Timestamps are in GMT/BST.
[0:03] * sanity (n=ian@) has joined #freenet
[0:03] * ChanServ sets mode +o sanity
[0:09] * christefano_ (n=christef@) has joined #freenet
[0:22] * zerwas (n=zerwas@) has joined #freenet
[0:27] * christefano (n=christef@) Quit (Connection timed out)
[0:32] * ahuxley (n=ahuxley@) Quit ()
[0:40] * christefano_ (n=christef@) Quit (Connection timed out)
[0:42] * zerwas (n=zerwas@) Quit ()
[0:45] * TranceAm (n=tranceam@) has joined #freenet
[0:46] <TranceAm> Eh, any admins hanging around?
[0:46] * TranceAm is 8/ the ~~~~ of Tenthu on Purple Sands (Original Mix) from Purple Sands dated 2008
[0:47] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[0:50] * TranceAm (n=tranceam@) has left #freenet
[2:18] * gasi_ (n=gasi@) has joined #freenet
[2:27] * christefano (n=christef@) has joined #freenet
[2:32] * gasi (n=gasi@) Quit (Read error: 110 (Connection timed out))
[3:16] * StefftheEmperor_ (n=Steffthe@) has joined #freenet
[3:16] * caytchen (n=caytchen@) Quit ("Verlassend")
[3:32] * StefftheEmperor (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[3:54] * Zarggg (n=z@) Quit ()
[4:44] * Vooch (n=IceChat7@) has joined #freenet
[4:45] * Nande (n=KVIrc@) Quit ("byes, tengan la bondad de ser felices. ;)")
[5:09] * Vooch (n=IceChat7@) Quit ("Not that there is anything wrong with that")
[5:58] * christefano (n=christef@) Quit ()
[6:46] * FrinkC (n=FrinkC@) has joined #freenet
[6:56] * Hikaru`_ (n=plop@) has joined #freenet
[7:32] <FreenetLogBot> r21659 (1155) was built successfully on emu, mirrors are updating
[7:50] * n0ob (n=travis@) Quit ("Now onto something more interesting")
[7:55] * TheBishop_ (n=TheBisho@) Quit (Remote closed the connection)
[8:38] * radden (n=radden@) has joined #freenet
[8:42] <radden> Just started using freenet... installed frost but can't seem to get any messages or connect to any boards
[8:45] <radden> am i going about this the right way? should i just post questions like this?
[8:45] <dbkr> do the boards show up red?
[8:46] <dbkr> if they do, frost has detected that they're being DoSed
[8:46] <dbkr> (I think)
[8:46] <radden> no, not red
[8:46] <dbkr> frost isn't really used anymore because of this vulnerability
[8:46] <radden> should i use fms?
[8:47] <dbkr> probably a better bet
[8:48] <radden> so when its working properly freenet feels a bit like using newsgroups?
[8:48] <radden> despite reading all the help i could find i'm still not sure what to expect
[8:49] <dbkr> FMS does, not freenet
[8:49] <dbkr> is that what you meant?
[8:50] <radden> sort of.... i'm not sure how the 2 interact / relate to each other
[8:50] <dbkr> well Freenet is just a networked data store
[8:51] <dbkr> various apps run on top of Freenet to provide different services
[8:51] <radden> and fms is how you interact with other users of freenet
[8:52] <dbkr> fproxy (the web interface) is the main one, and is actually built into Freenet
[8:52] <dbkr> FMS is one (relatively new) application
[8:53] <radden> i had a browse around the web interface, it had a few lists of indexes, flogs etc
[8:55] <radden> but i wanted a way to search around what other users had
[8:59] <dbkr> ah, yeah - search isn't an easy thing to do on Freenet
[9:00] <dbkr> you could check out the XMLLibrarian though, I don't know what it's like nowadays
[9:00] <dbkr> it's a plugin that downloads search indexes
[9:01] <radden> so do you try to find groups of users who have the same interests
[9:01] <radden> and then exchange stuff with them?
[9:03] <dbkr> yep, that works well
[9:03] <radden> thanks for your help by the way
[9:04] <dbkr> Freenet isn't really one of those pieces of software that offers you something that'll make your life better, more a way of solving a problem you might already have :)
[9:04] <radden> so how do you go about finding stuff like that?
[9:04] <dbkr> no problem :)
[9:05] <dbkr> frost used to be fairly good for browsing content, nowadays FMS might not be bad at the samr job
[9:06] <dbkr> index sites are not bad as well, but suffer a bit from only being maintained by one person, which is fine until said person goes and gets a girlfriend (or boyfriend)
[9:06] <radden> well yeah, isn't that problem usually anonymity
[9:07] <dbkr> I was thinking of specifically what you want to do anonymously
[9:07] <dbkr> it doesn't offer 'anonymity' so much as a means of doing things anonymously
[9:08] <radden> doing things apart from file swapping?
[9:08] <radden> and chatting etc
[9:10] <dbkr> hum, I was about to suggest http://wiki.freenetproject.org/FreenetZeroPointSevenApps, but it's horrendously out of date
[9:12] <radden> i'm downloading fms now
[9:12] * sanity (n=ian@) has joined #freenet
[9:12] * ChanServ sets mode +o sanity
[9:19] * infinity0 (n=infinity@) has joined #freenet
[9:21] * infinity0 (n=infinity@) Quit (Read error: 104 (Connection reset by peer))
[9:22] * infinity0 (n=infinity@) has joined #freenet
[9:24] <radden> fms won't install.. is says side by side configuration is incorrect... any idea what that means?
[9:27] <dbkr> hmm, absolutely no idea whatsoever, I'm afraid
[9:27] <dbkr> that's a wonderfully meaningless error message
[9:32] <radden> when i go to download fms through fproxy it has a few different files to download... binary, source & runtime files
[9:32] <radden> have you installed fms? should it just be a simple program install?
[9:34] <dbkr> I compiled it from source, I haven't tried any of the binaries
[9:35] <radden> ahhh... it needed both binary and runtime files
[9:36] <radden> i'm probably boring you to death... so do you hang around here to help people out?
[9:37] <dbkr> more-or-less. I maintain Freemail, so it's also useful to be on here to discuss things
[9:43] <radden> so i've opened up the fms web interface, how do find some boards
[9:44] <dbkr> I think it should find some automatically after a while
[9:44] * skxpl (i=skx@) Quit ("changing servers")
[9:47] * batosai (n=julien@) has joined #freenet
[9:47] * ChanServ sets mode +v batosai
[9:49] <radden> there is a list of boards, science, police etc but its just a list in html format, not something i can read and post to
[9:52] * infinity0 (n=infinity@) Quit ("Leaving.")
[9:55] <dbkr> yeah, you can't post or read messages from the web interface
[9:59] <radden> can you recommend any interesting boards to add?
[10:03] <dbkr> depends what you're interested in :)
[10:05] <FreenetLogBot> r21660 (1155) was built successfully on emu, mirrors are updating
[10:10] * skxpl (i=skx@) has joined #freenet
[10:28] * toad_ (n=toad@) has joined #freenet
[10:28] * ChanServ sets mode +o toad_
[11:04] <FreenetLogBot> r21662 (1155) was built successfully on emu, mirrors are updating
[11:07] <FreenetLogBot> r21663 (1155) was built successfully on emu, mirrors are updating
[11:07] <FreenetLogBot> r21664 (1155) was built successfully on emu, mirrors are updating
[11:08] * nextgens is reluctant to wire-in the changes to FOAF before it has been tested with appropriate metrics
[11:08] <nextgens> but now it's trivial to do
[11:11] <nextgens> toad_> you might want r21663 in your branch as well... unless you've already done something about it of course
[11:14] * radden (n=radden@) Quit ()
[11:20] <toad_> nextgens: agreed, we need the testing code
[11:20] <toad_> sback said he might do that, said he'd say something today
[11:22] <nextgens> toad_> so, what was the final word again? I don't remember:
[11:22] <toad_> on FOAF?
[11:22] <nextgens> limit to 19 foaf-locations per peer on opennet ....
[11:22] <nextgens> and 30% of requests for all peers, on both darknet and opnennet
[11:23] <nextgens> missroute if needed
[11:23] <toad_> yeah
[11:23] <toad_> but 30% of *locally originated* requests
[11:23] <toad_> possibly with a remote limit as well
[11:23] <nextgens> I've implemented the code to average on 5mins
[11:23] <nextgens> I've implemented the code to average on 5hah
[11:23] <nextgens> well, that's not what the current code does then
[11:23] <kork> nextgens: just a quick question, do you know anything about the legality of freenet in france? I seem to recall cryptography was outlawed, or something :/
[11:24] <nextgens> kork> it's not anymore for our purpose (since 2001)
[11:24] <nextgens> but that doesn't mean that freenet is legal in France :)
[11:24] <toad_> but DADVSI may be a problem
[11:24] <nextgens> afaic it's not because of DADVSI
[11:24] <nextgens> yeah
[11:25] <nextgens> and like if it wasn't enough they are going to add yetAnotherLayer of bullshitIPLaw soon
[11:25] <nextgens> probably early next year
[11:25] <kork> nextgens: hmm :) I'll start running a node on a server hosted in france (OVH), but since I rented it as a german from a german office of OVH, I'm not sure french law even applies. guess I'll just run it and wait till the FBI kicks in my doors :D
[11:25] <toad_> since merging the trunk changes, the db4o node has had a lot more object churn, a lot more GC, and generally gotten bogged down
[11:26] <toad_> did anything happen on trunk that might have caused severe performance problems / more object churn?
[11:26] <toad_> since i branched 2 months 2 weeks ago?
[11:26] <nextgens> kork> I'm aware of at least 5 nodes running in their datacenter :)
[11:27] <toad_> kork: does the german data retention de-anonymizer law apply to freenet nodes?
[11:27] <nextgens> yours won't be alone if that reassures you
[11:27] <kork> toad_: I don't have the slightest idea, really :(
[11:27] <toad_> :|
[11:27] <toad_> might be interesting to find out :|
[11:27] <nextgens> kork> and there is no FBI here :)
[11:27] <kork> nextgens: yeah, I guess there's quite a bunch of tor nodes there, too, with their traffic flat rate and everything
[11:27] <toad_> nextgens: can you be arrested in one Schengen state for a crime committed in another that isn't a crime in the one you're hiding in?
[11:28] <nextgens> I don't think so
[11:28] <toad_> I know they can pursue you across borders ... but surely they have no power of arrest in another country if it's not a crime there?
[11:28] <nextgens> and even... it doesn't qualify as a crime anymore here
[11:28] <nextgens> thanks to DADVSI
[11:29] <nextgens> but well, I'm not up to date anymore on the local IP laws
[11:29] <nextgens> they are evolving way too fast
[11:29] <nextgens> and generally speaking in the wrong direction imo
[11:29] <toad_> well DADVSI does lay out some criminal offences that might apply to us ...
[11:29] <toad_> they're probably not arrestible in any case though
[11:30] <toad_> so I don't need to keep out of the entire schengen area just because some devs i work with are in france and flagrantly violating local law?
[11:30] <nextgens> toad_> let me think about what went in since 1153
[11:30] <toad_> nextgens: please do
[11:30] <toad_> jmap -histo isn't very helpful, i'll probably have to profile trunk
[11:30] <toad_> that could take significant time
[11:31] <nextgens> what I've just commited introduces more churn for sure but it hasn't been merged yet
[11:36] <toad_> nextgens: it's not necessarily just churn, it's memory usage in general too
[11:36] <toad_> eventually (within 12 hours or so) it does OOM
[11:37] <nextgens> so, we are looking at a leak ?
[11:37] <toad_> very possible
[11:37] <toad_> i can't rule out one in the db4o code
[11:37] <toad_> but it seems much worse since merging
[11:38] <nextgens> what I've just commited is memory hungry but definitly not leaking
[11:40] <nextgens> toad_> my guess goes to the chainnedbucket thingy
[11:40] <nextgens> it's the only thing which has changed on the client-layer
[11:40] <toad_> nextgens: i already had that
[11:40] <toad_> i'm not sure it's a client layer problem
[11:41] <nextgens> hmm
[11:41] <toad_> cos the client layer is all different on the db4o branch anyway, very few changes kept from trunk in that
[11:41] <nextgens> well, the rest is just cosmetic changes & bugfixes
[11:41] <toad_> hmmm well maybe i'm wrong then ...
[11:41] <toad_> there doesn't seem to be anything obvious from the client layer occupying lots of RAM in the jmap -histo dumps
[11:42] <nextgens> unless you use heavily plugins & fproxy
[11:42] <toad_> but memory usage continues to increase ...
[11:42] <nextgens> I don't get why we would want to make the distinction in between locally originating requests and remote ones
[11:43] <nextgens> imo we need to apply the same rule for both
[11:43] <toad_> nextgens: because if we don't make the distinction, the countermeasure is useless
[11:43] <nextgens> otherwise it's yetAnotherDiscrimitator you can use to do correlation attacks
[11:43] <nextgens> sure
[11:44] <nextgens> okay, fair point
[11:44] <toad_> nextgens: locally originated requests are evenly distributed; remotely originated requests are specialised
[11:44] <nextgens> but my bet is that it's not of much use in any case
[11:44] <nextgens> nah
[11:44] <toad_> therefore, if we don't make a distinction, an attacker could use its 30% quota to get almost all locally originated requests
[11:44] <nextgens> that's not what the stats say
[11:44] <toad_> oh? you said the stats said remote requests were specialised?
[11:44] <nextgens> apparently both are almost evenly distributed
[11:45] <nextgens> incomming yes
[11:45] <nextgens> outgoing no
[11:45] <toad_> Outgoing Local Request Distribution
[11:45] <toad_> shows up as evenly distributed
[11:45] <nextgens> because the cache is effective
[11:45] <toad_> more or less
[11:45] <toad_> on my node
[11:45] <nextgens> here too
[11:45] <toad_> nextgens: woah
[11:45] <nextgens> but Outgoing Request Distribution is too
[11:45] <toad_> i see what you mean ... that's really impressive!
[11:45] * toad_ sees it on his node ... awesome
[11:45] <nextgens> or a bug :)
[11:46] <nextgens> hence I asked if you did review the stats' code
[11:46] * toad_ sees the same specialised distro outgoing as incoming ... except most of the central peak is gone
[11:46] <toad_> because the cache WORKS
[11:46] <toad_> anyway, it is still essentially specialised
[11:46] <toad_> although not very strongly
[11:47] <nextgens> that's also why I don't think that FOAF is gonna help much :)
[11:47] <toad_> w.r.t. discriminators, what about ignoring whether the request is *actually* local, and figuring out whether to treat it as such based on its location in the keyspace vs our specialisation?
[11:47] <toad_> nextgens: eh?
[11:47] <toad_> nextgens: why do you think FOAF won't help much?
[11:47] <nextgens> and why I think we *NEED* simulations before switching to a non-LRU caching scheme
[11:47] <toad_> nextgens: we've had them
[11:47] <toad_> nextgens: mrogers committed his simulations
[11:48] <nextgens> because cache can satisfy most of specialized requests
[11:48] <nextgens> foaf helps cutting down path of specialized requests
[11:48] <nextgens> ie: it's more effective as you get closer to the target
[11:48] <nextgens> toad_> he simulated only the store
[11:48] <toad_> no, read the code
[11:48] <toad_> he assumed the store, and simulated the cache
[11:49] <toad_> the store is fixed, an insert goes to one node's store, the most specialised node
[11:49] <nextgens> ah?
[11:49] <toad_> and then he simulates the cache for various cache types
[11:49] <nextgens> I'll have a look
[11:49] <toad_> please do
[11:50] * Luke771 (n=luke@) has joined #freenet
[11:50] <nextgens> what about introducing a kill switch for the cache?
[11:51] <nextgens> so that we can see if the statistic code works
[11:51] <toad_> ????
[11:51] <toad_> well, just delete it
[11:51] <nextgens> If we disable caching we should see a difference in between locally originating requests and non-local ones
[11:51] <nextgens> yeah or I can do that too
[11:51] <toad_> you don't?
[11:52] <nextgens> well, it's not big enough imho
[11:52] <toad_> i see a reasonably specialised outgoing distro but with the middle spike truncated to below its direct neighbours
[11:52] <nextgens> if the cache is too effective we will eventually end up transmitting only locally originating requests
[11:52] <nextgens> which is good in terms of efficiency
[11:52] <toad_> but the spike is at 0.6 and there's a large factor between 0.5/0.7 and 0.1/0.0/0.2
[11:52] <nextgens> but bad for anonymity
[11:53] * sbc (n=ca@) has joined #freenet
[11:55] <nextgens> maybe we should supperpose the two histograms
[11:56] <toad_> that would be interesting yeah
[11:57] <toad_> w.r.t. anonymity, if finding stuff takes less hops, we can spend more hops on other things such as tunnels
[11:58] <nextgens> sure
[11:58] <nextgens> but we have to take care about the scalling effect
[11:58] <toad_> hmm?
[11:59] <nextgens> if cache helps on a small network with not much content it won't be as affective on a big one
[12:00] <nextgens> anyway
[12:00] <nextgens> toad_> is there anything I should spend time on?
[12:00] <nextgens> I've fixed the "installer distributes update.sh even on windows bug"
[12:01] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[12:01] <nextgens> I could implement the remaining bits of the code we will need for foaf
[12:01] * cdent (n=cdent@) has joined #freenet
[12:01] <toad_> nextgens: if it's highly specialised then it will help a lot on a big network
[12:01] <nextgens> but I'm not convinced that the design I plan to use is sound
[12:02] <toad_> nextgens: good idea
[12:02] <toad_> what design?
[12:02] <nextgens> atm I sore in a sorted map timestamps of when a request has been done (one map per peernode)
[12:03] <nextgens> so that I can get a quick estimate of how many samples I have over a period of time
[12:03] <toad_> ah
[12:03] <toad_> well that's not an estimate, it's precise
[12:03] <toad_> but imho that sort of impl will lead to oscillations and so on
[12:03] <nextgens> but now I just learnt that I need to differentiate local vs remote
[12:03] <toad_> some sort of decaying average is preferable
[12:03] <nextgens> the easiest way of doing it is to keep two maps
[12:04] <nextgens> oscillations?
[12:04] <toad_> yeah
[12:04] <nextgens> how could it oscillate?
[12:04] <toad_> you get a spike, it takes a certain amount of time to leave the averaging interval
[12:04] <nextgens> I've a bigLock on PeerManager
[12:04] <toad_> and then the value goes down, more requests are allowed, you get another spike
[12:05] <nextgens> ah
[12:05] <toad_> we've had this before, it's a classic problem in lots of fields
[12:05] <toad_> so something based on a klein filter is probably preferable
[12:05] <nextgens> yeah, of course that's what gonna happen when we will missroute
[12:05] <toad_> a BDRA on the request interval / request rate (convert it before reporting it)
[12:06] <toad_> would be much less overhead, and wouldn't oscillate
[12:06] <toad_> well it would oscillate much less
[12:06] <nextgens> atm the stats were meant to be displayed, not used in the routing algorithm :)
[12:06] <toad_> :)
[12:06] <nextgens> BDRA ?
[12:06] <toad_> BootstrappingDecayingRunningAverage
[12:06] <nextgens> ??RunningAverage
[12:07] <nextgens> ok
[12:07] <toad_> we also have TDRA, but i dunno how well that works, there have been doubts about it over the years
[12:07] <nextgens> and what do we want here?
[12:08] <nextgens> I thought about something time-based
[12:08] <toad_> yeah either use a BDRA on the rate, or use a TDRA
[12:09] <nextgens> and how do you cheaply compute the rate for all peers?
[12:09] <toad_> add them all up?
[12:09] <nextgens> I mean, we know when we have a tick
[12:09] <toad_> isn't that cheap enough?
[12:09] <nextgens> it's O(N), not exactly cheap
[12:09] <toad_> N is a very small number
[12:09] <toad_> and BDRAs return a result very quickly
[12:10] <toad_> it's nothing, don't worry about it
[12:11] <toad_> one worry with BDRA on rate is that a node might get a request after a short period and then never get one again ... with a BDRA we'd probably treat that as a high rate because of the short initial interval ...
[12:11] <toad_> i dunno how we've handled this elsewhere
[12:11] <nextgens> and what about debugging/fixing TDRA ?
[12:11] <toad_> great idea :)
[12:12] <nextgens> I mean, were the concerns on the underlying idea or on the implementation itself?
[12:12] <toad_> the easiest thing would be to have a BDRA on the rate *and* keep the time-last-seen, so we can ignore the BDRA if the time-last-seen is past some period
[12:12] * nextgens doesn't remember of any story about TDRA
[12:13] <toad_> it's probably okay, it's just that it's home grown statistics; it's had mathematical bugs in the past; its behaviour hasn't always been what we'd want
[12:14] <toad_> both BDRA and TDRA in this application have the issue i just mentioned, i.e. they don't tick, they only get updated when a sample comes in
[12:14] <toad_> when it does it takes into account the ticks they would have done ...
[12:15] <toad_> i do think that keeping every sample in the interval and working on that basis would lead to oscillations; we do need something smoother
[12:15] <nextgens> that sounds way too complex compared to the gain imo
[12:15] <nextgens> we are talking about a boder case here
[12:16] <toad_> it would certainly be possible to make currentValue() do a calculation to work out what the current value *should* be given no samples have come in in the last time...
[12:16] <nextgens> no peer should get 30% of the requests in normal circumpstances
[12:16] <toad_> true
[12:16] <nextgens> imo we don't even need such a scheme
[12:16] <toad_> well, if you want simple, report the rate to a BDRA, record the last time we sent a query as well, and return 0 if the last time is over some threshold, else the value of the BDRA
[12:16] <nextgens> if we are under active attack we are doomed anyway
[12:16] <toad_> that should work adequately
[12:17] <toad_> and not oscillate, as long as the threshold is high enough
[12:17] <toad_> i dunno whether rate or interval gives more accurate results, you'll have to guess there ... i think mostly we use interval in other places though
[12:18] <nextgens> okay, I'll see what I can come up with
[12:19] <nextgens> next topic: what's next task?
[12:19] <toad_> you mean after you've implemented everything necessary for FOAF?
[12:19] <nextgens> you've found someone to implement the test framework, you work on your branch...
[12:19] <nextgens> what do we want to go into trunk next?
[12:19] <toad_> turn it on i suppose ... but we need the test framework first
[12:20] <toad_> if sback doesn't do it then i will
[12:20] <nextgens> well, that's a two lines patch anyway
[12:20] <nextgens> what else?
[12:20] <toad_> nextgens: any important bugs?
[12:20] * infinity0 (n=infinity@) has joined #freenet
[12:20] <nextgens> hmm, I've fixed those I was aware of
[12:20] <nextgens> the windows updating script would need some work
[12:20] <nextgens> but I'm not gonna do it
[12:20] <toad_> hmmm the bug tracker is sorting by bug number ... wtf?
[12:21] <toad_> nextgens: several people have reported that the windows updating script contains the unix updating script
[12:21] <nextgens> I can't do the needed administration on emu because of my crappy inet. connection
[12:21] <toad_> nextgens: any idea?
[12:21] <nextgens> it's fixed
[12:21] <toad_> ok
[12:21] <toad_> mantis is still sorting by bug number
[12:21] <toad_> even though it's set to last update ascending
[12:21] <toad_> has something broken on the server?
[12:23] <nextgens> dunno
[12:23] <nextgens> works for me
[12:23] <nextgens> I sort by "updated" here
[12:23] <toad_> it makes it very difficult to see what bugs have been used recently!
[12:24] <toad_> ah
[12:24] <toad_> last update *DESCENDING*
[12:24] <toad_> doh
[12:24] <toad_> nextgens: are you interested in tackling streams?
[12:24] * caytchen (n=caytchen@) has joined #freenet
[12:25] * toad_ has a look at the post-summit todo...
[12:25] <nextgens> I could do that
[12:25] <nextgens> but imho we need to do it on a branch
[12:25] <toad_> what happened to Mercurial?
[12:25] <nextgens> either that or working on the content-filter
[12:25] <nextgens> well, I need to make the repository conversion on a clean-room
[12:26] <nextgens> and I can't get enough memory on emu to do it without taking down critical services
[12:26] <toad_> :|
[12:26] <nextgens> ...
[12:26] <nextgens> and of course I don't have enough ram to do it locally
[12:26] <toad_> maybe you should do it anyway
[12:26] <toad_> servalan has enough memory
[12:27] <toad_> it's not been terribly stable recently mind you
[12:27] <toad_> okay, lets see, summary of conclusions from the freenet mini-conf:
[12:27] <nextgens> I need a proper inet connection to do it anyway
[12:27] <toad_> 3 things for vive to do
[12:27] <toad_> plus...
[12:27] <toad_> finish the db4o branch, FOAF
[12:27] * nextgens is holding his laptop to the balcony in order to surf here
[12:27] <toad_> installer usability issues
[12:28] <toad_> non-convergent splitfile encryption
[12:28] <toad_> that i suppose will probably have to wait until db4o lands, since it will involve significant client layer changes??
[12:28] <nextgens> yeah
[12:28] <toad_> throttle *EVERYTHING*
[12:28] <nextgens> not a good idea to tackle it now
[12:28] * toad_ wonders if that can be done in isolation
[12:28] <toad_> if so, it's a great mini-feature
[12:28] <nextgens> throttling is your domain :)
[12:28] <toad_> oh?
[12:28] <nextgens> but imho it's not as important as the new packet format or streams
[12:29] <toad_> you don't like throttling?
[12:29] <nextgens> well, sure I could do it
[12:29] <toad_> imho throttling everything would be a key part of 0.8's "reduce the footprint" agenda
[12:29] <toad_> other parts include reducing memory and cpu usage
[12:29] <nextgens> but I'm not up to date yet with the math code behind... I've just learnt about klein filters :)
[12:29] <toad_> whether these are contradictory ... i dunno
[12:29] <toad_> nextgens: throttling can be hard
[12:30] <toad_> nextgens: it's not meant to be a long term average, it's more like what TCP does
[12:30] <toad_> we want to throw everything into congestion control too
[12:30] * infinity0 (n=infinity@) Quit ("Leaving.")
[12:30] <toad_> we have a token bucket now, but we only use it for Big Packets
[12:30] <nextgens> when is work on upper level stuff planned for?
[12:30] <toad_> reinstate uninstall survey, automatic bandwidth limit calibration ...
[12:31] <nextgens> the content filter could use some work imo
[12:31] <toad_> that's the end of the 0.8 todo
[12:31] <toad_> nextgens: if you're interested in the content filter then by all means get on with it
[12:31] <nextgens> automatic bandwidth limit calibration is bullshit and won't work anyway
[12:31] <toad_> nextgens: well what about auto-calibrating from UPnP then?
[12:31] <nextgens> the uninstall survey is something I could do if I had a stable connection
[12:31] <nextgens> toad_> when up&p works it's doable and that would work
[12:32] <toad_> nextgens: well, you can use SVK for most of this ...
[12:32] <nextgens> toad_> I'm more into streams & new packet format
[12:32] <toad_> nextgens: streams and new packet format would be awesome
[12:32] <toad_> i do think that throttling everything is higher priority
[12:33] <nextgens> the problem is that it would take me ages to find out what we aren't thottling yet ... whereas you know it :)
[12:33] * cdent (n=cdent@) Quit ()
[12:33] <toad_> well yeah but you'll have to do the same work if you tackle streams
[12:34] <toad_> you might be able to do new packet format without touching throttling if you're lucky
[12:34] <toad_> but streams is definitely going to involve throttling work
[12:34] * infinity0 (n=infinity@) has joined #freenet
[12:35] * 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.")
[12:35] <nextgens> what about implementing vivee's new swapping algorithm so that we could test it with the real code?
[12:35] <nextgens> or is that premature? (not even worht worth simulating)
[12:36] <toad_> i dunno
[12:36] <toad_> it's worth simulating certainly
[12:36] <toad_> vive is supposed to be doing that, and has turned up again recently ...
[12:36] <toad_> if you want to work on it then go ahead, but don't deploy it yet
[12:37] <toad_> really we'd want a simple simulation though, something we could simulate a fair sized network on
[12:37] <nextgens> well, I'll let vivee do it then
[12:37] * nextgens would have done it in node.simulator with real nodes
[12:39] <nextgens> bbiab
[12:39] * infinity0 (n=infinity@) Quit (Remote closed the connection)
[12:39] * infinity0 (n=infinity@) has joined #freenet
[12:41] <toad_> bbiab
[12:56] <vivee> nextgens: i am working on it parttime, please ask me if you need any help doing it with the real code
[12:56] <vivee> are more seednodes needed at the moment?
[13:45] * infinity0 (n=infinity@) Quit ("Leaving.")
[13:58] * Hikaru`_ (n=plop@) Quit (Read error: 104 (Connection reset by peer))
[13:59] * Hikaru`_ (n=plop@) has joined #freenet
[14:02] <toad_> vivee: yes
[14:08] <toad_> vivee: imho the opennet swapping issue is more urgent than targeted swapping
[14:14] * kork (n=kork@) Quit ("brb")
[14:17] * korken (n=kork@) has joined #freenet
[14:21] <toad_> hi korken-the-kork-plural
[14:22] <korken> oops
[14:22] * korken is now known as kork
[14:22] <toad_> any sign of sback? not so far ...
[14:23] * toad_ wonders if he should work on automated network testing in that case...
[14:23] * phrosty (i=phrosty@) has joined #freenet
[14:24] * pmpp\pas-la is now known as pmpp
[14:24] * Hikaru`_ (n=plop@) Quit (Read error: 104 (Connection reset by peer))
[14:28] * StefftheEmperor (n=Steffthe@) has joined #freenet
[14:34] <toad_> TOTALS:
[14:34] <toad_> NEVER_CONNECTED : 12
[14:34] <toad_> CONNECTED_SUCCESS : 10
[14:34] <toad_> oooh
[14:34] <toad_> that fix in 1155 seems to have really helped
[14:38] * sdiz (n=sdiz@) has joined #freenet
[14:38] * ChanServ sets mode +o sdiz
[14:38] * sdiz (n=sdiz@) Quit (Client Quit)
[14:38] * sdiz (n=sdiz@) has joined #freenet
[14:38] * ChanServ sets mode +o sdiz
[14:40] * Hikaru`_ (n=plop@) has joined #freenet
[14:44] * StefftheEmperor_ (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[14:59] * Johan^mlg (n=bllarf@) has joined #freenet
[15:02] * Luke771 (n=luke@) Quit ("Leaving")
[15:05] <FreenetLogBot> r21671 (1155) was built successfully on emu, mirrors are updating
[15:08] * sbc (n=ca@) Quit ("Ex-Chat")
[15:12] * sanity (n=ian@) has joined #freenet
[15:12] * ChanServ sets mode +o sanity
[15:18] * FrinkC (n=FrinkC@) Quit ("Bye")
[15:22] * Nande (n=KVIrc@) has joined #freenet
[15:45] * Hikaru`_ (n=plop@) Quit (Read error: 104 (Connection reset by peer))
[15:46] <kork> toad_: you there?
[16:06] * sanity_ (n=ian@) has joined #freenet
[16:06] * ChanServ sets mode +o sanity_
[16:06] * sanity (n=ian@) Quit (Read error: 104 (Connection reset by peer))
[16:12] * sich (n=sich@) has joined #freenet
[16:31] * sanity_ (n=ian@) Quit (Read error: 110 (Connection timed out))
[16:41] * sdiz (n=sdiz@) Quit ("Lost terminal")
[16:45] <FreenetLogBot> r21672 (1155) was built successfully on emu, mirrors are updating
[16:49] * sanity (n=ian@) has joined #freenet
[16:49] * ChanServ sets mode +o sanity
[16:52] * sanity_ (n=ian@) has joined #freenet
[16:52] * sanity (n=ian@) Quit (Read error: 104 (Connection reset by peer))
[16:52] * ChanServ sets mode +o sanity_
[16:55] <FreenetLogBot> r21673 (KeyExplorer) was built successfully on emu, mirrors are updating
[17:01] <toad_> kork: hi
[17:08] * StefftheEmperor_ (n=Steffthe@) has joined #freenet
[17:09] * StefftheEmperor (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[17:11] <FreenetLogBot> r21674 (1155) was built successfully on emu, mirrors are updating
[17:17] <FreenetLogBot> r21675 (1155) was built successfully on emu, mirrors are updating
[17:18] * sanity_ (n=ian@) Quit (Read error: 110 (Connection timed out))
[17:19] <toad_> okay... I can work on db4o or i can work on automated testing
[17:20] <toad_> automated testing is fairly urgent, but sback said he would say this morning whether he could do it
[17:20] <toad_> he hasn't been here, so i suppose i should assume he can't and do it myself
[17:20] * PhilKC (i=PhilKC@) has joined #Freenet
[17:34] * StefftheEmperor (n=Steffthe@) has joined #freenet
[17:35] * StefftheEmperor_ (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[17:37] * StefftheEmperor_ (n=Steffthe@) has joined #freenet
[17:46] * caytchen (n=caytchen@) Quit ("Verlassend")
[17:46] <kork> toad_: now I'm back... guess you're away again :D
[17:50] * kork (n=kork@) Quit ("brb")
[17:51] * mavest (n=mave@) has joined #freenet
[17:53] * Nyquist5 (n=Nyquist5@) has joined #freenet
[17:53] * StefftheEmperor (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[17:57] * mib_egq03d (i=5226f314@) has joined #freenet
[18:02] * sanity (n=ian@) has joined #freenet
[18:02] * ChanServ sets mode +o sanity
[18:02] * mib_egq03d (i=5226f314@) Quit ("http://www.mibbit.com ajax IRC Client")
[18:03] * kork (n=kork@) has joined #freenet
[18:07] * StefftheEmperor_ (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
[18:08] * mavest (n=mave@) Quit ()
[18:09] * toad_ confirms experimentally what we know anecdotally ... it takes a ridiculously long time to bootstrap a node at the moment
[18:11] <kork> toad_, question: is it okay to darknet-link my old and new seednode?
[18:11] <kork> (it's not up yet)
[18:11] <toad_> why wouldn't ti be?
[18:11] <toad_> Completed bootstrap (10 peers) in 432369ms.
[18:12] <kork> I don't know, that's why I asked
[18:12] <toad_> 7 minutes and 12 seconds? come on, that's insane!
[18:13] <kork> I'll bring up the second seed this weekend
[18:13] * toad_ tries with DISABLE_GROUP_STRIP = true
[18:14] <toad_> okay, it's nextgens fault :)
[18:15] <toad_> or the failure of the seednodes to upgrade ...
[18:15] <toad_> is the auto-updater working?
[18:15] <toad_> did it get 1155 successfully for people?
[18:15] <kork> dunno, I'm not using it :/
[18:16] <toad_> anyone else?
[18:17] <Nyquist5> my node says its downloading something now.
[18:17] <Nyquist5> but i'm currently on opennet
[18:18] <toad_> Completed bootstrap (10 peers) in 244527ms (4m4s)
[18:18] <toad_> hmmm
[18:18] <toad_> well it's an improvement ... but it still sucks
[18:19] * dbkr- (n=nnndavid@) has joined #freenet
[18:19] <Nyquist5> so, would anyone like to exchange references?
[18:25] <FreenetLogBot> r21676 (1155) was built successfully on emu, mirrors are updating
[18:27] <reflect> afraid not.. it's not done like that anymore
[18:27] <reflect> ref exchanges, as I understand it, are to be done between friends.. real life friends
[18:28] <toad_> they don't have to be IRL
[18:28] <toad_> but there does need to be some degree of acquaintance and ideally trust
[18:29] <toad_> random matchmaking causes bad performance and security similar to opennet
[18:29] <reflect> the routing works better, if some other friends of yours, also add the friend you added, I understand
[18:42] * dbkr (n=nndavidj@) Quit (Read error: 110 (Connection timed out))
[18:42] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[18:43] * dbkr- is now known as dbkr
[18:45] <kork> I wonder if I should put the datastore on its own lvm volume... dunno if it fragments the shit out of a filesystem as does for example mldonkey
[18:50] <toad_> the datastore won't but the temp files might
[18:52] <kork> I just created an extra volume for freenet
[18:52] <kork> won't hurt
[18:55] <toad_> bbiab
[19:31] * mib_ttvv19 (i=5226f314@) has joined #freenet
[19:32] <mib_ttvv19> hi, trying to use opennet, after downloading and putting the seednodes.fref in freenet folder I'm getting "There is no seednodes.fref file found, so the node will not be able to automatically bootstrap itself onto the opennet". What am I doing wrong?
[19:35] * |sich| (n=sich@) has joined #freenet
[19:41] <Tommy[D]> mib_ttvv19: Where did you download the file? Where exactly did you place it? How old is your install? Did you restart freenet?
[19:43] <mib_ttvv19> I've downloaded the file from http://downloads.freenetproject.org/alpha/opennet/. I put it in C:\program files\freenet folder. My install is fresh, I've just downloaded it 2 hours ago. I restarted the PC.
[19:50] * StefftheEmperor (n=Steffthe@) has joined #freenet
[19:52] * sanity (n=ian@) has joined #freenet
[19:52] * ChanServ sets mode +o sanity
[19:52] * sich (n=sich@) Quit (Read error: 110 (Connection timed out))
[19:58] * sanity_ (n=ian@) has joined #freenet
[19:58] * ChanServ sets mode +o sanity_
[19:58] * sanity (n=ian@) Quit (Read error: 104 (Connection reset by peer))
[20:04] <Tommy[D]> mib_ttvv19: How did you install freenet?
[20:05] * saces (n=saces@) Quit (Read error: 110 (Connection timed out))
[20:05] * saces (n=saces@) has joined #freenet
[20:05] * ChanServ sets mode +o saces
[20:07] * mib_ttvv19 (i=5226f314@) Quit ("http://www.mibbit.com ajax IRC Client")
[20:19] * sanity_ (n=ian@) Quit (Read error: 104 (Connection reset by peer))
[20:19] * sanity (n=ian@) has joined #freenet
[20:19] * ChanServ sets mode +o sanity
[20:21] * DaffyDuck_ (n=jan@) has joined #freenet
[20:50] * Johan^mlg (n=bllarf@) Quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
[20:51] * Hikaru` (n=Plop@) Quit (Read error: 113 (No route to host))
[21:03] * sanity (n=ian@) Quit (Read error: 110 (Connection timed out))
[21:10] * archypetro (n=hellequi@) has joined #freenet
[21:11] * caytchen (n=caytchen@) has joined #freenet
[21:13] <toad_> rehi
[21:15] <caytchen> rehi ;)
[21:16] * archangelpetro (n=hellequi@) Quit (Read error: 110 (Connection timed out))
[21:17] * |sich| (n=sich@) Quit ("KVIrc 3.2.6 Anomalies http://www.kvirc.net/")
[21:24] * n0ob (n=travis@) has joined #freenet
[21:33] * 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.")
[21:39] <FreenetLogBot> r21677 (1155) was built successfully on emu, mirrors are updating
[21:40] <toad_> anyone have any thoughts on how to insert data to an established node within a standalone test app which is part of the trunk? are we going to have to include jfcplib in order to build it?
[21:41] <reflect> bork-bork-bork?
[21:41] <toad_> hmmmm?
[21:41] * CeeDee (n=marc@) has joined #freenet
[21:41] <reflect> swedish chef joke, nevermind :)
[21:42] * toad_ supposes we only care about PutSuccessful and PutFailed ...
[21:42] <toad_> so it really wouldn't be so difficult
[21:42] <saces> toad_: use cmdline interface?
[21:43] <toad_> you mean TMCI? or some convenient FCP inserter tool?
[21:43] <saces> tmci
[21:43] <toad_> ah
[21:44] <toad_> i don't think that would be much easier to parse than FCP
[21:48] <kork> toad_: was there something special I had to do to become a seednode? I think you need my node's ref
[21:49] <toad_> you need to send me your opennet noderef
[21:49] <toad_> you need to be non-NATed (or port forwarded), and have either a static IP or a dyndns address
[21:50] <toad_> there's an option in the node config to enable/disable seednode support, it needs to be enabled
[21:50] <kork> mail is on its way
[21:50] <kork> aah I already have that one since I more or less copied the other node's config
[21:51] <kork> hm is there a way to make the node NOT use an address?
[21:51] <kork> physical.udp=87.98.244.250:10000;91.121.72.107:10000
[21:51] <kork> the 87 one is firewalled to hell
[21:51] <kork> :/
[21:53] <kork> got it
[21:53] <kork> node.opennet.bindTo
[21:53] * rstanile (n=washuu@) has joined #freenet
[21:54] <rstanile> hi all
[21:54] <CeeDee> stupid newbie question with regards 0.7 + Frost. The only main board messages me and the rest of the darknet have seen is a few in fuqid. The messages we send on our own board get through though (eventually). Is this because of the ddos I've read about? Any workarounds or alternative groups?
[21:56] <rstanile> I think my first install of freenet is successful, I'd like to add some friends. Does anyone care? I am on 4MBits symmetric link.
[21:57] <reflect> CeeDee, probably, yes.. many are using FMS now
[21:58] <CeeDee> gotcha.
[21:59] <kork> toad_: you've got another mail, please disregard the first one
[22:00] * Krhis (n=Krhis@) Quit ("Welcome to ASH, sorry you're here.")
[22:06] <Tommy[D]> rstanile: simply enable opennet, will get you the needed connections automaticly
[22:07] <rstanile> OK, I'm gone. My email is rstanile@tenbit.pl. Bye.
[22:07] <rstanile> Tommy[D]: opennet is working, already.
[22:08] <Tommy[D]> rstanile: so why do you ask for peers?
[22:08] <rstanile> I need to test port forwarding for darknet connections, and AFAIK i cannot do this wothout friends.
[22:09] <Tommy[D]> if port forwarding for opennet works and you did the same steps for darknet, why should it not work?
[22:09] <Tommy[D]> Any why should you care about darknet port forwarding, if you have no darknet peers?
[22:09] <rstanile> If it wuold be TCP instad of UDP, o would test it easily by telnet from another host.
[22:10] <kork> rstanile: you can try it with netcat if you have an external host somewhere
[22:10] <rstanile> kork: I have, but does netcat work with UDP?
[22:10] <kork> yep
[22:11] <toad_> rstanile: congrats
[22:11] <rstanile> ok, i'll RTFm. :) thx.
[22:11] <kork> nc -u host
[22:11] <kork> err
[22:11] <kork> nc -u host port
[22:11] <kork> enter something
[22:11] <kork> netcat ends after pressing enter: port closed
[22:11] <kork> netcat continues: port open
[22:12] <toad_> rstanile: do you have a static IP? if you get your ports sorted, and stick around long enough, you could be a seednode, that would be useful to us...
[22:13] <kork> toad_: did you get my mails?
[22:14] <rstanile> toad_: yes, it's static. But what do you mean by "sorted ports"? I set up port forwarding on 3 consecutive ports, starting with 6533.
[22:18] <rstanile> and I was asking about peers, because of the warning of insecure mode and "please forward your port UDP" for darknet.
[22:20] <toad_> kork: not looked at it yet
[22:20] <toad_> rstanile: i meant when you're sure the ports are forwarded
[22:35] <rstanile> ok, gotta go. I couldn't netcat on two of external servers, and i don't root access to install it, so I just wait some days and see how freenet is running.
[22:46] * rstanile (n=washuu@) Quit ("Tkans.")
[22:46] * Nande (n=KVIrc@) Quit (wolfe.freenode.net irc.freenode.net)
[22:46] * oierw (n=mathew@) Quit (wolfe.freenode.net irc.freenode.net)
[22:46] * nextgens (n=nextgens@) Quit (wolfe.freenode.net irc.freenode.net)
[22:46] * zeeb (n=zeebe@) Quit (wolfe.freenode.net irc.freenode.net)
[22:46] * archypetro (n=hellequi@) Quit (Remote closed the connection)
[22:48] * nextgens (n=nextgens@) has joined #freenet
[22:48] * Nande (n=KVIrc@) has joined #freenet
[22:48] * zeeb (n=zeebe@) has joined #freenet
[22:48] * oierw (n=mathew@) has joined #freenet
[22:48] * NullAcht15 (n=unknown@) Quit (Remote closed the connection)
[22:49] * archypetro (n=hellequi@) has joined #freenet
[22:54] * archypetro (n=hellequi@) Quit (Read error: 104 (Connection reset by peer))
[22:58] * archangelpetro (n=hellequi@) has joined #freenet
[23:02] * PhilKC (i=PhilKC@) Quit (Read error: 104 (Connection reset by peer))
[23:05] * NullAcht15 (n=unknown@) has joined #freenet
[23:06] * whooha (n=whooha@) Quit (Read error: 110 (Connection timed out))
[23:10] * whooha (n=whooha@) has joined #freenet
[23:15] * PhilKC (i=PhilKC@) has joined #Freenet
[23:15] * Zarggg (n=z@) has joined #freenet
[23:21] <FreenetLogBot> r21679 (1155) was built successfully on emu, mirrors are updating
[23:23] * ahuxley (n=ahuxley@) has joined #freenet
[23:26] <FreenetLogBot> r21681 (1155) was built successfully on emu, mirrors are updating
[23:27] * PhilKC_ (i=PhilKC@) has joined #Freenet
[23:28] * FrinkC (n=FrinkC@) has joined #freenet
[23:36] * PhilKC (i=PhilKC@) Quit (Connection reset by peer)
[23:44] * infinity0 (n=infinity@) has joined #freenet
[23:56] * StefftheEmperor_ (n=Steffthe@) has joined #freenet
[23:58] * StefftheEmperor (n=Steffthe@) Quit (Read error: 110 (Connection timed out))
Irc logs of #freenet : 2008 2007 2006 2005
These logs were automatically created by FreenetLogBot on chat.freenode.net using the Java IRC LogBot.