I2P dev meeting, November 2, 2004

Quick recap

  • Present:

BS314159, cat-a-puss, deer, dm, duck, janonymous1, jrandom, protok0l,

Registo Completo do IRC

13:37 < jrandom> 0) hi
13:37 < jrandom> 1) Net status
13:37 < jrandom> 2) Core updates
13:37 < jrandom> 3) Streaming lib
13:37 < jrandom> 4) mail.i2p progress
13:38 < jrandom> 5) BT progress
13:38 < jrandom> 6) ???
13:38 < jrandom> 0) hi
13:38 < jrandom> sorry for the delay, weekly status notes posted @ http://dev.i2p.net/pipermail/i2p/2004-November/000477.html
13:38 < dm> meeting in 24 or 84?
13:38 < jrandom> 0
13:38 < dm> oh.. -36?
13:39 < jrandom> yup, 9p GMT
13:39 < jrandom> but i forgot that, so we're starting... now ;)
13:39 < jrandom> 1) net status
13:39 < dm> good timing
13:39 < jrandom> well, no real change in the net status from my end - does anyone have anything they'd like to bring up about it?
13:41 < jrandom> if not, might as well move on to 2) core updates
13:41 < jrandom> i dont really have anything to add beyond whats in the email, so i'll give people a min to digest
13:42 < deer> <postman> arg
13:42 < jrandom> there've been 8 patches since the release, with another one or two pending.  we'll probably just tag those all up into a 0.4.1.4, as the streaming lib itself isn't ready
13:43 < deer> <jrandom> wb, its looking a bit bumpy over here
13:43 < deer> <postman> np, i am back :)
13:43 < protok0l> any word on aum's disappearance? i want stasher!
13:44  * dm likes knowing that stuff is being done under the hood to optimize I2P
13:44 < jrandom> as gott quoted, diy, do or die
13:45 < jrandom> yeah, the memory churn was getting to be a substantial portion of the CPU time
13:45 < jrandom> so it was finally worth the effort to optimize
13:45 < deer> <baffled> Sorry, have to catch a bus I'll read the logs later night.
13:45 < deer> <peer> hi just a bug report
13:45 < jrandom> (as its cut down streaming lib test time by a factor of 5)
13:45 < jrandom> cool baffled, ttyl
13:46 < deer> <peer> when your net connection goes down, i2p dies
13:46 < dm> These are the kind of things that creep up on you, good to get them out of the way while the project is still lean.
13:46 < deer> * postman noticed this too a few days ago
13:46 < deer> <postman> one of my servers lost its link
13:46 < deer> <postman> for a few mins - after that i2p was good for a complete restart
13:46 < jrandom> dies, as in, the JVM stops, or the router stops talking to peers?
13:47 < jrandom> (it obviously stops talking to peers, i mean, after the net is back up, does it recover?)
13:47 < deer> <postman> jrandom: in my case jvm was still running - but no connection lead to success for about 15 minutes
13:47 < deer> <postman> jrandom: after that i restarted
13:47 < jrandom> hmm, ok, cool
13:48 < jrandom> thanks peer, postman.  i'll do some debugging down there
13:48 < jrandom> what OSes, btw?
13:48 < deer> <postman> jrandom: np - wanted to write you a mail  but forgeot
13:49 < deer> <postman> jrandom: Linux 2.4.recent - glibc2.3.recent jvm 1.4.05
13:49  * jrandom suspects that this week will be the week of "break shit and make i2p handle it better"
13:49 < jrandom> word
13:50 < deer> <peer> jrandom: in my case the jvm went completely
13:50 < jrandom> did it say OutOfMemory or have any CRIT messages?  or did it create a hs_* file inyour i2p install dir?
13:52 < jrandom> perhaps we could dig through the details later, after the meeting
13:52 < jrandom> does anyone have anything else on 2) core updates?
13:52 < jrandom> if not, on to 3) streaming lib
13:53 < dm> yeah
13:53 < dm> this increased latency
13:53 < dm> do you have an estimated % increase per hop?
13:53 < dm> we talking a couple % points or 30-40%?
13:53 < jrandom> none, its just some situations it didn't send through an outbound tunnel
13:54 < dm> so negligeable... 'kay
13:54 < dm> (on average)
13:54 < dm> 3)
13:54 < jrandom> 0% per hop, but its as if the peer you talk to has tunnels 1 hop longer than before (on average)
13:55 < jrandom> not many real visible updates for the streaming lib so far
13:55 < jrandom> things work pretty well, and i've been doing a bunch of benchmarks to track the progress during the recent memory updating
13:55 < dm> oh throughput numbers!!!
13:57 < dm> ping
13:57 < deer> <Natalia> .
13:57 < jrandom> well, it varied on the message size and per-hop latency injected, but preliminary throughput has been 2-5x faster
13:57 < jrandom> it has been CPU bound though
13:57 < dm> hmmm, not bad.
13:58 < dm> cpu at which end?
13:58 < jrandom> the big benefit is in the reduction of data retransmission and the virtual elimination of failure ;)
13:59 < jrandom> dm: these tests were done with the sim, injecting random delays per hop
13:59 < jrandom> (e.g. 400ms each time, or 1000ms, or 2000ms)
13:59 < dm> Is there some kind of priority scheme so that forwarding of messages of tunnels won't be too affected by people trying to download at 30k/s and maxing out their CPU?
13:59 < jrandom> (well, the *big* benefit is the sliding window and reordering, but reduction of retransmit is good)
14:00 < jrandom> not sure i understand
14:00 < dm> Like if I'm downloading porn, will I inject a 3s lag to anyone who's going through me in their tunnels.
14:00 < jrandom> (and the transfer rates were much higher than 30KBps, but again, this was local-only with random injected delays)
14:01 < dm> I'm just wondering what happens in general if someone is maxing out their CPU, as far as their contribution to the network is concerned.
14:01 < dm> I guess it's not specific to abusing the streaming lib.
14:02 < jrandom> you're not going to be maxing your CPU doing streaming, the CPU load was something i run into when using the local sim running a ton of routers on a single box
14:02 < dm> ah alright, I thought the cpu was maxed with one router trying to encrypt all the bits going down the pipe.
14:02 < jrandom> nah, encryption is ReallyReallyFast
14:03 < dm> coo'
14:03 < jrandom> ok, anyone else have any questions wrt the streaming lib progress?
14:03 < jrandom> if not, 4) mail.i2p progress
14:04 < deer> <jrandom> postman, you 'round?
14:04 < deer> <postman> yo :)
14:04 < deer> <postman> ok
14:04 < deer> * postman waves
14:05 < deer> <postman> well, gentlemen. Some of you may have noticed that we have finally implemented in/out services
14:05 < jrandom> [w00t!]
14:05 < deer> <postman> please reas www.postman.i2p/inout.html
14:05 < deer> <postman> please test the system out
14:06 < deer> <postman> baffled will deliver the 2nd official mx
14:06 < jrandom> word
14:06 < deer> <postman> right now i am working on IMAP implementation
14:07 < deer> <postman> this means a switch to maildir format soon
14:07 < jrandom> we'll need to recheck various clients for that though, right?
14:07 < deer> <postman> right now i am evaluating/testing
14:07 < jrandom> cool
14:07 < deer> <Natalia> why IMAP and not pop3 ?
14:07 < deer> <postman> yeah, and the serverside as well
14:08 < deer> <postman> Natalia: we have pop3 already
14:08 < deer> <postman> pop3 can be used of course 
14:08 < deer> <postman> IMAP4 will make us more flexible for webmail systems (hopefully)
14:10 < deer> <postman> this is still open issue
14:10 < deer> <Natalia> okay.
14:10 < deer> <Natalia> you sounded like you were going to switch from pop3 to IMAP
14:11 < deer> <postman> no, of course not
14:11 < deer> <postman> jrandom: are there any news concerning locally run webmail?
14:12 < jrandom> not to my knowledge.  i havent had time to look into it at all
14:12 < deer> * postman neither
14:12 < jrandom> there were those discussions of atmail, but they're closed source
14:12 < deer> <postman> mmh, yes
14:13 < deer> <postman> but something jspish ?
14:13 < jrandom> 'twould be a really great way for a volunteer to jump in and do some legword :)
14:13 < deer> <Natalia> well, I've added this description to gott.i2p/sites.html
14:13 < deer> * postman is completely unable to do research on that matter
14:13 < deer> <Natalia> for www.postman.i2p
14:13 < deer> <Natalia> postman runs i2p's first mail-service, providing free and anonymous pop3 and SMTP 
14:13 < deer> <Natalia> accounts over i2p. Recently implemented is the ability to send and receive e-mails to and 
14:13 < deer> <Natalia> from outside of the i2p network, marking the services of www.postman.i2p as a nifty 
14:13 < deer> <Natalia> destination for any concerned e-mailer and soon a must-have, as mail.i2p e-mail accounts 
14:13 < deer> <Natalia> become the norm for eepsite-authors.
14:14 < deer> <Natalia> sound good ?
14:14 < deer> <postman> thanks Natalia :)
14:14 < deer> <postman> jrandom: i think it's not a urgent issue
14:14 < deer> * Natalia curtsies :)
14:15 < deer> <postman> jrandom: maybe we pick up the webmail issue later again :)
14:15 < jrandom> agreed postman
14:15 < deer> <postman> that's all from my side , thanks :)
14:15 < jrandom> word, thanks postman
14:15 < deer> * postman curtsie too and sits down again
14:15 < jrandom> ok, anything else on that, or shall we move on to 5) BT progress?
14:16 < deer> <jrandom> dinoman: you 'round?
14:16 < dm> Yeah, I'm still waiting for BT to reactivate my ADSL
14:16 < jrandom> !thwap
14:17 < deer> <duck> dino has done some good work
14:17 < deer> <duck> with Ragnarok to fix some ends
14:17 < deer> <duck> so far it looks like the current problems are:
14:17 < deer> <duck> - SAM unreliability
14:17 < deer> <duck> - Python SAM library issues
14:17 < deer> <duck> - Incorrect usage of the Python SAM lib
14:18 < deer> <duck> - Correct handleing of destination instead of host/ip/port
14:18 < deer> <duck> once those are fixed it should work
14:18 < jrandom> cool
14:19 < deer> <duck> I think it is needed to take a little step back though
14:19 < deer> <duck> and agree on how to modify the protocol to properly handle destinations
14:19 < deer> <duck> it will be incompatible anyway, so better break it good
14:19 < jrandom> i concur
14:20 < jrandom> perhaps someone can mock up an overall plan of what needs to be done to various apps/components to get it working
14:20 < deer> <duck> each peer has an unique peer_id of 20 bytes
14:20 < deer> <duck> it is normally derived from the host/ip
14:21 < deer> <duck> I think that using the full destination is a bit much
14:21 < deer> <duck> what globally unique thing should we use?
14:21 < jrandom> SHA1(destination)[0:19]
14:21 < jrandom> perhaps?
14:21 < deer> <Ragnarok> first twenty bytes of the dest? :)
14:22 < deer> <duck> a sha1 hash is 20 bytes
14:22 < jrandom> first 20 bytes of the dest should be pretty random too, enough to deal with random clashes, but not to deal with hostile colisions
14:22 < jrandom> even better 
14:22 < deer> <dinoman> if you lose the key how do peers find one another
14:22 < jrandom> a peer *is* a key
14:23 < jrandom> oh
14:23  * jrandom misinterpreted
14:23 < jrandom> the tracker must give peers the full destination, not the SHA1(destination)
14:24 < jrandom> is that the same peer_id in question?
14:24 < deer> <dinoman> i have fixed the php tracker to send the full key as the ip
14:24 < deer> <duck> actually the client generates the peer_id
14:24 < deer> <duck> (what do you mean with 'key'?)
14:25 < deer> <dinoman> destination
14:25 < dm> Sounds like a who's on first skit.
14:25 < dm> Use full sentences people!
14:26 < deer> <dinoman> ok fine :/ the tracker sends the Full destination as the ip
14:27 < jrandom> heh dont mind dm.  sounds great
14:27 < deer> <dinoman> peer id is just for the trackers
14:27 < deer> <duck> maybe we could use #i2p-bt
14:28 < jrandom> what i think would be useful though is if you (or someone else) could perhaps draft up a list of modifications that'll need to be made
14:28 < deer> <duck> so no religious wars start each time the name of the snake is dropped
14:29 < deer> <dinoman> works for me
14:29 < deer> <dinoman> i don't war if it works it works
14:29 < jrandom> (e.g. "tracker sends e full destination as the IP", "client interprets the IP as the full destination", "torrent contains the tracker's destination in the field 'trackerDest'", etc)
14:29 < deer> <duck> definitly
14:30 < deer> <dinoman> jrandom you got it
14:31 < deer> <dinoman> this is the sample output of the tracker 8:intervali300e12:min intervali30e5:peersld2:ip50:klkjlkfsdjfkljkfdhjkddfsjkldsfjlkjfdlkjsfdl;kj;sdf7:peer
14:31 < dm> copy/pastes jrandom's sentence into notepad and saves as "draft.txt"
14:31 < cat-a-puss> will bt over i2p be intercompatible with other clients that are not over i2p?
14:31 < jrandom> cool dinoman
14:31 < deer> <dinoman> at ip50 you will see a junk key
14:32 < jrandom> cat-a-puss: yes
14:32 < deer> <dinoman> yes
14:32 < cat-a-puss> then we should talk
14:32 < jrandom> welcome to the weekly meeting!  :)
14:32 < deer> <dinoman> it will need to be something like .i2ptorrent to make it live togeter
14:32 < deer> <dinoman> for filenames and links and what not
14:33 < jrandom> are you working on something similar cat-a-puss, or have some ideas for improvements?
14:33 < cat-a-puss> working on something similar
14:33 < cat-a-puss> in java
14:33 < jrandom> cool
14:34 < jrandom> is it necessarily java specific, or can some peers be in other langs?
14:34 < cat-a-puss> good question, I don't know how to work that sort of thing in java, I'll have to look into it
14:35 < deer> <duck> right
14:35 < deer> <duck> lets use ugha.i2p to write up some specs
14:35 < deer> <duck> .
14:35 < jrandom> or perhaps we need a "swarming data transfer" section in the forum so we can all discuss this stuff at our own pace?
14:35 < jrandom> or ugha.i2p, of course
14:36 < jrandom> (while we work through some bugs in the sam impl and libs :)
14:36 < deer> <duck> makes it all a challenge
14:37 < deer> <dinoman> hehe ok
14:38 < deer> <duck> ...
14:38 < deer> <duck> mo bt?
14:38 < deer> * dinoman gets back to work on Savane
14:39 < jrandom> http://ugha.i2p/SwarmingTransfer / http://ugha.ath.cx/SwarmingTransfer
14:39 < jrandom> word
14:39 < jrandom> ok, anything else on 5) BT progress?
14:39 < jrandom> or shall we hit 6) ???
14:39 < jrandom> and ask dinoman how the Savane progress is coming?  :)
14:40 < deer> * jrandom cracks whip
14:40 < deer> <dinoman> mail i am stuck on using the i2p mail system
14:40 < deer> <dinoman> i think i should just take the mail out
14:40 < jrandom> is there any way to tell it to use the SMTP server at a different port?
14:40 < jrandom> or is the problem authenticated SMTP?
14:41 < deer> <dinoman> auth
14:41 < protok0l> Uptime: 5d
14:41 < protok0l> ii own
14:41 < deer> <dinoman> it is not in the class Savane uses
14:42 < deer> <dinoman> i can put it in 
14:42 < protok0l> i'm "Ident: pxEI" can someone tell me my rating
14:42 < jrandom> ok, i bet we can just get postman to set you up with a custom SMTP destination that doesnt require authentication
14:42 < dm> I give you a 6/10
14:42 < dm> You could work on your ass a bit
14:42 < janonymous1> Whats savana
14:43 < jrandom> janonymous1: its like sourceforge
14:43 < deer> <dinoman> because i am looking at the I2P Public Domain Software Homepage in my browser now
14:43 < jrandom> w00t
14:45 < deer> <dinoman> that would be cool be what is being done on the server i don't want someone hacking me and the getting the info about the mail server
14:45 < deer> <dinoman> that is what bugs me
14:45 < jrandom> well, they wouldn't get any info on the mail server, they'd just be able to (at worst) spoof @mail.i2p
14:45 < janonymous1> Cool
14:46 < jrandom> but yeah, it'd be great to have authenticated SMTP support to prevent that
14:46 < jrandom> i dont know how much work that'd be though
14:46 < protok0l> well im glad i left my mailserver idea to postman
14:46 < protok0l> it seem more difficult than i imagined
14:47 < deer> <Ch0Hag> I wouldn't mind helping with that
14:47 < dm> protocol
14:47 < deer> <Ch0Hag> Got to do something. :-)
14:47 < deer> <dinoman> i will do auth :( it will take a little time but i will do it
14:47 < deer> <protokol> yes dm
14:48 < jrandom> see, you've got a volunteer already dinoman!  :)
14:48 < deer> <protokol> maybe i could host a nessus server
14:48 < deer> <protokol> and tunnel it through TOR on my side
14:49 < deer> <Ch0Hag> Plus I need a  good excuse to work on the rest of  my network.
14:49 < deer> <protokol> and i shall dedicate myself to learning python
14:49 < janonymous1> 'the i2p software foundation'.  I can see it now
14:49 < deer> <protokol> and how to correctly type
14:49 < dm> I shall dedicate myself to the pursuit of more money for myself and for those directly related to myself, who may be inclined to give me money in the near future.
14:50 < jrandom> ok, anyone else have anything to bring up for 6) ??? 
14:50 < dm> 7) $$$
14:51 < duck> Roger Dingledine (arma @ freenode) published a draft for a chapter of an upcoming O'Reilly book
14:51 < duck> http://freehaven.net/doc/wupss04/usability.pdf
14:51 < jrandom> ah, yeah, its pretty good
14:51 < duck> it is about anonymity and usability
14:51 < dm> chapter on usability?
14:51 < deer> <protokol> i can run the i2p software foundation
14:51 < deer> <protokol> lol
14:51 < duck> some interesting parts about negative imago
14:52 < deer> <protokol> give me the keys the the treasury
14:52 < duck> having good default
14:52 < deer> <protokol> NOW!
14:52 < duck> etc
14:52 < duck> .
14:52 < jrandom> and the importance of usability, even over security at times
14:52 < dm> protok0l: you're the user advocate aren't you? You should read that document.
14:52 < jrandom> 'k, anything else for the meeting?
14:52 < deer> <protokol> wow, im seeing 83 peers
14:52 < duck> now we know why there are so few known hidden sides on tor
14:53 < deer> <protokol> dm: i shall
14:53 < duck> arma is affraid for negative imago
14:53 < duck> .
14:53 < dm> "imago" ?
14:53 < duck> image
14:53 < deer> <duck> (psychoanalysis) an idealized image of someone
14:53 < dm> No mention of I2P in there :(
14:53 < duck> jrandom: aint we?
14:54 < jrandom> hm?
14:54 < dm> he means aren't we. He's dutch.
14:54 < duck> if some specific group now moves to i2p,
14:54 < duck> they could keep away much needed other users
14:55 < jrandom> oh, thats in there?  i didnt see that
14:55 < duck> no, I am saying that
14:55 < duck> but it is in there too, more or less
14:55 < duck> ofcourse andy anarchist doesnt give a fuck
14:56 < jrandom> well, i do think there is room for both i2p and tor
14:56 < duck> yes
14:56 < duck> but what about early negative image on I2P
14:56 < deer> <Natalia> this is the reason I am forced to be a somewhat mundane female on this IRC channel
14:56 < protok0l> haha, when i get the word every major anarchist listserv and forum will hear about i2p within a day or 2
14:56 < jrandom> oh, i dont give a flying fuck about that duck ;)
14:56 < deer> <Natalia> jrandom doesn't approve of got
14:56 < deer> <Natalia> *gott
14:57 < duck> jrandom: yeah, but well
14:57  * duck counts the amount of anarchy friendly regions on the globe
14:57 < deer> <Natalia> so I have to be Natalia, the loved female of the channel
14:57 < deer> <Natalia> ( lame )
14:57 < duck> somalia?
14:57 < duck> I bet they have flying fucks there
14:57 < protok0l> Chiapas, mexica
14:57 < duck> but not friendly ones
14:57 < protok0l> mexiico
14:58 < deer> <Ragnarok> bah, you want to be feminized
14:58 < jrandom> duck: when it comes time to be more public, i'm certain we can put on a reasonable joe sixpack friendly face
14:58 < duck> k
14:58 < jrandom> will people do "bad" things with i2p?  yeah
14:58 < dm> I think we should target joe beergut
14:58 < protok0l> good luck, i know gott is planning something
14:58 < protok0l> gott will destroy us
14:58 < duck> ok
14:58 < duck> .
14:58 < jrandom> the only way any worthwhile anonymity or security system can survive is to be content neutral
14:59 < deer> <Ragnarok> anonymous communication systems can only protect communication.  They don't interfere with good old police work if someone actually *does* something.
14:59 < duck> just saying that some links placed on http://127.0.0.1:7657/index.jsp could be bad
14:59 < dm> I2P is about technology.
14:59 < deer> <Natalia> yes
14:59 < jrandom> true enough duck
15:00 < duck> and yes, the sitelist.html will turn into a TFE discussion thing all over
15:00 < jrandom> well, mmhmm
15:00 < deer> <Natalia> content neutrality is something I write about in the latest eeplog entry
15:00 < deer> <Natalia> http://gott.i2p/eeplog.html
15:01 < jrandom> this is, however, the power of interactive eepsites, like wikis
15:01 < jrandom> (e.g. having people register their site with a sitelist.py or whatever)
15:01 < deer> <Natalia> jrandom: do you support not support the idea of eepsite crawlers linking to illegal material, being linked from the frontpage ?
15:01 < deer> <Natalia> +or
15:01 < deer> <Natalia> if you were going to link to the sitelist
15:02 < duck> from a moral point I dont give a flying fuck either
15:02 < deer> <Natalia> jrandom: none of these are registered
15:02 < duck> but from an usability point I might
15:02 < deer> <Natalia> the script checks host.txt
15:02 < deer> <Natalia> *hosts.txt
15:02 < jrandom> from a nontechnical perspective, i support whatever the user community requires
15:02 < deer> <Natalia> so everyone gets added to the list if they have a domain
15:03 < deer> <Natalia> ugh, bras are so uncomfortable.
15:03 < protok0l> yup, creepy
15:03 < deer> <cervantes> have you _seen_ the user community?
15:03 < cat-a-puss> The simplist solution would be to simply link to search pages, Everyone knows how to use them, they provide fast access, and nobody sees for anything they did not ask for.
15:04 < deer> <cervantes> :)
15:04 < protok0l> gott is a serial killer, i know it. he will be the first to offer live murders via webcam on i2p
15:04 < deer> <Natalia> the user community consists of rather strange people.
15:04 < jrandom> good point cat-a-puss, we could just link to files.i2p 
15:04 < deer> <Natalia> at the moment, I am forced to be a woman because the lead developer disapproves of the immoral behaviour of my other.
15:04 < duck> cat-a-puss++
15:04 < deer> <Natalia> we are united through common adventure.
15:06 < BS314159> I'm not convinced this is a good idea, but the I2P license is certainly broad enough for people to spin off their own versions, differing only in the local link pages
15:06 < deer> <Natalia> well.
15:06 < deer> <cervantes> lets hope DrWoo can keep his indices free of corruption
15:06 < jrandom> certainly BS314159 
15:06 < BS314159> not versions. distributions.
15:06 < deer> <Natalia> files.i2p should be one link
15:06 < jrandom> BS314159: people can even edit their own local link page
15:06 < deer> <Natalia> and then there should be a yahoo-style internet directory link
15:06 < protok0l> most people will be wise enuf to use the official version
15:06 < jrandom> (in docs/readme.html)
15:07 < deer> <Natalia> search engines and internet directories serve different roles
15:07 < deer> <Natalia> this is why the directory is there in the first place
15:07 < deer> <Natalia> it has been requested as independent of a search engine
15:07 < BS314159> so if you want e.g. to target an anti-pornography demographic, find an anti-pornography maintainer who maintains a filtered default start page set
15:07 < protok0l> unless they are willing to search for backdoors in third-party versions
15:07 < deer> <Natalia> by people
15:07 < deer> <Natalia> so I think the search-engine is good
15:07 < jrandom> right BS314159 
15:07 < deer> <Natalia> but should not be the limit
15:07 < deer> <Natalia> search engine, internet directory, wiki, help page
15:07 < deer> <Natalia> perhaps.
15:08 < jrandom> we already link to fproxy.i2p, and we know what scary evil content they have on that site ;)
15:08 < BS314159> I'm not sure I'm on topic, but that seems possible. Is there an open-source content filter that any search-engine maintainers would be willing to implement support for?
15:08 < BS314159> I have a feeling I'm not on topic
15:08 < protok0l> is the meeting still on?
15:08 < jrandom> yes protok0l 
15:08 < BS314159> sorry. (silences self)
15:08 < deer> <Natalia> jrandom: perhaps you shouldn't link to fproxy.i2p
15:08 < deer> <Natalia> it is almost always down
15:08 < jrandom> BS314159: i think a cntent filter in the search engine is excessive
15:08 < deer> <Natalia> it is down right now, it seems
15:09 < protok0l> it is
15:09 < deer> <Natalia> according to the recent run of the site-checking script
15:09 < jrandom> 'k
15:09 < jrandom> well, this has been a good discussion, lots of good ideas
15:09 < BS314159> not _the_ search engine. _someone_'s search engine
15:10 < deer> * Natalia smiles.
15:10 < deer> <cervantes> BS3: aol.i2p ;-)
15:10 < jrandom> ok, is there anything else for the meeting?
15:10 < deer> <cervantes> whoa...still in the meeting...
15:11 < deer> <cervantes> thought I'd missed that by an hour
15:11 < jrandom> nope, i was late
15:11 < jrandom> ok, if not..
15:11  * jrandom winds up
15:11  * jrandom *baf*s the meeting closed