Timestamps are in GMT/BST.
[0:03] * Zothar (n=Zothar@) Quit ("Chatzilla 0.9.74 [Firefox 1.5.0.7/2006090918]")
[0:06] <SinnerG> hmm
[0:06] <SinnerG> paypal has some nice stuff :)
[0:06] <SinnerG> they have a sandbox to play on :p
[0:07] * Cheri (n=Anonymou@) has joined #freenet
[0:15] * sanity (n=ian@) Quit ()
[0:17] * Werdna (n=Andrew@) Quit (Connection reset by peer)
[0:17] * Werdna_ (n=Andrew@) has joined #freenet
[0:26] * Cheri (n=Anonymou@) Quit ()
[0:27] * sim00 (i=MyIdent@) has joined #freenet
[0:28] <sim00> tor
[0:28] <sim00> tor gateway
[0:29] <sim00> zxczzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
[0:29] * raekwon (n=raekwon@) has joined #freenet
[0:40] * Werdna_ (n=Andrew@) Quit (Connection timed out)
[0:45] * NullAcht15 (n=NullAcht@) Quit (Remote closed the connection)
[0:46] * jimmmmm (n=james@) has joined #freenet
[0:46] <jimmmmm> hi, anyone want to exchange noderefs?
[0:51] * jimmmmm (n=james@) Quit ("Leaving")
[0:51] * mozillaman (n=chatzill@) Quit (Read error: 131 (Connection reset by peer))
[1:15] * Turakamu (n=Steve@) has joined #freenet
[1:17] * usablepi (n=usablepi@) has joined #freenet
[1:20] * usablepi (n=usablepi@) has left #freenet
[1:20] * usablepi (n=usablepi@) has joined #freenet
[1:21] * usablepi is now known as UsablePI
[1:32] * UsablePI (n=usablepi@) Quit ()
[1:34] * UsablePI (n=UsablePI@) has joined #freenet
[1:35] * UsablePI (n=UsablePI@) Quit (Client Quit)
[2:12] * Bombe (n=bombe@) Quit (Read error: 60 (Operation timed out))
[2:17] * tigotigor (i=NoNoBrut@) has joined #freenet
[2:17] * Gasi__ (n=chatzill@) has joined #freenet
[2:17] * Gasi__ is now known as Gasi
[2:33] * Gasi_ (n=chatzill@) Quit (Read error: 110 (Connection timed out))
[2:48] * Turakamu (n=Steve@) Quit ("Leaving")
[3:08] * raekwon (n=raekwon@) Quit ("leaving")
[3:14] * railk (n=railk@) Quit (Remote closed the connection)
[3:51] * dermoth (n=dermoth@) has joined #freenet
[4:12] * TheBishop_ (n=bishop@) has left #freenet
[4:19] * Skiver (n=Skiver@) has joined #freenet
[4:24] * tigotigor (i=NoNoBrut@) Quit ()
[4:50] * JustMe_ (i=JustMe_@) Quit ()
[4:50] * MakeSpace (n=jpp@) has joined #freenet
[5:09] * sandos (n=sandos@) Quit (Read error: 145 (Connection timed out))
[5:23] * noggly_ (i=noggly@) has joined #freenet
[5:26] * Werdna (n=Andrew@) has joined #freenet
[5:33] * noggly (i=noggly@) Quit (Read error: 110 (Connection timed out))
[5:43] * sanity (n=ian@) has joined #freenet
[5:43] * ChanServ sets mode +o sanity
[5:47] * sandos (n=sandos@) has joined #freenet
[5:51] * sanity (n=ian@) Quit ()
[6:31] * sim00 (i=MyIdent@) Quit (Read error: 104 (Connection reset by peer))
[6:35] * sanity (n=ian@) has joined #freenet
[6:35] * ChanServ sets mode +o sanity
[6:42] * sanity (n=ian@) Quit ()
[6:47] * sanity (n=ian@) has joined #freenet
[6:47] * ChanServ sets mode +o sanity
[6:50] * sanity (n=ian@) Quit (Client Quit)
[7:01] * Bombe (n=bombe@) has joined #freenet
[7:01] * ChanServ sets mode +o Bombe
[7:27] * ljn1981 (i=LJN@) Quit (Remote closed the connection)
[7:30] * dermoth (n=dermoth@) Quit ("Ex-Chat")
[7:39] * oCc (n=nodeOne@) has joined #freenet
[7:41] * hjubal (n=hjubal@) has joined #freenet
[7:42] <hjubal> hi
[7:52] * QshelTier (n=bombe@) has joined #freenet
[7:52] * ChanServ sets mode +o QshelTier
[7:53] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[8:03] * QshelTier is now known as Bombe
[8:22] * ph00_away is now known as _ph00
[8:28] * MakeSpace (n=jpp@) Quit ("using sirc version 2.211+KSIRC/1.3.12")
[8:33] * MakeSpace (n=jpp@) has joined #freenet
[8:53] * hdp (n=hdp@) has joined #freenet
[9:07] * Zincoid (i=Zincoid@) has joined #freenet
[9:10] * grapjeg (n=fred@) has joined #freenet
[9:12] <grapjeg> Help. How to use FreMulET? one guide?
[9:23] <nextgens> you can try on the "freemule" board of frost
[9:24] <SinnerG> lot of activity here ;p
[9:24] * phrosty (n=phrosty@) Quit ("baseball is wrong: man with four balls cannot walk.")
[9:24] <grapjeg> okay
[9:26] <nextgens> SinnerG> I'm helping him on the french channel
[9:26] <nextgens> btw, freemule sucks
[9:26] <nextgens> :p
[9:27] <Bombe> Hehe...
[9:34] * Tritman (i=Tritman@) Quit (Client Quit)
[9:35] * Tritman (i=Tritman@) has joined #freenet
[9:35] * ljn1981 (i=LJN@) has joined #freenet
[9:36] <ljn1981> Morning
[9:44] * cbreak (n=cbreak@) has joined #freenet
[9:46] * cbreak (n=cbreak@) Quit (Client Quit)
[9:47] * grapjeg (n=fred@) Quit (Remote closed the connection)
[9:49] * dl1bku (n=dl1bku@) has joined #freenet
[9:49] <SinnerG> stupid suphp.. I had to alter my script cause of it ;p
[9:49] <SinnerG> (aka add the index.php to urls pff ;p)
[9:56] * nextgens sets mode +v Tritman
[9:56] <Tritman> thx
[9:57] <Tritman> nextgens, thx but what is the best way with tor ?
[9:57] <nextgens> ask a staffer to get a host cloack
[9:57] <nextgens> you can reach them using /stats p
[9:57] <_ph00> I tried that, I didn't even get an answer...
[9:58] <nextgens> well, ljn1981 has succeeded, ask him how to do :p
[9:58] <_ph00> no, I tried asking on #freeneode and on #freenode-social. not using /stats p
[9:58] <_ph00> hm. OK
[9:58] <ljn1981> Heh
[9:59] <_ph00> so?
[9:59] <_ph00> how did you do that
[9:59] <ljn1981> I was around and asked the guy who suggested it as a solution when the ban was set up
[9:59] * NullAcht15 (n=NullAcht@) has joined #freenet
[10:02] <nextgens> it's written in motd
[10:02] <nextgens> anyway the ban will remain on #freenet-refs
[10:03] <nextgens> as it's silly to come there and exchange references in the channel if you're going there though tor
[10:05] <_ph00> I don't usually swap refs on #freenet-refs but I'm often there to help ppl: most noob questions get asked there
[10:05] <_ph00> are
[10:06] <nextgens> they shouldn't be :)
[10:08] * sandos_ (n=sandos@) has joined #freenet
[10:09] <_ph00> I know. but they are anyway. and when I can answer, it's easier to do that, than to say 'ask that on #freenet'
[10:09] <Tritman> the solution if I leave and come again is to use "/stats p" command and wait for a cloack from a staff man, true ?
[10:10] <_ph00> how do I use /stats p exactly? I type /stats p ? that's all? what then?
[10:11] <ljn1981> _ph00: How about aswering there and asking them to ask future questions in here, on the mailing lists or on relevant frost boards?
[10:12] <_ph00> ok
[10:12] <ljn1981> Just an idea
[10:13] <_ph00> adding 'next time ask question on #freenet'
[10:13] <_ph00> why not
[10:13] <MakeSpace> I have a semi-technical question. Is it possible to run two different freenet nodes on one machine? I would like to start building up a node with ref's I really can trust.
[10:13] <ljn1981> Sure if you got the resources for it.
[10:14] <ljn1981> You just need to change the port numbers of one of the installations
[10:14] <_ph00> I have these two peers, right, they're supposed to be 24/7 nodes, but they've been down fom more than 15 hrs one and some 7 hrs the other.
[10:14] <_ph00> you don't
[10:14] <_ph00> it will do it
[10:14] <_ph00> automatically
[10:14] <MakeSpace> resource == cpu, mem, (net bandwidth) and disk, right?
[10:14] <ljn1981> I think only one of them can be made to start automatically on windows but nextgens would konw more about that
[10:14] <_ph00> the second node's fproxy port will be 8889
[10:15] <_ph00> and the third node's fproxy port will be 9999 I think
[10:15] <MakeSpace> cool! auto magic!!
[10:15] <_ph00> I did that, it worked fine
[10:15] <ljn1981> Does that work for FCP too?
[10:16] <_ph00> and, I got *four* nodes running on an old crappy machine (1600MhZ, 512 RAM)
[10:16] * sandos__ (n=sandos@) has joined #freenet
[10:16] <ljn1981> MakeSpace: Resource. Yes.
[10:16] <_ph00> I don't really know about fcp, I didn't test thet
[10:16] <ljn1981> _ph00: I highly doubt that ran well at all.
[10:17] <_ph00> but it's simple enough, if you have the first one running and the second won't connect, then you have to manually change port. not so difficult
[10:18] <_ph00> I had cpu a 100% steady and launching an app would take several minutes... when it didn't crash on launch. but it did work
[10:18] <_ph00> anyway, two nodes ran almost fine
[10:18] <_ph00> but that's too much anyway for my machine, even overclocking to 2000
[10:18] <_ph00> actually, my box is kinda bad even for one node...
[10:18] * Zincoid (i=Zincoid@) has left #freenet
[10:19] <_ph00> but ok, one node runs fairly well
[10:19] <MakeSpace> I have lots of memory and disk. Fred runs 'nice'd on linux, so cpu shouldn't be a real issue...
[10:19] <_ph00> I'm gonna invest some 400 euros and upgrade some harware: I'm gonna go for a pentium 3200
[10:19] <_ph00> you need mem as well
[10:19] <MakeSpace> 4gb
[10:20] <_ph00> memory, ok
[10:20] <_ph00> you said that
[10:20] <_ph00> sorry
[10:20] <MakeSpace> np
[10:20] <_ph00> woah. that's enough to run some 6 nodes...
[10:21] <MakeSpace> i run linux with no swap, so I would probably keep it to two ;-)
[10:21] <_ph00> ok then, simply run the installer in another dir, and it will auto-set the new fproxy port to 8889
[10:21] <MakeSpace> I'll go try it...
[10:21] * sandos_ (n=sandos@) Quit (Read error: 60 (Operation timed out))
[10:21] <_ph00> you don't really need swap with 4 gig mem
[10:22] * sandos (n=sandos@) Quit (Connection timed out)
[10:22] <MakeSpace> swap == lower security. that's why I went crazy on memory and took a fairly cheap MB
[10:24] <Griffon26> _ph00: that depends on how you use your system. If you have apps that use a lot of memory but are not accessing it for longer periods of time, you'd rather have that in swap so you can use the memory for caching stuff that you _are_ working with.
[10:24] <_ph00> use the line metadata.allowLocalAddresses=true in the refs when connecting the nodes to each other
[10:26] <MakeSpace> griffon26: only if you don't mind leaving copies of your data laying around at random places in the swap partition.
[10:26] <Griffon26> guys, do I understand correctly that if bwLimitDelayTime and pInstantReject are usually about 0, I should be adding more peers?
[10:26] <Griffon26> MakeSpace: right
[10:27] <Griffon26> MakeSpace: if you're that paranoid, the performance loss is negligible compared to the overhead in accessing your encrypted file systems. ;)
[10:27] <MakeSpace> Can not bind fproxy to 8888: let's try 8889 insteed.
[10:27] <MakeSpace> Can not bind fcp to 9481: force it to 9482 insteed.
[10:27] <MakeSpace> Can not bind console to 2323: force it to 2324 insteed.
[10:27] <MakeSpace> magic!!
[10:28] <Griffon26> MakeSpace: running multiple instances?
[10:28] <MakeSpace> yeah, I'll try an encrypted swap space -- if I ever need to swap.
[10:28] <Griffon26> netstat's your friend
[10:28] <MakeSpace> I am trying to start a second instance
[10:28] <Griffon26> or at least, in linux it is
[10:28] <MakeSpace> linux is my friend
[10:29] <Griffon26> "n00b: should I install linux if I plan to run freenet? answer: yes... and also if you don't plan to run freenet"
[10:29] <MakeSpace> :)
[10:32] <_ph00> Q: 'ho do I do that on windows?' A:'format drive and install Linux. Warning:process may cause data loss'
[10:33] <MakeSpace> i'de be nice and tell them to d/l the live cd ubuntu.
[10:34] <Griffon26> hehe.. that's especially funny with questions like "How do I see what files are in a directory on windows?"
[10:34] <Bombe> I wouldn't recommend Linux to anybody who has no clue of Linux.
[10:34] <Griffon26> format drive, install linux, ls
[10:35] <MakeSpace> _phoo: that metadata.xxx goes in both refs when I connect the two nodes, right? I don't need to have that in the refs when I give them out to other nodes, right?
[10:35] <Griffon26> or possibly: format drive, install linux, echo "There are no files in this directory... anymore"
[10:35] <_ph00> OT: what to do in case of terror attck:
[10:35] <_ph00> http://www.houstonjusticenotwar.org/humour/terrorist_attack/index.asp
[10:36] * toad_ (n=toad@) has joined #freenet
[10:36] * ChanServ sets mode +o toad_
[10:36] <Griffon26> Bombe: unless they have no clue of Windows either.. could help reduce spam
[10:36] <_ph00> I usually recommend using ubuntu as your first distro. after a couple of months, switch to something less noobish, but for a many-years windoze user who never run linux before, ubuntu us good
[10:38] <sleon> i second Bombe
[10:39] <_ph00> I knew nothing about linux when I first tried it. How are you supposed to learn stuff if you never try
[10:40] <_ph00> <MakeSpace> _phoo: that metadata.xxx goes in both refs when I connect the two nodes, right?
[10:40] <_ph00> yes
[10:40] <CIA-5> nextgens * r10502 /trunk/freenet/src/freenet/ (3 files in 2 dirs): Refactor NodeClient.storeConfig() into FilePersistentConfig.
[10:40] <_ph00> (sorry I didn't see you question before)
[10:40] <sleon> _ph00: read a book about it ?
[10:40] <sleon> _ph00: then read another one
[10:40] <sleon> _ph00: and get a clue
[10:40] <_ph00> and then ger scared away
[10:41] <_ph00> get
[10:41] <sleon> you get scared away to or so
[10:41] <_ph00> I read a couple of distro's homepages
[10:41] <sleon> if you are typical windows click and run user
[10:41] <MakeSpace> _ph00. I tried to add the line but it messes up the signatures (which is a good thing). I tried without the lines and the two nodes are talking fine.
[10:41] <_ph00> a couple of posts, forums, etc about linux
[10:41] <_ph00> then I tried a couple of live CD's
[10:41] <_ph00> and then installed ubuntu
[10:42] <_ph00> oh. OK. I used that line when I had nodes on one machine, and it worked, but that was several weeks ago. if it works anyway, then OK
[10:43] <MakeSpace> Oops. Spoke too soon. With only the one connection, the new node isn't happy. "Error: Splitfile error<pre>18 Route not found - could not find enough nodes to be sure the data doesn't exist </pre>"
[10:44] <MakeSpace> Hm... have to wait till I have a few more nodes I can really trust...
[10:45] <MakeSpace> but they can ntn message each other.
[10:45] <_ph00> you don't ever get anything like that
[10:45] <nextgens> [10:26] < Griffon26> | guys, do I understand correctly that if bwLimitDelayTime and
[10:45] <nextgens> pInstantReject are usually about 0, I should be adding more peers?
[10:45] <_ph00> the best you get are 'nodes you don't completely distrust'
[10:45] <nextgens> no
[10:46] <Griffon26> nextgens: what's the criteria to determine whether to add more nodes then?
[10:46] <nextgens> hi toad_
[10:47] <nextgens> Griffon26> adding more nodes for adding more nodes is silly
[10:47] <nextgens> unless you just want to stress test freenet
[10:48] <nextgens> bback sent me a mail a while ago
[10:48] * nextgens is wondering whether implementing what he wants is worthy or not
[10:48] <nextgens> "
[10:48] <nextgens> I want to add a list of chk keys (newline separated) into the fetch
[10:48] <nextgens> line in fproxy (or into a new field), and the node adds all files into
[10:48] <nextgens> the global queue with filename and target = node download folder.
[10:48] <nextgens> Is this possible? The field could be on the queue page, beside the insert
[10:48] <nextgens> field.
[10:48] <nextgens> "
[10:48] <nextgens> is it usefull ?
[10:48] <Griffon26> nextgens: I did not claim adding more nodes is the goal. My question is when should you add more and when should you stop adding?
[10:49] <nextgens> you should add more if you don't have 3 connected peers at any given time
[10:49] <nextgens> you should stop adding more as soon as you do :p
[10:50] <nextgens> of course you can always add more "really trusted" peers
[10:50] <_ph00> OT: I found the 'duck and cover' of the 2000's
[10:50] <_ph00> http://www.ready.gov/
[10:50] <nextgens> but adding more untrusted peers won't help you in any way
[10:51] <nextgens> toad_> I'm currently re-processing pipermail archives ... to do it properly I'd better block the mailing lists
[10:52] <nextgens> otherwise we risk having inconsistent archives
[10:53] <nextgens> hmm
[10:53] <nextgens> in fact I'll do that
[10:53] <Griffon26> nextgens: it could well connect you into a part of the network containing most of the content you usually access
[10:53] <nextgens> don't worry if the mailing lists are laggy today
[10:53] <nextgens> Griffon26> you should NOT try to optimize that on your own
[10:54] <nextgens> that's the network and routing algorithm's work
[10:54] <nextgens> btw, how do you determine where the content you're playing with is ?
[10:54] <nextgens> you can have the location of the manifest ...
[10:55] <Griffon26> no idea, but I thought 3 nodes was a bit of a small number for freenet to work with
[10:55] <nextgens> but what about the keys inside it ?
[10:56] <nextgens> ok, I've blocked mailman's queue processing
[10:56] <nextgens> it will be restarted when I'm done with pipermail archives
[10:57] * Agent43958723490 (n=tootsie@) has joined #freenet
[10:57] <nextgens> Griffon26> btw, the network wouldn't work if it was working that way :)
[10:58] <nextgens> as the location of the content is function of its hash
[10:59] <nextgens> freenet is based on the fact that content's hash is distributed uniformly on the "keyspace"
[11:04] <MakeSpace> _ph00, I can't pm. but the answer is yes.
[11:04] <_ph00> k
[11:04] <_ph00> register you nick if you want to pm
[11:04] <Griffon26> lol.. just heard. Some kind of report was leaked. "... the french government has not confirmed the report, but does plan to investigate the leak."
[11:05] <sleon> Griffon26: ihihihi
[11:05] <_ph00> hehe
[11:06] <nextgens> Griffon26> what are you refering to ? ben laden's story ?
[11:06] <_ph00> that spokeperson should be looking for a new job already
[11:06] <Griffon26> yes
[11:07] <_ph00> what bin laden story? any link?
[11:07] <nextgens> http://news.google.fr/news/url?sa=t&ct=fr/0-1-0&fp=4515e8ddf2374f13&ei=bRUVRYLrNZf8wQHkh8jPDg&url=http%3A//info.france2.fr/monde/24545121-fr.php&cid=1102795738
[11:08] <nextgens> http://www.agoravox.fr/article.php3?id_article=13734
[11:08] <nextgens> the second one is better
[11:08] <sleon> nextgens: ? and in english?h?
[11:09] <nextgens> nope, that's in french
[11:10] <nextgens> well, to sum up, a newspaper has published some unverified intelligence from DGSE
[11:10] <nextgens> telling that Ben Laden is likely to be dead since the 23. of september
[11:11] <Griffon26> that'd be today
[11:11] <nextgens> hmm, let me check
[11:11] <MakeSpace> yep its today here (seattle washington)
[11:11] <nextgens> sorry, that's when the newspaper published it
[11:12] * nextgens is looking for the translation of "typho?de"
[11:12] <Griffon26> typhoid
[11:12] <nextgens> so he would have dead on the 23. of august from a typhoid crisis
[11:14] <nextgens> not sure I'm making sense ... anyway the important bit in the story is that it was classified intelligence
[11:15] <nextgens> and it has been divulgated whereas it shouldn't had
[11:15] <nextgens> doh, my ssh connection is soooo laggy
[11:18] <Agent43958723490> hi guys/girls/etc.
[11:23] <MakeSpace> hello
[11:23] <sleon> toad_: ping
[11:27] <_ph00> i think osama's been dead longer than that. maybe a couple of years
[11:30] <Agent43958723490> So how exactly does one go about finding trusted nodes?
[11:31] <MakeSpace> good q. I have no answer :(
[11:36] * MakeSpace is now known as makespacev2
[11:40] <sleon> i think osama and bush are the same person
[11:40] * makespacev2 is now known as MakeSpace
[11:42] <Agent43958723490> heh
[11:42] <Agent43958723490> I think Bush is beating Osama in terms of body count right now.
[11:42] <MakeSpace> that mean bush is dead !
[11:43] * MakeSpace (n=jpp@) Quit ("using sirc version 2.211+KSIRC/1.3.12")
[11:44] <Agent43958723490> Another question, the list of "trusted" nodes I build up, are those the only entry/exit points into Freenet, or will my node also connect to "untrusted" nodes I don't know about?
[11:45] * MakeSpace (n=jpp@) has joined #freenet
[11:45] <MakeSpace> did I get this set up right? (Messing with IRC client)
[11:45] <nextgens> Agent43958723490> they are the only entry/exit point for data
[11:46] <nextgens> Agent43958723490> your node might talk to STUN servers
[11:46] <nextgens> and do some dns queries
[11:47] <Agent43958723490> nextgens: ok, thanks, also means those nodes will be the bottle neck... wish the internet over here wasn't so pathetically slow.
[11:55] * mozillaman (n=chatzill@) has joined #freenet
[11:57] <nextgens> mozillaman> 977 is out
[11:58] * QshelTier (n=bombe@) has joined #freenet
[11:58] * ChanServ sets mode +o QshelTier
[11:59] <mozillaman> nextgens: OK, what's new?
[12:02] <nextgens> it's a bugfix release
[12:02] <nextgens> FINs
[12:10] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[12:12] <TheSeeker> yey, KSK@FrostInstantNote.test.html loaded for me today.
[12:27] * QshelTier (n=bombe@) Quit (Read error: 110 (Connection timed out))
[12:29] * Chade25 (n=t2oni@) has joined #freenet
[12:33] * sbc (n=sbc@) has joined #freenet
[12:37] * Bombe (n=bombe@) has joined #freenet
[12:37] * ChanServ sets mode +o Bombe
[12:40] * FooDerGrosse (n=kevin@) has joined #freenet
[12:41] <_ph00> toad_ check your pm's
[12:42] * mario69 (n=nn@) has joined #freenet
[12:43] * _ph00 is now known as ph00
[12:50] * SebaLion (n=sebalion@) has joined #freenet
[12:56] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[12:57] <ph00> toad_ did you see my new msgs?
[13:04] * ph00 is now known as ph00_away
[13:05] <ph00_away> BBL
[13:05] * Bombe (n=bombe@) has joined #freenet
[13:05] * ChanServ sets mode +o Bombe
[13:11] <mario69> I think that update
[13:11] <mario69> detection is acting strange
[13:11] <Tritman> mario69, huh ?
[13:11] <mario69> 977 was published yesterday, right?
[13:12] <mario69> my node, which is 24/7 did not detect it till now, but when I did a restart it detected the update right away.
[13:14] <Tritman> you're seem true, me too, the #977 was only found after a restart (after more than one week Up)
[13:14] <mario69> in my case it was 4+ days uptime, so similar.
[13:15] <Tritman> and i've auto update at true in my node config since long time
[13:16] <mario69> checking...
[13:16] <mario69> I have: Check for, and download new versions = true; Automatically install new versions = false
[13:18] <Tritman> i've both true, but my node didn't automaticaly find update since yesterday, I saw node in 977 hence decided to restart and the 977 was found and fetched
[13:20] <mario69> and mine did not inform me, that a new version is available (yes, I refreshed the page :) . Let's leave this for the devs to have a look through the backlog and spot this.
[13:20] <Tritman> it seems leaving node up and waiting for update is not as so effective than a stop/start ;)
[13:25] <Tritman> my node didn't inform me too that a new release was available yesterday, I'll check in the wrapperl.log too, and nothing. It seems that when you start the node, it checks for a new release more actively.
[13:30] * Werdna (n=Andrew@) Quit (Read error: 60 (Operation timed out))
[13:30] <mario69> looks like periodic update check scheduling was somehow affected by changes between 96x and 97x. At least I remember 96x to detect updates pretty soon.
[13:33] <Tritman> ok it's not a pb if the update is done after some time (hours), but not some days (hence we'll ask for a strange behavior)
[13:41] * QshelTier (n=bombe@) has joined #freenet
[13:41] * ChanServ sets mode +o QshelTier
[13:47] * sbc (n=sbc@) Quit ("Leaving")
[13:53] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[14:06] * Bombe (n=bombe@) has joined #freenet
[14:06] * ChanServ sets mode +o Bombe
[14:06] * QshelTier (n=bombe@) Quit (Read error: 110 (Connection timed out))
[14:13] * QshelTier (n=bombe@) has joined #freenet
[14:13] * ChanServ sets mode +o QshelTier
[14:16] * Agent43958723490 (n=tootsie@) has left #freenet
[14:18] * railk (n=railk@) has joined #freenet
[14:22] * Bombe (n=bombe@) Quit (Read error: 104 (Connection reset by peer))
[14:34] * Bombe (n=bombe@) has joined #freenet
[14:34] * ChanServ sets mode +o Bombe
[14:34] * QshelTier (n=bombe@) Quit (Read error: 110 (Connection timed out))
[14:38] * wilx (n=wilx@) has joined #freenet
[14:48] * NullAcht15 (n=NullAcht@) Quit (Nick collision from services.)
[14:48] * unknown_ (n=NullAcht@) has joined #freenet
[14:49] * mario69 (n=nn@) Quit ()
[14:50] * unknown_ is now known as NullAcht15
[14:56] * ljn1981 (i=LJN@) Quit ("Chatzilla 0.9.75-2006090801 [Firefox 1.5.0.7/2006090921]")
[15:06] * Chade25 (n=t2oni@) Quit (Client Quit)
[15:08] * MikeW (i=Mike@) has joined #freenet
[15:09] * QshelTier (n=bombe@) has joined #freenet
[15:09] * ChanServ sets mode +o QshelTier
[15:11] * Bombe (n=bombe@) Quit (Read error: 110 (Connection timed out))
[15:31] * TheShado (i=Nonofyou@) has joined #freenet
[15:36] <MakeSpace> my node rebooted about 1 hour 20 minutes ago. I didn't request the reboot, I don't have auto update on. Any idea what happened or how I could find out?
[15:36] <MakeSpace> s/reboot/restart/
[15:37] * nextgens sets mode +v TheShado
[15:37] <nextgens> MakeSpace> watch logs out
[15:37] <nextgens> especially wrapper.log
[15:37] <nextgens> stick it to a pastebin then maybe we can help
[15:38] <TheShado> thanks Can I get cloak for my host here
[15:38] <MakeSpace> I gunziped the logs in freenet/logs and could find nothing interesting around the time of the restart... I'll look for wrapper.log...
[15:38] <nextgens> TheShado> I'm effraid you'll have to ask a staffer
[15:39] <TheShado> okay how can I contact a staffer
[15:39] <nextgens> do /stats p
[15:40] <TheShado> okay thank you
[15:44] * TheShado (i=Nonofyou@) has left #freenet
[15:54] * xpx (n=dIQ9TqFB@) has joined #freenet
[15:56] * xpx (n=dIQ9TqFB@) has left #freenet
[15:56] * xpx (n=dIQ9TqFB@) has joined #freenet
[15:57] <xpx> #freenet-refs seems to be empty
[15:57] <xpx> i dunno where to get refs from
[15:58] <nextgens> they are 43 people in there
[15:58] * xpx (n=dIQ9TqFB@) has left #freenet
[15:58] * xpx (n=dIQ9TqFB@) has joined #freenet
[15:58] <xpx> so i just can't see them?
[15:59] * sanity (n=ian@) has joined #freenet
[15:59] * ChanServ sets mode +o sanity
[15:59] * xpx (n=dIQ9TqFB@) has left #freenet
[16:02] * ph00_away is now known as _ph00
[16:06] <_ph00> toad_ you got mail
[16:14] <toad_> actually, more than 3 connections is a good thing
[16:14] <toad_> oskar said you need reasonable "degree" i.e. connections per node
[16:14] <toad_> HOWEVER as many of them as possible should be vaguely-trusted as opposed to total-strangers
[16:16] <toad_> hmmm
[16:17] <toad_> mario69/Tritman: well 977 only came out yesterday ... nextgens probably inserted it quite late on (i forgot to) ...
[16:18] <toad_> Tritman: still here
[16:18] <toad_> ?
[16:19] <toad_> nextgens: when did you insert of 977 finish?
[16:20] <nextgens> toad_> yes it did a while ago
[16:20] <nextgens> and we had success reports
[16:21] <nextgens> though I do think we should reschedule the USK fetching process every once a while
[16:22] <Tritman> toad_, yes
[16:23] <Tritman> toad, nextgens I restarted my node 4h54m ago
[16:23] <Tritman> to update fro 976 to 977
[16:24] <Tritman> sorry not 4h54, (after a refresh of the frpoxy) it is 7h23m
[16:29] * toad_ (n=toad@) Quit (Read error: 104 (Connection reset by peer))
[16:29] * toad_ (n=toad@) has joined #freenet
[16:29] * ChanServ sets mode +o toad_
[16:30] <Tritman> hi toad
[16:31] <_ph00> toad_ disconnected. n2n msg sent anyway
[16:33] * toad_ (n=toad@) Quit (Read error: 104 (Connection reset by peer))
[16:34] * MikeW (i=Mike@) Quit (Read error: 113 (No route to host))
[16:47] * defnax (n=defnax@) has joined #freenet
[16:47] <Tritman> bbiab
[16:48] * defnax (n=defnax@) Quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
[16:53] <Tritman> back
[17:03] * toad_ (n=toad@) has joined #freenet
[17:03] * ChanServ sets mode +o toad_
[17:07] * MikeW (i=Mike@) has joined #freenet
[17:18] * mozillaman (n=chatzill@) Quit ("later")
[17:36] <toad_> rehi
[17:38] * nirm (n=nirm@) has joined #freenet
[17:38] * nirm (n=nirm@) has left #freenet
[17:42] <Tritman> toad_, an idea why the auto-update didn't view/check the new release this morning ?
[17:49] * ljn1981 (i=LJN@) has joined #freenet
[17:54] * ljn1981 (i=LJN@) Quit (Remote closed the connection)
[17:55] <Tritman> ... hence bbl for any answer ;)
[17:55] <toad_> Tritman: :|
[17:55] <toad_> Tritman: no idea :|
[17:55] * phrosty (n=phrosty@) has joined #freenet
[18:20] * Turakamu (n=Steve@) has joined #freenet
[18:23] * Nextgens_ (i=xterm@) has joined #freenet
[18:27] <_ph00> we got a sp00fer ladies and gentlemen!!!
[18:29] * Nextgens_ (i=xterm@) Quit (Excess Flood)
[18:30] <_ph00> and *that*, believe it or not, is the reason why Tor users must get viced aby a channel op: this sp00fer is not going to annoy anyone
[18:30] <_ph00> becasue
[18:30] <_ph00> HE CAN'T TALK!!!!
[18:30] <_ph00> nice job, ops
[18:30] <_ph00> viced = voiced*
[18:36] <toad_> anyone know how to set firefox to not google stuff when i middle click on the wrong place?
[18:36] <phrosty> hmm? you must have a weird extension, that doesn't happen to me
[18:38] <_ph00> you probably need to reconfigure tab mix
[18:39] <_ph00> tab mix and tab mix plus do some weird stff by default on middle click
[18:43] * Turakamu (n=Steve@) Quit ("Leaving")
[18:45] <Tritman> rehi
[18:45] <Tritman> ok toad_
[18:45] <toad_> how long had the node been up?
[18:45] <Tritman> we'll see with the next release how the auto-update works
[18:45] <Tritman> mine ?
[18:45] <toad_> when it failed to auto-update
[18:45] <Tritman> more than 2 weeks
[18:45] <toad_> and it didn't even start to fetch it?
[18:46] <Tritman> it was in 976 since ther last update and never down
[18:47] <_ph00> m2m message sent: this time I did get a 'sent' feedback from my node, so probably those messages are not stored and sent on connection, they can be sent only to nodes that are already connected
[18:48] <Tritman> and never found the 977, as I've seen other node in 977 I decided to test a restart
[18:48] * wilx (n=wilx@) has left #freenet
[18:48] <toad_> _ph00: they're supposed to be stored
[18:48] * SinnerG (i=SinnerG@) Quit (Read error: 110 (Connection timed out))
[18:48] <Tritman> and in the wrapper.log I've seen with a tail -f wrappre.log that my node was fetching the 977
[18:48] <Tritman> during the restart
[18:48] <_ph00> well, when I tried to send to a non-connected peer, it didn't work
[18:49] <_ph00> (but it's not impossible that I did something wrong)
[18:52] * Urs_ShPo (n=gaim@) has joined #freenet
[18:53] <Tritman> toad_, that's all that I can say for the moment, I'll prompt the dev team if I note that the next update won't be automaticaly found after your announce (frost or irc) : hence don't forget to make an announce when it's available ;)
[18:53] <toad_> Tritman: :)
[18:53] <toad_> Tritman: thanks
[18:53] <toad_> Tritman: there's a bug filed for it, please get an account on the bugtracker so you can interact with us via that
[18:54] <Tritman> i need a login/pass to access bugtrack no ?
[19:01] <Tritman> ok, i withdraw my question (https://bugs.freenetproject.org/view.php?id=764)
[19:01] <Tritman> bbl all, good night
[19:09] <_ph00> BRB
[19:20] * blibbet (n=blibbet_@) has joined #freenet
[19:46] <sleon> guys. what is your misrouting distance ? (in advanced view of darknet in fproxy)
[19:46] <sleon> i have missRoutingDistance: 0.0556
[19:46] <sleon> is there thaw sourcecode available _
[19:46] <sleon> who is developing it again_
[19:46] <agsarite> missRoutingDistance:??0.0068
[19:47] * Ash-Fox (i=UNKNOWN@) has joined #FreeNET
[19:49] <sleon> agsarite: thx
[19:52] <_ph00> back
[19:52] <_ph00> *no* msgs in the meantime?
[19:52] <_ph00> almost one hour with no msgs no joinings no leavings
[19:52] <_ph00> (am I still connected?)
[19:52] <agsarite> uh, your file is 98.2% complete?
[19:57] * agsarite decides to sleep the rest of the day way.
[20:26] <hjubal> bye
[20:26] * hjubal (n=hjubal@) Quit (Read error: 104 (Connection reset by peer))
[20:30] <nextgens> toad_> around ?
[20:30] <toad_> nextgens: hi
[20:30] <nextgens> cool
[20:31] <toad_> nextgens: any idea where zothar or jogy might be contacted, other than their @freenetproject.org's?
[20:31] * sandos_ (n=sandos@) has joined #freenet
[20:31] <nextgens> I need your input : what's the most important ? improving the usability through implementing a "first time wizard" or finishing StS ?
[20:31] <toad_> or blured75 for that matter
[20:31] <nextgens> have you looked at the /etc/aliases files ?
[20:32] <nextgens> mantis should contain a valid mail address too
[20:32] <toad_> nextgens: STS at the moment
[20:32] <nextgens> ok
[20:32] <toad_> nextgens: first time wizard is important, but the peasants are still revolting ;)
[20:33] <toad_> there's a lot of crap being put out about 0.7 being far less secure than 0.5, and one of the big things which is actually true is the lack of STS
[20:34] <nextgens> the thing is that I don't like StS :p
[20:34] <toad_> hmmm?
[20:34] <toad_> what do you mean?
[20:34] <nextgens> I would prefer to do something DoS resistent
[20:34] <toad_> you think the protocol sucks because you can DoS it?
[20:34] <nextgens> yes
[20:34] <toad_> JFK will take significantly longer to do
[20:34] <nextgens> agreed
[20:34] <toad_> and DoS doesn't matter on a darknet
[20:34] <toad_> on an opennet, yeah, that could be a problem
[20:35] <toad_> but opennet has much bigger issues to deal with
[20:35] <toad_> it's like spending months securing the random number generator ; there are probably easier attacks
[20:35] <toad_> but if you want to do JFK after you've done STS, that's fine
[20:35] <toad_> but we DO need STS asap
[20:35] <nextgens> ^^-^^
[20:35] <nextgens> we haven't got around implementing it since .7 has started ...
[20:36] <nextgens> almost a year ago now
[20:36] <nextgens> but I agree, we need it
[20:36] <toad_> we need it
[20:37] <nextgens> though I'd prefer to be sure about our current crypto code before doing anything
[20:37] <nextgens> anything "forward" I meant
[20:37] <nextgens> hence I'd like to have scott's input on the topic
[20:38] <nextgens> maybe the protocol I've read is something optionnal added to the protocol after it has been designed
[20:39] <nextgens> but in its current state, I'm definitly not sure of what we are/have been doing with the DSA code
[20:39] <toad_> well...
[20:39] <nextgens> if provent missimplemented, it's likely to be even more urgent than sts
[20:39] <toad_> we need to talk to scott, yes
[20:39] <nextgens> -t
[20:39] <toad_> maybe you can get ian to phone him or something :)
[20:40] <toad_> but my inclination is:
[20:40] <toad_> if it's obvious that we've half-implemented the same algorithm as is described on wikipedia, then just complete the implementation
[20:40] <toad_> is it?
[20:41] <nextgens> well, we have half-implemented something said to be the DSA algorithm on wikipedia
[20:41] <nextgens> but it doesn't mean we were aiming to implement "real" DSA
[20:41] <toad_> well, is it a strict subset of the real DSA algorithm ?
[20:41] * nextgens will have a look on how it was implemented on .5
[20:41] <toad_> we copied the code
[20:41] <nextgens> yes it's definitly a subset of the "real" dsa algorithm
[20:42] <toad_> okay then fix it
[20:42] <toad_> if it suddenly breaks everything we know there's a deeper problem
[20:42] <nextgens> the problem beeing that it won't be backward compatible ...
[20:42] <nextgens> we have been ommiting to do some tests
[20:43] <nextgens> meaning that we might have selected some DSA groups/keys that aren't "compatible"
[20:43] <toad_> i thought it was during verification?
[20:43] <nextgens> meaning that if I "fix" the dsa signature verification code, some nodes might need to be rekeyed
[20:43] <nextgens> both
[20:44] <toad_> if it's during verification then it shouldn't be a problem
[20:44] <toad_> hmmm
[20:44] <toad_> what do we miss out during key generation?
[20:44] <nextgens> well keys durring verification has to match some properties ...
[20:44] <toad_> well there are two operations
[20:44] <nextgens> and it ensures the verification will go well
[20:44] <toad_> signing and verification
[20:44] <toad_> right?
[20:44] <nextgens> yes
[20:44] <toad_> what do we miss out during signing exactly?
[20:45] <nextgens> nothing, but we are missing that out during the key generation
[20:45] <nextgens> I mean the proceedure is
[20:45] <toad_> are you sure?
[20:45] <nextgens> 1) generate the key
[20:45] <nextgens> 2) verify that it matches some properties
[20:45] <nextgens> 3) sign
[20:45] <nextgens> 4) go through the wires
[20:45] <toad_> yes but what are the properties that it should match that we don't already check?
[20:45] <nextgens> 4) verify that the pubkey/signature matches some properties
[20:46] <nextgens> 5) verify the signature
[20:46] <nextgens> we are currently missing steps 2 and 4
[20:46] <toad_> we DO check for example that x < g
[20:46] <nextgens> let me have a look
[20:46] <nextgens> it's not enough
[20:46] <nextgens> that's the point
[20:46] <toad_> what else is there?
[20:47] <nextgens> we need to ensure that the number of bits of the L prime is in between 512 and 1024
[20:47] <nextgens> and that it's dividable by 64
[20:47] <toad_> that's the group generation
[20:48] <nextgens> we need to verify that q is 160 bits long and a Sophie Germain prime
[20:48] <toad_> p, q and g make up the group
[20:48] * sandos__ (n=sandos@) Quit (No route to host)
[20:48] <toad_> and we do such verifications when generating the group
[20:48] <nextgens> yes
[20:49] <toad_> so what's the problem?
[20:49] <nextgens> but we need to check that the group is valid with the selected private key
[20:49] <nextgens> iirc
[20:49] <nextgens> let me check
[20:49] <toad_> how do we do that? i thought it was simply 0 < x < q
[20:50] <toad_> i wonder if we check that x != 0
[20:51] <nextgens> HUH
[20:51] <toad_> we don't
[20:51] <toad_> we should, although it's extremely unlikely
[20:51] <nextgens> the wikipedia's pages are wrong!
[20:51] <nextgens> http://en.wikipedia.org/wiki/Digital_Signature_Algorithm
[20:51] <toad_> nextgens: what?
[20:51] <nextgens> http://fr.wikipedia.org/wiki/Digital_Signature_Algorithm
[20:51] <nextgens> the french one is wrong
[20:51] <nextgens> see the verification
[20:51] <toad_> nextgens: I was looking at the english version :)
[20:51] <nextgens> they forget the "Reject the signature if either 0<r<q or 0<s<q is not satisfied." bit as we do
[20:52] <nextgens> I wasn't hence I had problem remembering what we are forgetting to do ;)
[20:52] <toad_> lol
[20:54] <nextgens> ok, I'll fix the french version
[20:55] <nextgens> in fact according to the history of the page it wasn't there previously
[20:55] <nextgens> that check has been added
[20:55] <nextgens> we should add it
[20:59] <toad_> sure, fix it
[21:00] <_ph00> dude! I guess this is my reward for hanging around on an irc channel for the third saturday night in a row: all my 13 peers are connected, no back offs, no disconnected
[21:00] <toad_> bbiab
[21:11] * hdp (n=hdp@) Quit ("Leaving")
[21:17] <nextgens> http://code.bulix.org/vqc31y-18065
[21:17] <nextgens> :<
[21:17] <nextgens> sun's jvm is crap
[21:20] <nextgens> toad_> yetAnotherQuestion : shall we do the verification when instanciating the dsagroup object or just when verifying the signature ?
[21:20] <MakeSpace> does anyone know (or can they point me at the source) if the udp traffic has any freenet specific headers sent in plain text?
[21:21] <_ph00> it's encrypted afaik
[21:21] <nextgens> MakeSpace> the outgoing traffic is cyphered/padded with "random" bits
[21:21] <nextgens> there is no known fingerprint
[21:21] <_ph00> so, it looks like random traffic, right?
[21:22] <nextgens> almost
[21:22] <nextgens> _ph00> then we will have to debate on what randomness is
[21:22] <nextgens> and how random is our RNG implementation :D
[21:22] <_ph00> you may want to get some more udp traffic to mix up with that, just to make it even more difficult for an eventual spy
[21:23] <nextgens> it's not fingerprintable according to size either
[21:23] <nextgens> well they are limits
[21:23] <nextgens> the MTU and the min. packet size
[21:23] <_ph00> ok, I didn't really menat 'random', more like 'non specifically identified traffic'
[21:23] <nextgens> but that's about matching almost anything
[21:24] <nextgens> moreover port numbers are "unidentified" as well
[21:24] <nextgens> MakeSpace> btw, why are you asking ? aren't you a bad guy trying to match freenet's traffic ? :p
[21:25] <nextgens> in fact it's matchable if you do know both of the node's id
[21:25] <nextgens> and you agree with spending time trying to match the protocol bits decyphering the flow
[21:26] <nextgens> as the outer encryption key is just a XOR of both node's id iirc
[21:27] <_ph00> you men both nodes' IDs?
[21:27] <_ph00> like the IDs of both nodes?
[21:27] <_ph00> so, a really good bad guy
[21:27] <_ph00> hhmm....
[21:27] <_ph00> you really want to trust your peers...
[21:27] <_ph00> swapping refs with random users on #freenet-refs is *bad*
[21:28] <MakeSpace> I'm just a paranoid cryptogeek. right now for fun I am implementing a dm-crypt which swaps blocks to prevent update analysis...
[21:28] <MakeSpace> nextgens: JUST XOR!!! surely we can do better...
[21:29] * zhaan (n=zhaan@) has joined #freenet
[21:29] * zhaan (n=zhaan@) has left #freenet
[21:30] <nextgens> MakeSpace> what for ? :)
[21:30] <MakeSpace> _ph00: swapping refs with random ppl is bad, like u said. that's why I am running a second instance where the only folks I add aren't random freenet-refs folks
[21:30] <nextgens> MakeSpace> I'm talking about the outer encryption layer :)
[21:30] <nextgens> afterwards we do a DH exchange
[21:31] <nextgens> and we achieve PFS on inner encryption layers
[21:31] <MakeSpace> nextgens> swap blocks? just for fun. See (public WWW) http://dataquality.i2r.a-star.edu.sg/hhpang/publications/StegFS-traffic.pdf#search=%22update%20analysis%20steganographic%22
[21:31] <nextgens> _ph00> sending your references in plaintext is bad indeed
[21:32] <_ph00> I do worse than that
[21:32] <MakeSpace> nextgens> is it: frenet-header encrypt(payload) random-pad?
[21:32] <_ph00> I had my ref pubblished on a pastebin
[21:32] <_ph00> so basically the *badguys* know who ph00 is and know that he runs a freenode
[21:33] <MakeSpace> nextgens> or is it encrypt( freenet-header, payload, random-pad) ?
[21:33] <_ph00> np, as long it's not illegal, and as long as I am not recognizable on frost
[21:33] <_ph00> or on the rest of freenet
[21:35] <nextgens> MakeSpace> I don't remember, check the code :)
[21:36] <MakeSpace> nextgens> which files?
[21:36] <nextgens> it's likely to be in freenet.node.FNPPacketMangler
[21:36] <nextgens> iirc
[21:36] <MakeSpace> coo
[21:36] <nextgens> "Reject the signature if either 0<r<q or 0<s<q is not satisfied."
[21:36] <nextgens> that's silly isn't it ?
[21:36] <nextgens> hmm, sorry
[21:36] <nextgens> pasted the wrong line
[21:39] * SinnerG (i=SinnerG@) has joined #freenet
[21:42] <MakeSpace> btw, whats StS ?
[21:43] * ljn1981 (i=LJN@) has joined #freenet
[21:45] <ljn1981> Hey everyone
[21:46] <nextgens> Station to Station
[21:46] <nextgens> authenticated DH
[21:47] <MakeSpace> nextgens: thanks
[21:48] <ljn1981> toad_: Just read the latest mails on the list. It's scary how much it look like a campagne is being run aganst us.
[21:51] <ljn1981> So anyway what's up people?
[22:04] * Ash-Fox (i=UNKNOWN@) Quit (Read error: 145 (Connection timed out))
[22:05] <_ph00> campaing? against us? where? where? (WHERE is my gun?!?)
[22:07] <ljn1981> Hehe
[22:08] * steviewonder (n=dee@) has joined #freenet
[22:08] <steviewonder> where can i get the fuqid that works with freenet 0.7?
[22:08] <nextgens> bbiab
[22:08] <ljn1981> All the people on frost and now unjustified questions and accusations about 0.7 security on the mailing lists.
[22:08] <steviewonder> the 1 i was using doesn seem to connect to the new freenet
[22:09] <ljn1981> steviewonder: I don't know but I think there's a site on freenet. If you installed Thaw with freenet you could try that, it's trying to be a java implementation that does most of what FUQID does.
[22:11] <steviewonder> ok thanks
[22:11] <ljn1981> USK@DWB7AD1mBr5HNatdWvbqBe5FdZ3Zt~eS3nPHdp1Yg~g,HK1~x4rUgH6NyIT87CymSWcIHLP2Yc6dPQ417n1gnBY,AQABAAE/fuqid/1/
[22:11] <ljn1981> Found that in the tools section of http://freenetproject.org/
[22:11] <steviewonder> thanks
[22:11] <steviewonder> it will work with the new version?
[22:11] <steviewonder> any reason why the older one doesnt?
[22:11] <ljn1981> You need a new frost too if you have the old one and want to use it too. You can import your old identities.xml is you want.
[22:12] <ljn1981> Yes
[22:12] <steviewonder> i have the new frost but i have lost all my messages and boards
[22:12] <ljn1981> The FCP protocol is completely rewritten. FCP being the one clients use to talk to the node.
[22:12] <SinnerG> there, my host comp will be called : mystichost.net :P
[22:13] <SinnerG> aka ifreed.net will stay for freenet even tho I dont know how long I'll be able to run the node
[22:13] <SinnerG> maybe on low cpu
[22:13] <ljn1981> We have new boards anyway, the old ones wouldn't have worked.
[22:13] <steviewonder> ok
[22:13] <_ph00> #freenet-chat has gone almost 36 hours without ONE message.
[22:13] <_ph00> maybe more than that, I've been on since yesterday @ 3pm and no one sent anything so far
[22:14] <ljn1981> :P
[22:14] * ljn1981 points to #freenet-chat
[22:23] * FooDerGrosse (n=kevin@) Quit ("W??hlt die PiratenPartei Deutschland!!!")
[22:34] * MikeW (i=Mike@) Quit ()
[22:38] * Werdna (n=Andrew@) has joined #freenet
[22:38] * timmy2chk (n=Vincent@) has joined #freenet
[22:43] * shuttler (n=cc@) has joined #freenet
[22:45] <CIA-5> dbkr * r10503 /trunk/apps/Freemail/src/freemailgui/WizardAskGenKeys.java: Whoops
[22:46] <Griffon26> $ cat commit.log|grep -i oops|wc -l
[22:46] <Griffon26> 9.47837E9372
[22:47] <CIA-5> dbkr * r10504 /trunk/apps/Freemail/src/freemailgui/WizardAskGenKeys.java: Add the LGPL header
[22:50] * steviewonder (n=dee@) has left #freenet
[22:54] <toad_> <nextgens> toad_> yetAnotherQuestion : shall we do the verification when instanciating the dsagroup object or just when verifying the signature ?
[22:54] <toad_> no idea, whatever's best
[22:55] <toad_> <MakeSpace> does anyone know (or can they point me at the source) if the udp traffic has any freenet specific headers sent in plain text?
[22:55] <toad_> nope
[22:56] <toad_> MakeSpace: what's "update analysis" ?
[22:56] <MakeSpace> itoad_ ...
[22:57] * oCc (n=nodeOne@) Quit ("s?ren")
[22:57] <MakeSpace> toad_> if the attacker has two immages of your disk they can tell which blocks were updated. that way they can see if data is being updated in 'unused' blocks. and that way you loose plausible denighability about the existance of that data. But...
[22:58] <MakeSpace> toad_> the block swaping algorithm has another benefit. It is essentially continuously 'wiping' the drive by flipping bits and location of data with out loss of data.
[22:58] <toad_> <ljn1981> toad_: Just read the latest mails on the list. It's scary how much it look like a campagne is being run aganst us.
[22:58] <toad_> sometimes it feels that way
[22:58] <toad_> i suppose the answer is just to implement opennet...
[22:58] <ljn1981> Especially lately
[22:59] <MakeSpace> which emails? what campaign?
[22:59] * randomselect (n=chatzill@) Quit ("no reason")
[23:00] <toad_> MakeSpace: nice
[23:00] <toad_> MakeSpace: on the tech and support lists...
[23:00] <MakeSpace> (I am here you can stop beeping)
[23:00] <toad_> apparently as a result of some posts on frost/0.5
[23:01] <MakeSpace> (or have I done something dumb with my irc client?)
[23:01] <toad_> there seems to be a fairly active campaign to get people to use 0.5, on the grounds that it's more secure
[23:01] <toad_> except it isn't
[23:01] <toad_> well it is in a few minor ways
[23:01] <MakeSpace> OIC. How is it more secure?
[23:01] <toad_> i dunno, smaller blocks may make correlation attacks easier, but they're pretty easy anyway
[23:02] <toad_> well, 0.7 has some major benefits for hostile-environment; it's invisible (indistinguishable from random data), unlike 0.7, for example
[23:02] <toad_> and if you use it in true darknet mode, it's significantly more secure, because it limits exposure
[23:02] <MakeSpace> I think correlation attacks can be beat by matching the distribution of your response times with the distribution of times of responses to your queries.
[23:03] <toad_> no, we're not talking about datastore probing
[23:03] * shuttler (n=cc@) Quit (Read error: 145 (Connection timed out))
[23:04] <MakeSpace> oops. give me a pointer, i'll educate myself about correlation attacks (and shutup about them until then :-S )
[23:04] <toad_> we're talking about attacks based on active splitfile requests - you see x% of a splitfile in requests coming from X, you calculate that it's very likely to have come from that node itself rather than being relayed
[23:04] <MakeSpace> hmm....
[23:04] <sanity> hi toad
[23:04] * sanity is having teeth issues
[23:05] <toad_> sanity: hi
[23:05] <sanity> i think i might need surgery to have a wisdom tooth removed on Monday
[23:05] <toad_> sanity: do you think there's a deliberate campaign to defame freenet 0.7?
[23:05] <toad_> sanity: hmmm :(
[23:05] <ljn1981> sanity: Ouch, not fun at all.
[23:05] <sanity> toad_: i think lack of opennet is the root cause of the frustration - and i can understand it
[23:05] * Ash-Fox (i=UNKNOWN@) has joined #FreeNET
[23:05] <sanity> ljn1981: yup
[23:06] <sanity> ljn1981: at least i get a few days off work :-)
[23:06] <sanity> toad_: speaking of opennet, where is everything? are you still going to send a status report?
[23:06] <toad_> sanity: i'm not quite sure what the status _is_
[23:06] <toad_> but i may send something on monday]
[23:07] <ljn1981> I don't understand this wish for a less secure system.
[23:07] <sanity> toad: how can you not know what the status is?
[23:07] <ljn1981> sanity: Hehe, YAY, nothing is so bad it's not good for something.
[23:07] <toad_> ljn1981: they think opennet is MORE secure. that's the scary part.
[23:07] <toad_> ljn1981: that will hurt us long-term.
[23:07] <ljn1981> toad_: I know, and most won't listen to reason either.
[23:08] <sanity> a designed opennet is more secure than a haphazard one such as what we have now
[23:08] <sanity> and it is certainly more convenient for users
[23:09] <toad_> it is marginally more secure than what we have now
[23:09] <MakeSpace> re: correlation attacks. I suppose you onion-shell route your requests and responses... ugh to do that a node needs to know info about non-peer nodes.each peer.
[23:09] <sanity> toad_: marginally more secure, but *far* easier for users to get started with freenet
[23:09] <toad_> it is much more convenient, and yields a significantly better topology
[23:10] <sanity> toad_: you didn't answer my question : how can you not know the status?
[23:10] <toad_> sanity: well i suppose i must know better than anyone else does :)
[23:10] <sanity> toad_: i should think so
[23:10] <MakeSpace> re opennet: after joining system via open net (like I have essentially done) you can get to know folks, build trust, and select peers for the dark-net
[23:11] <toad_> okay, i'll try to get some sort of status report out on monday
[23:11] <sanity> toad_: i am concerned that not a lot seems to be happening - despite our funding windfall
[23:12] <toad_> well, i had a long period off, at least a week, not long ago
[23:12] <sanity> toad_: sure, but what about mrogers and the load balancing?
[23:12] <toad_> i don't know
[23:12] <ljn1981> MakeSpace: But how many users will bother? It works fine without and then it's going to get blocked and there wiwll be a big mess essenially like we have now but we'll have spent significant effort creating the now useless opennet
[23:12] <toad_> i do know he's making progress on his original roadmap
[23:12] <toad_> what was his last commit batch?
[23:13] <toad_> 1sec...
[23:13] <ljn1981> At least that's my main problem with opennet, waste of resources for a less efficient result.
[23:13] <ljn1981> s/result/product/
[23:13] <sanity> toad_: he is offering no transparency into what he is doing - at least that i have seen
[23:13] <toad_> 14 september he put a big batch in
[23:14] <sanity> toad_: yeah, but apart from commit messages with terse comments, we don't actually know if he has learned anything
[23:14] <toad_> well, he regularly commits his changes, which is good ... but I haven't seen any actual simulation results
[23:14] <sanity> toad_: exactly
[23:14] <toad_> also I asked him for a summary of the low-level changes we had discussed, so we could implement that
[23:14] <toad_> which hasn't materialized
[23:14] <sanity> doesn't it defeat the point for us to implement it before he has simulated it?
[23:15] <toad_> he's simulated the low level stuff
[23:15] <toad_> pretty much
[23:15] <toad_> phase 6 is CHKs and SSKs ... that's getting very high level
[23:15] <sanity> simulated it and discovered what exactly?
[23:16] <toad_> some of his commits say X works better than Y
[23:16] <toad_> but I agree we need a summary of his results so far and the proposed low level changes
[23:17] <sanity> toad_: can you email him and politely express our concerns?
[23:17] <toad_> yes
[23:17] <sanity> toad_: if you would like me to look over your email before you send it, i'm happy to
[23:17] <sanity> toad_: and either way, could you cc me?
[23:17] * timmy2chk (n=Vincent@) Quit ()
[23:17] <toad_> yeah
[23:21] <toad_> sent
[23:22] <toad_> there are a number of people talking about big freesite inserts etc; there are some things I need to do to facilitate that - restartable freesite inserts, multi-container freesites etc
[23:22] <toad_> content -> users -> content -> users
[23:22] <sanity> i see priorities are:
[23:22] <sanity> 1) opennet
[23:22] <sanity> 2) speed
[23:23] <sanity> hopefully load balancing will address 2
[23:23] <toad_> https://bugs.freenetproject.org/view.php?id=765
[23:23] <sanity> i really think that working on opennet will be a catalyst for progress generall
[23:23] <toad_> we have loads of volunteers right now
[23:23] <toad_> even by historical standards
[23:24] <sanity> toad_: true, but i'm worried that they are a self-selecting group with a high tolerance for usability issues
[23:24] <toad_> well, there's been a lot of work on second-order usability stuff
[23:24] <sanity> toad_: second-order?
[23:24] <toad_> i.e. all usability issues other than opennet
[23:25] <sanity> including the installer?
[23:25] <toad_> the installer is pretty good
[23:25] <toad_> if we had opennet it'd be far easier than 0.5
[23:25] <sanity> toad_: i know
[23:25] <sanity> toad_: i just think that implementing opennet should be relatively straight-forward
[23:26] * Werdna (n=Andrew@) Quit (Read error: 110 (Connection timed out))
[23:26] <sanity> toad_: do you have a sense of why some requests take ages?
[23:26] <toad_> implementing it won't be _that_ easy, but is feasible
[23:27] <toad_> making it _WORK_ is a lot harder
[23:27] <toad_> sanity: some requests but not others?
[23:27] <sanity> i think you are being excessively pessimistic
[23:27] <toad_> sanity: USK@.../-1/ links are slow because it has to wait for lots of DNFs...
[23:27] <sanity> toad_: i waited over 5 minutes while trying to retrieve a freesite the other day
[23:27] <toad_> was it on a /-1/ link?
[23:27] <sanity> perhaps that was the problem
[23:28] <sanity> it was that new index that someone mentioned a while back
[23:28] <toad_> maybe we should cache them for a while
[23:28] <ljn1981> I'd want more speed before opennet if it was up to me. More speed is not really worth any outside announcement but opennet is. Announcing opennet before it get speedier might make a good chunk of potential users leave again because it's too slow.
[23:28] <toad_> and prefetch them
[23:28] <MakeSpace> re opennet: why do anything more sophisticated than a web page with a randomized list of registered refs. beside each ref is a button which generates the same querry to localhost:8888 that pushing the add button would?
[23:28] <toad_> pong :)
[23:29] <sanity> toad_: really, assuming 700ms round-trip time, no request should take more than 3-4 seconds
[23:29] <toad_> sanity: some requests will be more than 5 hops
[23:29] <sanity> toad_: some, but not much more
[23:29] <toad_> sanity: most freesites will cause more than one request (or none at all)
[23:29] * SebaLion (n=sebalion@) Quit ("( www.nnscript.de :: NoNameScript 4.03 :: www.XLhost.de )")
[23:29] <toad_> well it might be 10
[23:29] <ljn1981> MakeSpace: Too centralized, we don't like single points of failure.
[23:29] <toad_> shouldn't be more than that
[23:29] <toad_> i mean hops
[23:29] * jLadd (n=jladd144@) has joined #freenet
[23:29] <sanity> toad_: it just worries me that surfing 0.7 doesn't seem must faster than surfing 0.5 (back when 0.5 worked)
[23:30] <sanity> even with UDP and all that
[23:30] <toad_> yes, and THE reason (apart from issues with /-X/ searches) is that load limiting is screwed right now
[23:30] <toad_> people do say it's faster than 0.5 incidentally
[23:31] <sanity> toad_: is there a doc describing the current load balancing/limiting algo?
[23:31] <toad_> I dunno
[23:31] <sanity> toad_: but you didn't write one
[23:31] <toad_> not an up to date one no
[23:31] <ljn1981> I don't think so, but there were a lot of talk on the mailinglists when the new algo was being planned.
[23:31] <sanity> toad_: load balancing always just gets more and more complicated until nobody understands it and it becomes voodoo
[23:32] <toad_> we have the AIMD on the client, we have pre-emptive rejection on nodes ...
[23:32] <sanity> ljn1981: yeah, lost of ideas, but which ones actually made it into the code...
[23:32] <toad_> sanity: well, token passing should simplify matters, it's conceptually very nice
[23:32] <toad_> although I'm still not entirely clear how we determine how many tokens should be in flight
[23:32] <sanity> toad_: is there a doc or email describing that? i never really grokked it
[23:32] <toad_> sanity: there was a thread about it
[23:33] <toad_> sanity: TCP/IP has more than AIMD
[23:33] <toad_> sanity: it also has a receive window, or something, can't remember what it's called; anyway, it tracks how much is actually in flight, and if the window in flight is too big, it doesn't send any more data
[23:33] <toad_> it doesn't just calculate a rate
[23:34] <toad_> it adjusts instantly to temporary pipe clogging up
[23:35] <toad_> that principle is a big part of token passing
[23:35] <toad_> don't start requests faster than they can complete
[23:35] <sanity> well, lets see what mrogers says
[23:35] <toad_> for a while i was worried about pre-emptive rejection not working well... it's fine on my node at the moment though...
[23:36] <sanity> hmm, "don't start requests faster than they can complete" - i like that, its simple
[23:36] <toad_> the problem is we don't have actual packet loss on freenet; if we let it come to timeouts, that's very bad
[23:36] * sanity is currently on vicodin
[23:36] <sanity> ...for wisdom tooth
[23:37] <toad_> oh the thing I was worried about ... we pre-allocate bandwidth so that requests can't be started too often for our bandwidth usage
[23:37] <toad_> which should mean that bwlimitDelayTime never gets really high
[23:37] <toad_> because bwlimitDelayTime only goes high when there's a shortage of output bandwidth
[23:37] <toad_> yet bwlimitDelayTime *does* go high
[23:37] <toad_> and then pre-emptive rejection kicks in
[23:38] <toad_> sanity: another part of token passing is that we have a maximum number of requests in flight, which is something I've been thinking about doing anyway
[23:38] <sanity> i wonder if there are minor changes we can make to improve what we have now in lieu of the new algo
[23:39] <toad_> but we can do it cleanly with token passing
[23:39] <toad_> if we do it before that, it's just another layer of complexity, just another reason for requests to be rejected
[23:39] <sanity> yeah, i guess
[23:39] <toad_> hmmm
[23:39] <sanity> ok, lets see what mrogers response is
[23:39] <toad_> load limiting is at least limiting load
[23:40] <toad_> all but one of my nodes have probability of rejection < 5%
[23:40] <toad_> actually all but 2
[23:40] <toad_> 1 has 28%, one has 18%
[23:40] <toad_> the others are very low
[23:40] <sanity> toad_: oh yeah, can we connect? http://locut.us/ref
[23:40] <toad_> 6.3, 4.7 ...
[23:40] <sanity> i have 21% and 14%
[23:40] <toad_> sure, how do you want my ref?
[23:41] <sanity> url?
[23:41] <toad_> i refuse to stick it on the web ...
[23:41] <sanity> email then
[23:41] <toad_> on windows, .fref files are registered so if you double click them they're added automatically
[23:41] <sanity> toad_: do you normally find mrogers to be responsive?
[23:42] <sanity> toad_: i use a mac
[23:42] <sanity> toad_: but that is cool, i didn't know that
[23:42] <toad_> reasonably so; sometimes he disappears for up to a week...
[23:42] <sanity> what do you know about him?
[23:42] <sanity> location etc?
[23:43] <toad_> not a great deal; i know his later simulation work coincides with his academic research interests
[23:43] <toad_> so hopefully will be completed
[23:43] <sanity> m.rogers@cs.ucl.ac.uk
[23:43] <toad_> 1sec i find which uni he's at...
[23:43] <toad_> right
[23:43] <sanity> so he is in your neck of the woods
[23:43] <toad_> so we know where he is :)
[23:44] <sanity> http://www.cs.ucl.ac.uk/staff/mrogers/
[23:44] <sanity> seems to know his stuff
[23:45] <toad_> looks cool yeah
[23:45] <toad_> researchers usually ignore or ridicule us, this is a refreshing change :)
[23:46] <toad_> having said that they cite