq> of course peers could change info about their mtu's
[15:57] <paveq> and use lowest common
[15:57] * nextgens goes back working on his AI project
[15:57] <Zothar_Work> well, the path MTU will be per connection rather than common to the whole node
[15:57] <nextgens> mail me if something else needs fixing
[15:58] <toad_> ok
[15:58] * Ralith (n=ralith@) has joined #freenet
[15:59] <paveq> hmm why is aol using 1400
[15:59] <paveq> doesn't make any sense to me
[15:59] <toad_> AOL in germany use 1400 or 1454 according to frost
[15:59] <toad_> compuserve germany use 1400
[15:59] <toad_> lots us 1454
[16:01] <paveq> "The AOL network runs at an MTU of 1450. "
[16:01] <paveq> http://info.aol.co.uk/broadband/faqHomeNetworking.adp
[16:02] <paveq> some talks about 1362..
[16:02] <paveq> :/
[16:06] <nextgens> http://zgp.org/pipermail/p2p-hackers/2005-May/002649.html
[16:07] <toad_> nextgens: :|
[16:09] <Zothar_Work> TCP/IP stacks should have a "what's the path MTU to that IP" interface :)
[16:09] <paveq> could it be discovered by just sending some package with specific mtu and receiving end detect if it was fragmented
[16:10] <toad_> no, they should get rid of fragmentation :)
[16:10] <toad_> IPv6 does this - fragmentation becomes a link level matter
[16:10] <Zothar_Work> paveq: fragments are at a lower level than we're at
[16:10] <paveq> thats what I thought too
[16:10] <paveq> maybe some way to hack :P
[16:11] <toad_> when it has to go over an ethernet, it's split up into ethernet sized packets, but then it's reassembled before it reaches the next hop
[16:11] <Zothar_Work> ...so we can't detect them directly (AFAIK UDP fragments are reconstructed before bumped up)
[16:11] <toad_> obviously this produces some extra latency, but it simplifies matters for the user no end
[16:11] <nextgens> toad_> we could store a per-peer MTU and tweak it on the fly based on the packetloss
[16:11] <Zothar_Work> without native, we'd have to send packets with differeing MTUs until we saw a noticable performance difference and assume we found the path MTU
[16:11] <toad_> yes and packet loss can also be caused by congestion
[16:12] <toad_> what we need to do is to be able to send packets with Dont Fragment set
[16:12] <paveq> Zothar_Work: doesn't sound very reliable way
[16:12] <toad_> then there are algorithms for picking up the MTU, even without ICMP feedback
[16:12] <Zothar_Work> paveq: I didn't say it was :)
[16:12] <toad_> but the only way to do that is to call out to some external program
[16:12] <toad_> ping, for example
[16:12] <paveq> hmmh
[16:13] <paveq> how does example azureus solve this
[16:13] <paveq> its completely java afaik
[16:14] <toad_> probably by setting an arbitrary limit, like we do
[16:14] <paveq> dunno, would it be any easier with other than java anyways?
[16:14] <toad_> if they want to go to more effort, they'll make it configurable and exchange it on connection setup
[16:17] <toad_> nextgens: hmmm, you didn't delete te header at the top
[16:27] * gruffy (n=daniel@) has joined #freenet
[16:37] * Fennes (i=1234@) Quit ("But in the end the prince is and will always be just a prince")
[16:40] <toad_> okay, the OOM was caused by me running an older version of the plugin ...
[16:45] <Ralith> gah
[16:45] <Ralith> I'm getting 'connection refused' on fproxy
[16:46] <Ralith> nevermind
[16:49] * railk (n=railk@) Quit ("Cya, wouldn't want ta be ya!")
[16:51] * cantthinkofaname (n=witchdoc@) has joined #freenet
[16:56] * cantthinkofaname (n=witchdoc@) Quit ("BitchX: the quilted quicker picker upper")
[16:56] * railk (n=railk@) has joined #freenet
[16:56] * NullAcht15 (n=NullAcht@) has joined #freenet
[16:59] * Ralith (n=ralith@) Quit (Read error: 104 (Connection reset by peer))
[17:09] * Fennes (i=1234@) has joined #freenet
[17:10] * Ralith (n=ralith@) has joined #freenet
[17:12] * K-roy|DjKaos| (i=K-roy@) Quit ()
[17:14] * ralith_ (n=ralith@) has joined #freenet
[17:15] * Ralith (n=ralith@) Quit (Nick collision from services.)
[17:15] * ralith_ is now known as Ralith
[17:16] * K-roy|DjKaos| (i=K-roy@) has joined #freenet
[17:19] * Urs_ShPo (n=gaim@) Quit ("Leaving.")
[17:25] * OctobersDark (n=October@) has joined #freenet
[17:25] * K-roy|DjKaos|2 (i=K-roy@) has joined #freenet
[17:31] * timmy2chk (n=violent@) Quit ("so what?")
[17:37] * K-roy|DjKaos| (i=K-roy@) Quit (Read error: 110 (Connection timed out))
[17:42] * gruffy (n=daniel@) Quit ("Ex-Chat")
[18:07] * bobdufour (n=louis@) has joined #freenet
[18:13] * tetaworx (n=freenet@) has joined #freenet
[18:32] * Xenith_ (i=cronus@) has joined #freenet
[18:59] * Ash-Fox (i=UNKNOWN@) Quit ("Do not meddle in the affairs of kitsune, for you are crunchy and good with ketchup.")
[18:59] * IMCensored1 (n=KMIntern@) Quit ("Peace and Protection 4.22")
[19:02] * nineus (n=ninus@) has joined #freenet
[19:02] <nineus> hey - willing to help me a bit ?
[19:03] <toad_> hi
[19:03] <toad_> problems?
[19:03] <nineus> sort of
[19:03] <nineus> trying to get Freenet to work on gentoo 64 bit
[19:04] * nextgens bets nineus is using the 3 years old ebuild
[19:06] <nextgens> toad_> so; what are you working on ? Librarian, RSS or binary blobs ?
[19:06] * nineus probably agrees with nextgens
[19:07] <nineus> in other news - perhaps nextgens will share some of his epic wisdom :)
[19:08] * Ash-Fox (i=UNKNOWN@) has joined #FreeNET
[19:10] <nextgens> http://bugs.gentoo.org/show_bug.cgi?id=32779
[19:10] <nextgens> I suggest you forget about it :)
[19:10] <nextgens> http://bugs.gentoo.org/show_bug.cgi?id=128783
[19:10] <nextgens> nineus> the ebuild is setting up freenet 0.5 : we are currently using freenet 0.7
[19:11] <nineus> yes - can i run it without the ebuild ?
[19:11] <nineus> eg just the files ?
[19:11] <nextgens> nineus> I suggest you follow instructions on http://freenetproject.org/download.html and forget about the ebuild
[19:12] <nextgens> there is a tarball available
[19:12] <nineus> yeh
[19:12] <nineus> but its complaining about my 1.4 jvm
[19:12] <nextgens> huh ?
[19:12] <nextgens> freenet is fully 1.4 compatible
[19:12] * RAKER (n=a@) has joined #freenet
[19:12] * K-roy|DjKaos|2 (i=K-roy@) Quit ()
[19:13] <nextgens> except one plugin : jstun
[19:13] <Zothar_Work> don't use the JSTUN plugin and you'll probably be OK
[19:13] <nineus> ok - hmm, brb
[19:15] <nineus> what does jstun do anyway ?
[19:16] <nineus> java.lang.RuntimeException: 1.4 based 64 bit JVM not supported
[19:17] <Zothar_Work> JSTUN detects your external IP address by consulting 3rd party servers and checks things like firewallas and NATs and such
[19:17] <nineus> ahh, ok
[19:17] <Zothar_Work> ah, then you gotta go to a 1.5 JVM or run it 32 bit
[19:17] <nineus> damn
[19:17] * green_qween (n=losergrl@) has joined #Freenet
[19:17] * Yukishiro (n=zhaan@) has joined #freenet
[19:18] <toad_> nextgens: fixing the "plugins disappear on restart" bug
[19:18] <nineus> woohoo
[19:18] <nineus> it runs
[19:19] <toad_> isn't the solution to "there are two plugin systems" simply to delete the one that isn't used?
[19:19] <nineus> just had to switch to sun-1.5 jre ... *sigh*
[19:20] * IMCensored1 (n=KMIntern@) has joined #freenet
[19:20] <paveq> nineus: gentoo can handle both 1.4 and 1.5 at the same time
[19:20] <paveq> read the java docs
[19:20] <paveq> you gotta set which vm to use
[19:21] <nineus> well - it runs
[19:21] <nineus> changed the path to include the sun jre, and bang...
[19:22] <nineus> doesn't seem like it has problems with jstun either
[19:22] <CIA-14> toad * r11282 /trunk/freenet/src/freenet/config/Option.java: Document method.
[19:23] <Zothar_Work> toad_: they're both used by different stuff is the problem I believe
[19:23] <CIA-14> toad * r11283 /trunk/freenet/src/freenet/node/Node.java:
[19:23] <CIA-14> We weren't checking aggressiveGCModificator.
[19:23] <CIA-14> (Maybe we should remove it?)
[19:24] * green_qween (n=losergrl@) has left #Freenet
[19:25] <CIA-14> toad * r11284 /trunk/freenet/src/freenet/plugin/PluginManager.java:
[19:25] <CIA-14> Plugins: (plugin/)
[19:25] <CIA-14> Catch Throwable loading plugin.
[19:25] <CIA-14> Don't queue starting plugin until it's been added.
[19:25] <CIA-14> Don't write freenet.ini until plugin has been loaded.
[19:25] <CIA-14> Logging.
[19:26] <CIA-14> toad * r11285 /trunk/freenet/src/freenet/pluginmanager/PluginManager.java: Minor cleanup
[19:27] <CIA-14> toad * r11286 /trunk/freenet/src/freenet/config/ (StringArrOption.java SubConfig.java):
[19:27] <CIA-14> Fix isDefault() - fixes "plugins lost on restart" bug.
[19:27] <CIA-14> Don't try to load a plugin named "" if none are specified.
[19:27] <CIA-14> Logging.
[19:30] <toad_> hmm, in fact, that may have caused other settings to get lost too ...
[19:39] <zorton> wow, all my freenet ref nodes have gone "busy"
[19:40] * lattt (n=haha@) has joined #freenet
[19:40] <CIA-14> toad * r11287 /trunk/plugins/Librarian/Librarian.java: Librarian: Return results in order.
[19:41] * whiterabbit (n=Whiterab@) has joined #freenet
[19:41] * toad_ waits for emu to regenerate the jar ...
[19:42] <lattt> hi, I've seen that someone asked a few days if he should kick a node who won't update from 999 ... I think it was answered that he should wait?! ... so kick him/her or wait ... asked for upgrade 2 days ago but no answer, but he was online for at least 2 times since them
[19:42] <toad_> lattt: I'd wait
[19:43] <lattt> how long? he is a newbie because I got his ref from #freenet-refs maybe 1-2 weeks ago
[19:43] <lattt> newbie like me .. but I've done the update :)
[19:43] <toad_> :)
[19:44] <toad_> i dunno
[19:44] <Zothar_Work> something like 1003 I think it is will be mandatory a week from tomorrow
[19:44] <toad_> i wouldn't worry about it until his node is actually blocked
[19:45] <lattt> ah I forgot that ... is anywhere stated what the last revision is which will work?
[19:46] <Zothar_Work> I think it's mentioned in the build release announcment about the mandatory
[19:46] <Zothar_Work> once 1007 comes out, I believe it'll be making 1000 mandatory
[19:47] <zorton> Zothar_Work: how goes the work on your bot? bot2bot coming along?
[19:47] <lattt> hm, ok then I'll subscribe the mails if I haven't done it already, thx
[19:48] <Zothar_Work> zorton: it's coming; I did a bunch of hacking on it last night; I might have it doing bot2bot ref trading sometime in the next week or two; I haven't decided if I'm going to implement cooperative bot announcements first though as I see the ref trading could increase bot usage and otherwise clutter the channel
[19:49] <zorton> you could knock down channel clutter by having the bot2bot traffic in a seperate channel...
[19:49] <zorton> might make things easier to code too
[19:49] <zorton> #freenet-nodebot or something
[19:49] <Zothar_Work> zorton: the bot2bot traffic is all PMs; I'm talking about the number of bots in the channel, which are there because they will also be interacting with humans (for now)
[19:50] <Zothar_Work> (and last I looked, there were some architectural changes I might need to make before I could do multi-channel)
[19:52] <RAKER> does anyone know of a tutorial for freenet, or would like to help me get it configured?
[19:54] <zorton> Zothar_Work: k, I got ya. I wouldn't worry so much about the number of bots in channel. I don't know many people who care much about how big their /names list is
[19:55] <zorton> whiterabbit: msg it help for nfo
[19:55] <zorton> wrong chan :)
[19:56] * lattt (n=haha@) Quit ("Verlassend")
[19:58] * hagman (n=hstahlm@) has joined #freenet
[19:58] <RAKER> why are all my peers "never connected"
[20:01] <Xenith_> cuz they have to want to connect to you too
[20:01] * nineus (n=ninus@) Quit (Remote closed the connection)
[20:04] * IMCensored1 is now known as IMCensored_
[20:05] * IMCensored_ is now known as IMCensored
[20:06] * IMCensored is now known as IMCensored1
[20:09] * hagman (n=hstahlm@) Quit ("Leaving")
[20:09] * toad_ hmmm
[20:10] <toad_> any dangerous or especially heavy operation should be done via a POST form...
[20:10] <toad_> anything we don't want to happen from outside should be confirmed via a hidden parameter
[20:11] <toad_> but that applies equally to accessing plugins from freesites and from external web sites
[20:11] <toad_> and right now we don't support POST forms for plugins ...
[20:12] <Zothar_Work> zorton: well the reason the number of bots matters without cooperative bot announce is because the more bots there are the larger percentage of the channel traffic will be various bots greeting the channel every 20-30 minutes per bot
[20:12] <toad_> so maybe we can just pass through /plugins/ GET forms and direct links?
[20:13] <toad_> external sites can already link to / inline plugins running locally
[20:15] <Zothar_Work> RAKER: both sides have to add each other before you get a peer connection
[20:15] <sleon> nextgens: ping
[20:22] <RAKER> Zothar_Work: ok, how do get them to add me now?
[20:23] * IMCensored1 (n=KMIntern@) Quit ("K-LINED kornbluth.freenode.net (I'm the shit, so imma ban myself for a while)")
[20:23] <Zothar_Work> RAKER: ask for anyone wanting to trade refs on #freenet-refs
[20:23] <sleon> RAKER: go to #freenet-refs and talk to ppl you want to connect with and provide them with your reference
[20:25] * hagman (n=hstahlm@) has joined #freenet
[20:28] * IMCensored1 (n=KMIntern@) has joined #freenet
[20:29] <hagman> hi out there. anybody to exchange a ref? http://dark-code.bulix.org/otdu6s-25394?raw
[20:30] <Zothar_Work> hagman: see #freenet-refs
[20:31] <hagman> Zothar_Work: Thanks
[20:31] * Apophis2_ (n=Apophis@) has joined #freenet
[20:31] * MikeW (i=Mike@) has joined #freenet
[20:37] * Apophis2 (n=Apophis@) has joined #freenet
[20:38] * Tritman (i=Tritman@) Quit (Remote closed the connection)
[20:39] * Tritman (i=Tritman@) has joined #freenet
[20:48] * xxxxx (n=Apophis@) Quit (Read error: 110 (Connection timed out))
[20:49] <CIA-14> toad * r11288 /trunk/freenet/src/freenet/clients/http/filter/HTMLFilter.java: No target attribute on <form>.
[20:53] * Apophis2_ (n=Apophis@) Quit (Read error: 110 (Connection timed out))
[20:58] * railk_ (n=railk@) has joined #freenet
[20:59] * _ph00 (n=z@) Quit ("Leaving")
[21:08] <CIA-14> toad * r11289 /trunk/freenet/src/freenet/clients/http/filter/ (4 files):
[21:08] <CIA-14> Move the decision on whether a <form> is allowed to the filter callback, which is where it should be.
[21:08] <CIA-14> Force the enctype and accept-encoding.
[21:08] <CIA-14> Drop support for accept until we get around to filtering it properly.
[21:10] * whiterabbit (n=Whiterab@) Quit ("KVIrc 3.2.5 Anomalies http://www.kvirc.net/")
[21:14] * railk (n=railk@) Quit (Read error: 110 (Connection timed out))
[21:24] <CIA-14> toad * r11290 /trunk/freenet/src/freenet/clients/http/filter/ (3 files):
[21:24] <CIA-14> Allow freesites to post to plugins.
[21:24] <CIA-14> External sites can do this, after all.. so it's not a great security risk.
[21:24] <CIA-14> Any dangerous operations should 1) be POSTs and 2) be confirmed via the formPassword (which isn't known to either external sites or freesite authors).
[21:36] * jamie (n=jamie@) has joined #freenet
[21:36] * Yukishiro (n=zhaan@) Quit ("Chatzilla 0.9.75 [Firefox 1.5.0.8/2006102516]")
[21:37] <CIA-14> toad * r11291 /trunk/plugins/Librarian/Librarian.java:
[21:37] <CIA-14> Fix ClassCastException.
[21:37] <CIA-14> Follow newURI's.
[21:38] <CIA-14> toad * r11292 /trunk/plugins/Librarian/Librarian.java: Doh
[21:39] * MikeW (i=Mike@) Quit ()
[21:42] * railk_ is now known as railk
[21:42] <CIA-14> toad * r11293 /trunk/plugins/Librarian/Librarian.java: Another HashSet
[21:42] * hjubal (n=hjubal@) has joined #freenet
[21:44] <toad_> FreenetLogBot: oi!
[21:49] * green_qween (n=losergrl@) has joined #freenet
[21:50] * greycat (i=rfc1413@) Quit ("This time the bullet cold rocked ya / A yellow ribbon instead of a swastika")
[21:51] <toad_> hmmm
[21:51] <toad_> emu appears to be recompiling the old version of Librarian each time
[21:52] * Ralith (n=ralith@) Quit (Remote closed the connection)
[21:53] * xyz (n=xyz@) has joined #freenet
[21:53] * xyz (n=xyz@) has left #freenet
[22:03] <CIA-14> toad * r11294 /trunk/plugins/Librarian/Librarian.java: Add class= to relevant elements to allow for CSS later on.
[22:03] * hagman (n=hstahlm@) Quit (Read error: 110 (Connection timed out))
[22:04] <toad_> FreenetLogBot: hello?
[22:07] <CIA-14> toad * r11295 /trunk/plugins/Librarian/Librarian.java: Return results in order. Hopefully.
[22:08] <CIA-14> toad * r11296 /trunk/plugins/Librarian/Librarian.java: Doh
[22:27] <CIA-14> toad * r11297 /trunk/plugins/Librarian/Librarian.java: Allow an external stylesheet to be specified in the form.
[22:27] * ShipHead (i=ShipHead@) has joined #freenet
[22:29] <CIA-14> toad * r11298 /trunk/plugins/Librarian/Librarian.java: - not . in class names
[22:33] * bobdufour (n=louis@) Quit ("Ex-Chat")
[22:34] * fader (n=segfault@) Quit (Nick collision from services.)
[22:34] * fader_ (n=segfault@) has joined #freenet
[22:34] * fader_ is now known as fader
[22:39] <CIA-14> toad * r11299 /trunk/freenet/src/freenet/ (node/NodeClientCore.java pluginmanager/PluginRespirator.java): New support code for plugins: Allow them to create a [GenericRead]FilterCallback.
[22:40] * _ph00 (n=z@) has joined #freenet
[22:40] <toad_> FreenetLogBot: hey
[22:40] <toad_> FreenetLogBot: hey
[22:41] <CIA-14> toad * r11300 /trunk/plugins/Librarian/Librarian.java: Force recompile of plugin
[22:42] <toad_> FreenetLogBot: you there?
[22:50] * Xenith_ (i=cronus@) Quit (Read error: 104 (Connection reset by peer))
[22:53] * green_qween (n=losergrl@) Quit ()
[22:53] * PraiseChaos (n=kcecil@) Quit (Remote closed the connection)
[22:53] * makomk (n=aidan@) Quit (Read error: 110 (Connection timed out))
[23:04] * railk (n=railk@) Quit (Remote closed the connection)
[23:05] * mk_ (n=mk@) has joined #freenet
[23:05] <CIA-14> toad * r11301 /trunk/freenet/src/freenet/ (3 files in 3 dirs): Make stylesheet parameter work.
[23:05] <mk_> how will premix routing be implemented without making the darknet harvestable?
[23:08] <toad_> bbiab
[23:08] <nextgens> mk_> part of the topology will be shown
[23:09] <mk_> how will only part be shown withot making it harvestable?
[23:09] <nextgens> we haven't implemented that yet ... don't know
[23:09] <nextgens> but we don't need to publish the whole topology to set up a mixnet
[23:11] <mk_> that's true, but you do need to publish some of it... I'm wondering how you can publish some of it yet ensure that the entire thing doesn't become available
[23:13] <nextgens> hmm
[23:14] <nextgens> we will probably use in between node messenging system
[23:14] <mk_> what do you mean?
[23:14] <nextgens> and "broadcast" to our peers some of our links ... including some kind of "hops to live"
[23:15] <mk_> I don't think I understand
[23:15] * nextgens is almost sure he isn't clear enough
[23:16] <nextgens> :)
[23:16] <nextgens> well
[23:16] <nextgens> each peer will disclose peers it is aware of ... including a counter set to some value
[23:17] <nextgens> each time the message is going through a neighbor, the counter will be deacreased
[23:17] <mk_> so like sharing your friends,
[23:17] <nextgens> hence at some point some peers will disapear from the "announce"
[23:17] <mk_> (with your friends) but also sharing the friends that have been shared with you
[23:17] <nextgens> hence it won't publish the whole topology
[23:18] <mk_> but differentiating between the two, so that you don't share 3rd hop friends or somesuch
[23:19] <nextgens> we will take care not to disclose who are the "real" neighbors of peer in question
[23:19] <mk_> one immediate problem I see with that is for every 'covert agent' or traitor, you can harvest 1000 nodes with 3 hops and 10 peers
[23:20] <nextgens> yes, that's likely
[23:20] <mk_> if you decrease that, then you decrease the pool, which makes you less safe
[23:21] <nextgens> it's even worst than that
[23:21] <nextgens> we will be building an untrusted premix network
[23:21] <mk_> what do you mean?
[23:21] <nextgens> ie: we don't have exit points and we can't ensure what their pubkey is
[23:21] <mk_> right. there's no way to verify it
[23:22] <mk_> one of the biggest problems here is that the darknet is untrusted from the start
[23:22] <nextgens> not really ...
[23:22] <nextgens> we do trust our peers
[23:23] <mk_> do we?
[23:23] <nextgens> to some extend at least
[23:23] <mk_> the reason I say that it's untrusted is because right now people are getting refs from a public location
[23:23] <nextgens> we don't trust them for not snooping at our requests
[23:23] <nextgens> but we will have to trust them
[23:24] <mk_> if you only connected to your real-life friends, then I would say that it is trusted
[23:24] <mk_> but I don't think that freenet is anywhere near that widespread for this to be practical
[23:24] <nextgens> :|
[23:25] <nextgens> you're probably right
[23:25] <Zothar_Work> as far as anonymity, our peers are our most trusted of all nodes on Freenet
[23:25] <mk_> yes, but that's only when we're connecting to people we've established trusting relationships with
[23:26] <nextgens> sure, but if we are implementing premix routing, it's to prevent them from doing statistical attacks on us
[23:26] <Zothar_Work> I'm just saying that the design assumes they're the most trusted; unfortunately, it doesn't say anything about how much we trust them
[23:26] * NullAcht15 (n=NullAcht@) Quit (Remote closed the connection)
[23:26] <nextgens> mk_> anyway, premix routing will be hard to implement on freenet : and we are not anywhere near having solved all the problems related to its usage yet ... hence it's planned for 0.8 :)
[23:27] <mk_> well, right, but what I'm getting at is that there's a technical problem (the entire mechanism *assuming* we have trusted friends), and then the social problem of actually having 20 trusted friends
[23:27] <nextgens> mk_> I do have trusted peers ... most of them are real IRL friends
[23:27] <nextgens> others are freenet developpers
[23:27] <mk_> how many people are there like you?
[23:28] <nextgens> arguably my connection set isn't representative
[23:28] <mk_> I wouldn't think that it was if we have a channel set up to publicly exchange refs
[23:28] <sleon> nextgens: i have a question to grsecurity
[23:29] <nextgens> sleon> go ahead
[23:29] <sleon> nextgens: can i forbit to root access to a specified file using rbac?
[23:29] <nextgens> sure
[23:29] <sleon> nextgens: it should be also not be possible by any means except rebooting to non rbac mode
[23:29] <mk_> also, you said most... does that mean that you have a peer that you don't trust that well? because it only takes one
[23:29] <sleon> to trick it, (sudo su <other user> and so on)
[23:30] <sleon> nextgens: did you also recompile kernel with grsecurity patch for debian?
[23:30] <nextgens> mk_> no, it doesn't ... even if one of your peers is evil, it won't be able to harm so much
[23:31] <nextgens> mk_> it has to either know to whom you're connected to or to force you to grab some specific content in order to be really harmfull
[23:31] <mk_> nextgens, really? what about correlation attacks on splitfiles?
[23:32] <sleon> nextgens: and also how big is performance hit for io operations compared to normal debian kernel?
[23:32] <nextgens> if you've got enough peers and the overall content on freenet is large enough, the probability they can do something is low
[23:32] * jamie (n=jamie@) Quit ("Leaving")
[23:32] <nextgens> on the other hand, if they manage to know to whom you're connected to and they DoS most of them, you're caught
[23:33] <nextgens> sleon> I don't use debian kernels at all