Timestamps are in GMT/BST.
[0:06] * batosai (n=batosai@) Quit (Read error: 113 (No route to host))
[0:07] * nOgAnOo (i=nOgAnOo@) Quit (Read error: 104 (Connection reset by peer))
[0:30] * nOgAnOo (i=nOgAnOo@) has joined #freenet
[0:34] * zerwas_ (n=zerwas@) has joined #freenet
[0:44] * zerwas__ (n=zerwas@) has joined #freenet
[0:47] * zerwas__ (n=zerwas@) Quit (Client Quit)
[0:48] * Sevy (i=sevy@) Quit (wolfe.freenode.net irc.freenode.net)
[0:49] * zerwas (n=zerwas@) Quit (Read error: 113 (No route to host))
[1:05] * zerwas (n=zerwas@) has joined #freenet
[1:06] * zerwas_ (n=zerwas@) Quit (Read error: 113 (No route to host))
[1:08] * zerwas_ (n=zerwas@) has joined #freenet
[1:08] * zerwas (n=zerwas@) Quit (Nick collision from services.)
[1:08] * zerwas_ is now known as zerwas
[1:31] * ShadowofShinobi (i=Shadowof@) Quit (Read error: 54 (Connection reset by peer))
[1:31] * ShadowofShinobi (i=Shadowof@) has joined #freenet
[1:49] * Mr_Giraffe_ (n=dickbloo@) Quit (Remote closed the connection)
[2:02] * Zarggg (n=z@) Quit ()
[2:17] * gasi__ (n=gasi@) has joined #freenet
[2:33] * gasi_ (n=gasi@) Quit (Read error: 110 (Connection timed out))
[2:43] * caytchen_ (n=caytchen@) Quit ("Verlassend")
[2:44] * Tarantuka (n=eddie207@) has joined #freenet
[2:54] * kakakap (n=chatzill@) has joined #freenet
[2:54] * kakakap (n=chatzill@) Quit (Client Quit)
[3:01] * zerwas (n=zerwas@) Quit ("Keep on flowing ...")
[3:17] * n0ob (n=travis@) has joined #freenet
[3:25] * Tarantuka (n=eddie207@) Quit ("Quitte")
[4:06] * ShadowofShinobi (i=Shadowof@) Quit (Read error: 54 (Connection reset by peer))
[4:07] * ShadowofShinobi (i=Shadowof@) has joined #freenet
[4:23] * Pichu0102 (n=pichu010@) Quit (Read error: 110 (Connection timed out))
[4:23] * mrdmx (i=mrdmx@) has joined #freenet
[4:23] * JustMe (n=me@) has joined #freenet
[4:24] * mrdmx is now known as mrdmx-brb
[4:25] * Pichu0102 (n=pichu010@) has joined #freenet
[4:33] * Pichu0102 (n=pichu010@) Quit (Remote closed the connection)
[4:41] * mrdmx-brb is now known as mrdmx
[4:44] * Pichu0102 (n=pichu010@) has joined #freenet
[5:03] * mrdmx is now known as mrdmx-bbl
[5:07] * mrdmx-bbl is now known as mrdmx
[6:13] * ShadowofShinobi (i=Shadowof@) Quit (No route to host)
[6:17] * Pichu0102 (n=pichu010@) Quit (Read error: 110 (Connection timed out))
[6:18] * Pichu0102 (n=pichu010@) has joined #freenet
[6:24] * mikusr (i=mikusr@) has joined #freenet
[6:25] * mikusr (i=mikusr@) has left #freenet
[6:29] * ed__ (n=Ed@) has joined #freenet
[6:30] * edt (n=Ed@) Quit (Read error: 110 (Connection timed out))
[6:36] * n0ob (n=travis@) Quit ("Now onto something more interesting")
[6:42] * sbc (n=ca@) has joined #freenet
[7:00] * batosai (n=batosai@) has joined #freenet
[7:22] * mrdmx is now known as mrdmx-bbl
[7:45] * sbc (n=ca@) Quit (Read error: 104 (Connection reset by peer))
[7:50] * mrdmx-bbl is now known as MRDMX
[7:50] * MRDMX is now known as mrdmx
[7:51] * lugark_ (n=lugark@) has joined #freenet
[7:52] <lugark_> hi
[8:14] * Ratchet (n=xxxx@) has joined #freenet
[8:29] * christefano (n=christef@) has joined #freenet
[8:31] * Nande (n=KVIrc@) Quit ("byes, tengan la bondad de ser felices. ;)")
[9:08] * Mc2` (n=fred@) Quit (Read error: 110 (Connection timed out))
[9:13] * Mathiasdm (n=Mathias@) has joined #freenet
[9:18] <nextgens> hi
[9:19] * nextgens changes topic to 'http://freenetproject.org/download.html (1153 mandatory), please read before asking for help here. #freenet-refs is no more, turn on insecure mode if you don't have any friends using Freenet | 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-it #freenet-de | paste at http://code.bulix.org'
[9:21] * Pichu0102 (n=pichu010@) Quit (Connection timed out)
[9:21] <nextgens> hrrrmmmm
[9:21] <nextgens> apparently there is a problem
[9:21] <nextgens> did your peers update to 1154?
[9:22] <nextgens> and did your node use UoM to do so?
[9:22] <nextgens> (if it did you'll find "fetching main jar from XXX" in wrapper.log)
[9:43] * jan_ (n=jan@) has joined #freenet
[9:44] <Cooo> haven't been running my normal node, starting it now.
[9:52] * DaffyDuck_ (n=jan@) Quit (Read error: 110 (Connection timed out))
[9:52] * jan_ (n=jan@) Quit ("Leaving")
[9:53] * DaffyDuck_ (n=jan@) has joined #freenet
[9:56] * Pichu0102 (n=pichu010@) has joined #freenet
[9:59] <Cooo> Yep. Node is using UoM to update.
[10:00] <Cooo> Gotta go now. later
[10:08] <nextgens> that's bad news
[10:08] <nextgens> okay, will have a look
[10:10] * phrosty (i=phrosty@) Quit ("He who asks a question is a fool for five minutes; he who does not ask a question remains a fool forever.")
[10:14] * phrosty (n=phrosty@) has joined #freenet
[10:19] * Mc2` (n=fred@) has joined #freenet
[10:23] * toad_ (n=toad@) has joined #freenet
[10:23] * ChanServ sets mode +o toad_
[10:29] * mrdmx is now known as mrdmx-bbl
[10:32] <nextgens> hey toad_ !
[10:33] <toad_> hi
[10:33] <nextgens> toad_> I'm not sure how the updater is supposed to proceed anymore:
[10:33] <nextgens> do we want "normal" updates to be fetched by UoM or not?
[10:33] <nextgens> atm it seems to fetch everything through UoM
[10:33] <nextgens> imho it's not what we want
[10:33] <toad_> why?
[10:34] <nextgens> because it hogs one of our peer's bandwidth for a while
[10:35] <nextgens> whereas we could use freenet's caching to do it smothly over all of our peer's bandwidth
[10:35] <nextgens> would you object to me changing that behaviour?
[10:35] <toad_> well...
[10:35] <toad_> it's essential that we get the update reasonably quickly
[10:36] <toad_> you could make it attempt to download it for some period and then fall back to UOM perhaps
[10:36] <nextgens> yeah that's something I can do
[10:36] <toad_> but you'd need to be careful of uptime issues
[10:36] <nextgens> how long would be reasonable?
[10:37] <nextgens> persisting the time we attempted to fetch the update seems insane
[10:37] <toad_> why?
[10:38] <toad_> if we never reach an uptime of an hour, and routing is broken in some subtle way, we'll fail
[10:38] <toad_> well we'll fail until the mandatory kicks in on other nodes...
[10:39] <toad_> maybe we should try less than an hour - 10 minutes?
[10:39] <nextgens> updates are 2.5M big atm
[10:39] <nextgens> I wouldn't be surprised if it took longer than that to fetch
[10:39] <toad_> okay so half an hour? an hour?
[10:39] <nextgens> an hour sounded fair
[10:40] <toad_> okay... do we persist the time at which we started to fetch it?
[10:40] <nextgens> if the uptime is below one hour there is no way the node is going to be useful anyway
[10:40] <toad_> the uptime might be low because of a nasty bug
[10:40] <toad_> UOM is fast and reliable
[10:40] <nextgens> ttrue
[10:40] <nextgens> hmm
[10:41] <nextgens> my concern is that it's not "smooth"
[10:41] <nextgens> I mean, lots of nodes are likely to be disturbed when servicing updates to others
[10:42] <toad_> yes but it ALWAYS works
[10:42] <toad_> and it shouldn't use more than the bandwidth limit unless there is a bug
[10:42] <nextgens> okay; let's take the problem from the other hand
[10:42] <toad_> what is your suggestion?
[10:42] <nextgens> in which case would the normal updater fail ?
[10:43] <nextgens> 1) we broke the build at client's layer level
[10:43] <toad_> if routing is broken
[10:43] <nextgens> 2) we broke the build "at connection" layer
[10:43] <nextgens> 3) we can't keep the node long enough
[10:43] <nextgens> uom is safe up to level 2
[10:43] <nextgens> whereas normal update is safe up to level 1
[10:43] <nextgens> both fail in case 3
[10:44] <toad_> no, UOM works if routing is fscked
[10:44] <nextgens> sure; that's level 1
[10:44] <nextgens> level1 kind of problems I should say
[10:44] <toad_> no
[10:44] <toad_> client layer is different
[10:44] <toad_> UOM uses a blob iirc
[10:44] <toad_> so it needs client layer
[10:45] <toad_> normal updating needs routing
[10:45] <nextgens> and client layer
[10:45] <toad_> right
[10:45] <nextgens> so in any case, if client layer is broken we are doomed
[10:45] <toad_> in terms of severe uptime problems, UOM may work when normal updating doesn't, because it's much faster
[10:46] <nextgens> unless we have request resuming ;)
[10:46] <toad_> maybe
[10:46] <nextgens> in which case it's less likely to succeed
[10:46] <toad_> what is?
[10:46] <nextgens> because blobs have to be transfered fully
[10:46] <nextgens> you can't resume a blob transfert, can you?
[10:47] <toad_> no
[10:47] <toad_> but resuming implies making the update a persistent download
[10:47] <toad_> and depends on more things -> is less reliable
[10:47] <nextgens> okay
[10:48] <nextgens> hmm, and what's the problem with updating only if the new version is self-mandatory again?
[10:48] <nextgens> even if the propagation is slow we shouldn't care
[10:48] <toad_> routing might be broken
[10:49] <nextgens> because we time-bomb normal mandatories
[10:49] <nextgens> if we break something the only thing we have to do is to release a self-mandatory to force all nodes to use UoM
[10:49] <nextgens> sure, but even if routing is broken nodes will eventually hear about the new version from their peers
[10:49] <toad_> right, but the logic to decide whether to use UOM doesn't know whether it is self-mandatory or not
[10:49] <nextgens> atm it's what happens
[10:50] <nextgens> the whole "normal" updating process isn't used at all
[10:50] <nextgens> because it's faster for the full network to update from uom
[10:50] <toad_> if routing is broken on every node, how is the process going to start?
[10:50] <nextgens> uom is about a peer sending announces when it connects
[10:50] <toad_> one node has to download it and advertise it ... the originator, of course
[10:50] <nextgens> iirc we do send what's "current mandatory" for this peer
[10:50] <toad_> the insertor node picks it up after it has inserted it
[10:51] <toad_> and then advertises it to its peers
[10:51] <nextgens> so it works; doesn't it?
[10:51] <toad_> afaics any change to UOM means less reliable updating means more likelihood of losing users
[10:52] <toad_> maybe we need some way to indicate to the node whether a build is urgent or not
[10:52] <nextgens> I would agree with "slower" updating but that's on purpose
[10:52] <nextgens> sorry, I gtg
[10:52] <toad_> ... and that mechanism is the mandatory build number ...
[10:52] <nextgens> that's what the mandatory thingy is about :)
[10:52] <nextgens> brb
[10:52] <toad_> gah, i was hoping to discusss FOAF routing countermeasures with you
[10:54] <toad_> nextgens: okay, so if the build is urgent, then the insertor will run it on the originator node, all his peers will be marked TOO NEW, he will insert it only into his own datastore, and then advertise it to his peers who will immediately deploy it through UOM
[10:54] <toad_> nextgens: if not, he can insert it
[10:54] * Luke771 (n=luke@) has joined #freenet
[10:54] <toad_> nextgens: okay, i agree with you; give it say an hour, but get it from UOM if it's self-mandatory
[10:55] <toad_> i.e. if lots of our peers are TOO NEW
[11:11] <toad_> nextgens: i've written to oskar and vive, hopefully they are awake :)
[11:13] * Doutordor (i=PJIRCWeb@) has joined #freenet
[11:14] * Doutordor (i=PJIRCWeb@) Quit (Client Quit)
[11:25] <nextgens> rehi
[11:25] <nextgens> cool
[11:26] * toad_ thinks we can 1) limit opennet peers' locations to 20, 2) limit darknet peers' locations to 50???, 3) limit any one node to 20% of outgoing requests
[11:26] <nextgens> I'm not convinced that 3) won't f*ck routing up
[11:26] <toad_> why would it
[11:26] <nextgens> and well, 2) sounds like completly arbitrary
[11:27] <toad_> well yeah it is
[11:27] <nextgens> moreover I doubt that it will be effective
[11:27] <nextgens> even on opennet
[11:27] <toad_> why?
[11:27] <nextgens> the properties of small world network is a lot of small links...
[11:27] <nextgens> meaning that most of my peer's peers and peer's locations will be "close" to mine
[11:27] <toad_> yes, and most incoming traffic is close to our specialisation
[11:28] <toad_> what the attacker wants is most of our traffic, or most of our locally originated traffic?
[11:28] <nextgens> ie: if I put broader offsets to locations I will catch more nodes with less locations
[11:28] <nextgens> depending on the attack scenario, he might need only one of them
[11:29] <toad_> he might only want local requests, right
[11:34] <toad_> does that mean that 3) doesn't work either?
[11:34] <toad_> because local requests may only be 20% of the total outgoing requests?
[11:37] <nextgens> I doubt that the naive implementation of 3) can work
[11:38] * nextgens found a bug in the UoM code
[11:38] <nextgens> I wonder how to fix it in a backward compatible way
[11:38] <nextgens> in UOMAnnounce messages we send pingTime and bandwidth
[11:38] <nextgens> both are stored as double on the node's side
[11:39] <nextgens> and casted to an int when then in the message
[11:39] <nextgens> that's obviously not working when the bandwidth limit is high
[11:39] <nextgens> what happens if I change the DMT message type on new nodes?
[11:39] <nextgens> it will break, right?
[11:40] <toad_> yes
[11:40] <nextgens> okay
[11:40] <toad_> what's the problem, ints go up to 4G
[11:40] <nextgens> so I'd better use Math.min(Integer.MAX_VALUE)
[11:41] <nextgens> toad_> I've got a 10Gb/s link here :p
[11:41] <toad_> if you cast a double which is bigger than MAX_VALUE, it will be cast down to MAX_VALUE
[11:41] <toad_> lucky you
[11:43] * nextgens didn't know that
[11:43] <nextgens> are you sure?
[11:43] <nextgens> it won't end up being a negative value
[11:44] * nextgens found a negative value in his logfiles
[11:44] <nextgens> arguably I did have a silly value for my bwLimit
[11:57] <toad_> :)
[12:12] * zerwas (n=zerwas@) has joined #freenet
[12:25] * tomahasamoot (n=kvirc@) Quit (Read error: 110 (Connection timed out))
[12:27] * tomahasamoot (n=kvirc@) has joined #freenet
[12:42] * caytchen (n=caytchen@) has joined #freenet
[12:45] * christefano (n=christef@) Quit ("time for a coffee")
[12:59] * archangelpetro (n=hellequi@) Quit ("Build a hundred bridges and suck a single cock, and you're not a bridge builder.. you're a cocksucker")
[13:10] * mrdmx-bbl is now known as mrdmx
[13:24] * Dieppe (n=clement@) Quit (Remote closed the connection)
[13:25] * Dieppe (n=clement@) has joined #freenet
[13:32] * zerwas (n=zerwas@) Quit (Remote closed the connection)
[13:50] * zerwas (n=zerwas@) has joined #freenet
[13:54] * DaffyDuck_ (n=jan@) Quit ("Leaving")
[14:33] * Mathiasdm (n=Mathias@) Quit ("Invisible Internet Project: http://www.i2p2.de")
[14:59] * zerwas (n=zerwas@) Quit ("Keep on flowing ...")
[15:22] <nextgens> rehi
[15:23] <toad_> rehi
[15:24] <nextgens> svn 1.5 has entered testing
[15:24] <nextgens> cool
[15:25] <nextgens> toad_> when is the db4o merge planned for? before/after sdiz's one?
[15:26] * nextgens would like to know how urgent is the upgrade of the repository
[15:26] <nextgens> apparently we have two options
[15:26] <nextgens> either we do an "upgrade"
[15:26] <nextgens> or we dump/reload the whole thing
[15:27] <nextgens> the second option allows more optimization
[15:27] <nextgens> but requires everyone to nuke his current working copy
[15:38] <toad_> dump/reload? not sure i understand
[15:38] <toad_> the current db4o code supports reading downloads.dat.gz and auto-migrating
[15:39] <toad_> it would probably be best to merge sdiz's code first since it's simpler
[15:41] <nextgens> toad_> to take advantage of the new features we have to convert the repository
[15:41] <nextgens> the svn guys have developped an upgrading tool
[15:41] <sanity_> gah, another 5 minute wait to reconnect to opennet :-/;
[15:41] <nextgens> but they recommand to use the current dump/reload functionnalities instead
[15:42] <nextgens> sanity_> is that 5mins of bootstrapping or 5mins for the node to start up because it's awaiting for entropy
[15:42] <nextgens> ?
[15:42] <sanity_> nextgens: no, the node is up, its 5 minutes to successfully reestablish a connection
[15:42] <sanity_> ok, got two now
[15:44] <nextgens> is your node doing any request?
[15:45] <nextgens> did you get rid of all bookmarks or something?
[15:45] <sanity_> yes, i'm running fms
[15:45] <sanity_> no
[15:46] <nextgens> and is it really "reestablishment of an existing connection" or "5mins awaiting for the seednode to return a peer which connects" ?
[15:46] <nextgens> iirc the opennet code keeps something like 50 "old" references
[15:46] <nextgens> and tries all of them before announcing
[15:46] <sanity_> if so, why does it take 5 minutes to test 50 references?
[15:47] <sanity_> especially since they could be tested in parallel
[15:47] <nextgens> hmm
[15:47] <nextgens> it's a bit trickier than that
[15:48] <nextgens> well, in theory they could
[15:48] <nextgens> but the current implementation of JFK doesn't allow it
[15:48] <sanity_> JFK?
[15:48] <nextgens> Just Fast Keying (the DiffieHellman scheme we use at link-layer)
[15:49] <sanity_> doesn't allow what exactly?
[15:49] <nextgens> if the authenticator's cache reach 30 entries we flush it, changing the key
[15:49] <nextgens> effectively preventing any connection from happening with those keys
[15:49] <nextgens> hmm
[15:50] <nextgens> well, in fact it's used only on the responder's side
[15:50] <nextgens> so it shouldn't prevent connections
[15:51] <nextgens> okay, here you've a point: we should try to handshake more aggressively when we don't have enough connections on opennet
[15:51] <nextgens> atm we don't
[15:52] <nextgens> probably because we wanted to avoid detection on isp's side
[15:52] <nextgens> aggressive handshake attemps could be regarded as "port scanning" or "flooding"
[15:55] <sanity_> it seems to me that its rather rare for pre-existing nodes to reconnect
[15:55] <sanity_> what about just re-announcing every time?
[15:56] <nextgens> that sucks because it would make freenet trivial to detect for an ISP
[15:57] <nextgens> of course we could have a setting for that
[15:57] <nextgens> but what matters is the default configuration value
[15:59] <nextgens> they are many optimizations we could do if we assume that we don't need to hide from our isp
[16:00] <nextgens> but imho it's not an assumption we can make
[16:00] <nextgens> sooner or later ISPs will want to fingerprint freenet's traffic
[16:00] <nextgens> at the moment it's rather hard to do
[16:02] <sanity_> is it?
[16:02] <sanity_> i'm not so sure about that, with a stateful packet monitor I think it would be quite easy to recognize Freenet traffic
[16:03] <nextgens> stateful packet analysis doesn't help
[16:03] <nextgens> we have no pattern to match
[16:03] <nextgens> we use random length packets
[16:03] <nextgens> we have various send-delays (including for handshakes)
[16:03] <nextgens> we use packet coalescing
[16:04] <nextgens> no, really it's not trivial to match
[16:04] <nextgens> the only "working" criteria I know of is "doesn't match something we know about" :p
[16:04] <nextgens> someone suggested using packet's entropy a while ago
[16:04] <sanity_> how long are the packet delays?
[16:04] <nextgens> that would work: freenet's packets entropy is high
[16:04] <sanity_> could that be why requests are so slow?
[16:05] <nextgens> higher than any protocol we found (including skype/ssh/...)
[16:05] <nextgens> I don't think so
[16:05] <nextgens> we coalescing is 200ms iirc
[16:05] <nextgens> but there is also a "size" trigger
[16:07] <nextgens> s/we/our/
[16:08] <nextgens> well 200ms is the base coalescing
[16:08] <nextgens> and we have a random modifier on top of it iirc
[16:20] <toad_> <sanity_> it seems to me that its rather rare for pre-existing nodes to reconnect
[16:20] <toad_> agreed
[16:21] * sanity (n=ian@) has joined #freenet
[16:21] * ChanServ sets mode +o sanity
[16:21] <toad_> random delays are on handshaking, not on data iirc
[16:21] <toad_> i dunno whether we randomize the coalescing delay or not off hand
[16:22] <toad_> we do need to deal with the entropy issue
[16:25] <toad_> and the amount of padding we do is a performance issue, but there are ways to deal with that (streams)
[16:25] <toad_> anyway, i have no idea why it takes so long to seed a new node / reseed an old one
[16:26] <toad_> traditionally the answer has been "you ****ing idiot, you should run 24x7 or the whole network will implode!!12121!!!!"
[16:26] <toad_> :)
[16:27] <toad_> it might be a shortage of seednodes again... some of them may have moved ...
[16:27] <toad_> nextgens: weren't you working on a seednode validator tool?
[16:27] <toad_> or was i?
[16:27] * wingnutamj (n=chatzill@) has joined #freenet
[16:28] <nextgens> toad_> I was working on a seednode flooding tool iirc
[16:29] <nextgens> but I guess it could serve both purposes
[16:29] * nextgens doesn't remember how far he got on that
[16:30] <TheSeeker> Lst time I announced, I think I announced to all of 2 IPs.
[16:30] * sanity_ (n=ian@) Quit (Read error: 113 (No route to host))
[16:31] <nextgens> jvm 1 | We received a valid UOMAnnounce : (isOutdated=false version=1154 whenToTakeOverTheNormalUpdater=-3588145
[16:31] <nextgens> jvm 1 | Starting the update process: found the update (1154), now fetching it.
[16:31] <nextgens> jvm 1 | Starting fetch for 1154
[16:31] <nextgens> yay
[16:31] * sanity (n=ian@) Quit ()
[16:31] <nextgens> let's see if the normal updater still works :)
[16:44] * sanity (n=ian@) has joined #freenet
[16:44] * ChanServ sets mode +o sanity
[16:47] <FreenetLogBot> r21289 (1154) was built successfully on emu, mirrors are updating
[16:48] * zerwas (n=zerwas@) has joined #freenet
[16:50] <nextgens> okay, worksForMe(TM)
[16:52] <zerwas> ™
[17:10] * Mathiasdm (n=Mathias@) has joined #freenet
[17:15] * zerwas (n=zerwas@) Quit ("Keep on flowing ...")
[17:19] * sanity_ (n=ian@) has joined #freenet
[17:19] * ChanServ sets mode +o sanity_
[17:32] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[17:38] * sanity_ (n=ian@) Quit ()
[17:52] * sanity (n=ian@) has joined #freenet
[17:52] * ChanServ sets mode +o sanity
[17:57] * Johan^mlg (n=bllarf@) has joined #freenet
[18:00] * wingnutamj (n=chatzill@) Quit ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]")
[18:24] * NEOatNHNG (n=michael@) has joined #freenet
[18:32] * christl_s (i=christin@) Quit (Read error: 104 (Connection reset by peer))
[18:32] * christl_s (i=christin@) has joined #freenet
[18:40] * christl_s (i=christin@) Quit (Read error: 104 (Connection reset by peer))
[18:44] * mrdmx is now known as mrdmx-bbl
[18:54] * mrdmx (i=mrdmx@) has joined #freenet
[18:55] * mrdmx-bbl (i=mrdmx@) Quit (Read error: 104 (Connection reset by peer))
[19:06] * Nande (n=KVIrc@) has joined #freenet
[19:10] * sbc (n=ca@) has joined #freenet
[19:10] * Iso-Nisse (n=iso_niss@) Quit (Read error: 104 (Connection reset by peer))
[19:11] * zerwas (n=zerwas@) has joined #freenet
[19:11] * DaffyDuck_ (n=jan@) has joined #freenet
[19:17] * FrinkC (n=FrinkC@) has joined #freenet
[19:18] * Iso-Nisse (n=iso_niss@) has joined #freenet
[19:28] * sanity (n=ian@) Quit ()
[20:03] * saces (n=saces@) Quit (Read error: 110 (Connection timed out))
[20:03] * saces (n=saces@) has joined #freenet
[20:03] * ChanServ sets mode +o saces
[20:10] * Krhis (n=Krhis@) has joined #freenet
[20:26] * Mathiasdm (n=Mathias@) Quit ("Invisible Internet Project: http://www.i2p2.de")
[20:31] * Luke771 is now known as luke771_afk
[20:39] * batosai (n=batosai@) Quit ("KVIrc 3.4.0 Virgo http://www.kvirc.net/")
[21:37] * Johan^mlg (n=bllarf@) Quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
[21:43] * luke771_afk (n=luke@) Quit ("Leaving")
[21:45] * FrinkC (n=FrinkC@) Quit ("Bye")
[21:48] * Zarggg (n=z@) has joined #freenet
[22:05] * JustMe (n=me@) Quit ()
[22:30] <toad_> rehi
[22:36] <Tommy[D]> huh, how comes a peer of me has many locations shown?
[22:37] <toad_> how many?
[22:38] <Tommy[D]> 1 bold, 9 italic
[22:41] <nextgens> toad_> how can I check that a connection to a seednode is alive? sending a ping?
[22:42] * nextgens dug out his seednode pining code
[22:42] <nextgens> pinging
[22:42] <nextgens> of course I can make it announce
[22:42] <nextgens> but it will waste announcements, won't it?
[22:46] <nextgens> toad_> regarding the jfk's NPE: the code already exist in trunk :)
[22:48] <toad_> cool
[22:48] <toad_> pinging it would be fine yeah
[22:48] <toad_> for a first approximation
[22:50] <toad_> our initial problem is simply how to be notified when a seednode goes offline, and the second problem is automatic movement via ARKs
[22:57] * nOgAnOo (i=nOgAnOo@) Quit (Client Quit)
[23:02] * sbc (n=ca@) Quit ("Ex-Chat")
[23:10] <nextgens> heh
[23:10] <nextgens> if you think I'm gonna implement seednode harvesting you're dreaming
[23:11] <nextgens> the only thing I might do is a seednode flooding tool
[23:11] <toad_> :)
[23:11] <nextgens> what I have commited doesn't work as expected
[23:12] <nextgens> it connects to seednodes
[23:12] <nextgens> but then they don't reply to pings
[23:12] <nextgens> and it throws a NotConnectedException
[23:12] <toad_> curious...
[23:12] <nextgens> and does announce
[23:12] <toad_> ah
[23:12] <toad_> is ping handling conditional on being ReallyConnected/Routable ??
[23:12] <FreenetLogBot> r21299 (1154) was built successfully on emu, mirrors are updating
[23:12] <toad_> check NodeDispatcher
[23:13] * ahuxley (n=ahuxley@) has joined #freenet
[23:13] <nextgens> it should be handled even by them, yeah
[23:13] <toad_> hmmm
[23:13] <toad_> i dunno then until i get to the code...
[23:13] <nextgens> btw, why do we have two different kind of pings? :)
[23:13] <nextgens> LinkPingPong and LinkPing
[23:14] <toad_> good question
[23:14] <toad_> LinkPingPong?
[23:15] <nextgens> we have FNPPing and FNPLinkPingPong in DMT.java
[23:19] <toad_> ah
[23:19] <toad_> yeah
[23:19] <toad_> i think they may be different levels
[23:19] <toad_> that doesn't necessarily mean they are both necessary though
[23:21] * nextgens double checks he is sending the right one
[23:22] <nextgens> both should be handled
[23:25] <nextgens> okay, let's get rid of one of them
[23:26] * roo9 (i=roo9@) Quit (Remote closed the connection)
[23:27] <toad_> good idea
[23:31] <FreenetLogBot> r21300 (1154) was built successfully on emu, mirrors are updating
[23:36] <nextgens> bbl
Irc logs of #freenet : 2008 2007 2006 2005
These logs were automatically created by FreenetLogBot on chat.freenode.net using the Java IRC LogBot.