I2P dev meeting, October 5, 2004

Quick recap

  • Present:

cat-a-puss, cervantes, deer, dm, duck, jrandom, protok0l,

سجل IRC الكامل

14:05 < jrandom> 0) hi
14:05 < jrandom> 1) 0.4.1.1 status
14:05 < jrandom> 2) Pretty pictures
14:05 < jrandom> 3) 0.4.1.2 and 0.4.2
14:05 < jrandom> 4) Bundled eepserver
14:05 < jrandom> 5) ???
14:05 < jrandom> 0) hi
14:05  * jrandom waves
14:05 < jrandom> weekly status notes are available at http://dev.i2p.net/pipermail/i2p/2004-October/000461.html
14:06 < jrandom> (i cant believe its october)
14:06 < cervantes> it's december
14:06  * jrandom disconnects from cervantes.  excess clock skew
14:06 < deer> <baffled> could we have summer back now?
14:07 < cervantes> damn...lost your pr0n feed
14:07 < jrandom> sure. its a few thousand KM south of you baffled
14:07 < jrandom> ok, jumping into 1) 0.4.1.1 status
14:07 < deer> <baffled> will you let me know when I get there?
14:07 < cervantes> heh
14:07 < jrandom> click your heels three times...
14:08 < jrandom> ok, the 0.4.1 and 0.4.1.1 revs are out, and things are pretty much working again
14:08 < deer> <baffled> no, no, I don't want to go home it's cold there.
14:08 < jrandom> ;)
14:08 < jrandom> the autodetection of the external IP address seems to be working for the most part
14:09 < jrandom> (there have been a few quirks though, due to b0rked connections that don't hang up properly)
14:09 < jrandom> have people been using that, or had good/bad experiences with the autodetection?
14:10 < jrandom> guess not
14:10 < jrandom> ok, anyone have any comments/questions/concerns wrt 0.4.1.1?
14:11 < cervantes> no complaints here....
14:11 < dm> Haven't tried it yet, but it's on my agenda!
14:11 < jrandom> if not, swingin on to 2) pretty pictures
14:11 < jrandom> !thwap dm
14:12 < deer> <Jake> dunno about autodetection, but i tried using the 'guess' button or whatever on my natted windows box and it guessed the ip right...... if thats what wer're talking bout
14:12 < jrandom> ah ok, naw, the 'guess' button just tries to guess your IP by querying www.whatismyip.com
14:13 < jrandom> the autodetection is where you leave the IP address field blank and it figures it out by itself
14:13 < jrandom> most existing I2P users won't need it, since we're all used to either dyndns or static IPs anyway
14:13 < jrandom> it'll probably only matter for new users
14:14 < deer> <demonic_1> yea that worked a little slow for me
14:14 < deer> <demonic_1> but it did work
14:15 < jrandom> ok cool
14:15 < jrandom> anyway, i dont want to rehash what i posed in this weeks email wrt the stats gathered
14:16 < jrandom> instead, does anyone have any questions/comments/concerns about them?
14:17 < jrandom> i was pretty glad to see the 20h summary had only 500-something send failures out of 30,000-ish
14:17 < cervantes> how much load does the stats collecting generate?
14:17 < cervantes> I know the filesizes...but will it impact on performance having it ticking in the background
14:18 < jrandom> should be ~= 0.  there's no memory allocation in the stat gathering (as we use preallocated events) and everything is async
14:18 < cervantes> cool
14:18 -!- Sugadude [random@badfish.securityminded.net] has joined #i2p
14:18 -!- cat-a-puss [~tom@152.228.242.159] has joined #i2p
14:19 < jrandom> once 0.4.1.2 is outi'll probably nag some more people to gather various stats at times
14:19 < deer> <mule_iip> you're welcome
14:19 < cervantes> I'm happy to start collecting now... I'm already on 0.4.1.1-6
14:20 < jrandom> w3wt
14:21 < jrandom> ok, thats all i've got for the stats, unless anyone has anything to add?
14:21 < jrandom> if not, 3) 0.4.1.2 and 0.4.2
14:21 < deer> <baffled> You have my vote for streaming first.
14:22 < jrandom> cool
14:22 < jrandom> does anyone think we should keep the tunnel mods first?
14:22 < deer> <mule_iip> streaming first
14:23 < cervantes> doing the tunnel stuff now would likely cause more network distruption....it's probably good to have a breather ;-)
14:23 < jrandom> true
14:23 < deer> <mule_iip> all of those here today have been identified by the black hats anyhow :)
14:23 < jrandom> though i was thinking the other day about how we could do the tunnel mods without incompatabilities
14:23 < deer> <baffled> Comeon admit it, you just want to get your audio p0rn faster.
14:23 < duck> (me too on streaming first)
14:23 < jrandom> hehe
14:24 < cervantes> hehe
14:24 < cervantes> baffled: only if you source more of it ;-)
14:24 < dm> I think we should stick to the tunnel stuff first
14:24 < dm> get it out of the way...
14:24 < cat-a-puss> how is the new encryption stuff going to be different?
14:24  * jrandom kicks dm
14:25 < jrandom> cat-a-puss: right now, we have blanket tunnel encryption - messages passed within the same tunnel look the same at each hop
14:25 < deer> <baffled> I think I can get a bit more.
14:25 < cat-a-puss> oh!
14:26 < cervantes> http://www.i2p.net/todo#tunnelId
14:26 < jrandom> it isn't so bad since an alice-->bob message goes through two tunnels with different encryption, but it does b0rk us for colluding attackers
14:27 < jrandom> the per-hop tunnelId stuff is also necessary to keep harvesting from messing with predecessors (/etc)
14:27 < dm> Yeah, we should definitely fix that first.
14:27 < deer> <mule_iip> i vote for dm to do it
14:28 < deer> <fidd> did i miss the meeting? ;)
14:28 < jrandom> i was just about to suggest that mule :)
14:28 < cervantes> I vote for dm not to have anything to do with it
14:28 < jrandom> heh
14:28 < jrandom> nope fidd, we're on item 3 of the agenda
14:29 < jrandom> ok, if there are no objections to dm's suggestion (other than his own), i think we'll go ahead and move the streaming lib updates to 0.4.2 
14:29 < dm> sweet
14:30 < jrandom> ok, moving on to 4) Bundled eepserver
14:30 < jrandom> if you haven't noticed, there's a bundled eepserver.
14:30 < cervantes> "just put the war files in the webapps directory and you're ready to go"
14:30 < jrandom> heh
14:30 < jrandom> for sufficiently well coded .war files :)
14:31 < cervantes> ooh does such a think exist?
14:31 < cervantes> *thing
14:31 < jrandom> but from a practical perspective, "just edit ./eepsite/docroot/index.html"
14:31 < deer> <baffled> One question I have is are you wishing people would use the eepserver or use a standard httpd server?
14:31 < cat-a-puss> do the ones generated by kde work?
14:31 < jrandom> cervantes: phttprelay.war, i2ptunnel.war, routerconsole.war :)
14:31 < dm> ah yes.. war. One of those J2EE things that requires 20 years experience at manually editing xml files.
14:31 < cervantes> touche
14:32 < jrandom> baffled: i really don't care.  if people have a webserver installed that can accept requests from kooky Host: lines, great
14:32 < jrandom> the eepserver is just for convenience
14:32 < jrandom> cat-a-puss: hmm, kde .war files?
14:32 < protok0l> monoculture... monoculture...
14:33 < deer> <duck> when playing with wars, I miss the feature to only restart jetty; which is unfortunately needed for a lot of deployment stuff
14:33 < cat-a-puss> yeah, you need kdeaddons installed, just go to a webpage and then click archive and it makes a .war file
14:34 < jrandom> duck: ah, thats true.  simply pull out the lines starting the eepserver from clients.config and put them into a shell script
14:34 < jrandom> (with the same classpath as the router)
14:34 < dm> can we integrate i2p into jboss and bundle that before 1.0?
14:34 < jrandom> ooh, cool cat-a-puss 
14:35 < cervantes> I take it the missing webdefault.xml has been fixed in cvs?
14:35 < deer> <detonate> actually, jetty.xml has
14:35 < jrandom> find us a compelling .ear dm :)
14:35 < jrandom> cervantes: what detonate said.  (i messed up the jetty.xml)
14:36 < cervantes> yup... think I mentioned somewhere about removing the reference in the jetty.xml so it uses it the one inside the jetty archive
14:36 < jrandom> wr0d
14:37 < cervantes> just wanted to check that's been fixed in cvs ;-)
14:37 < jrandom> si sr
14:37 < cervantes> cool
14:37 < jrandom> (though the 0.4.1.2 release update will not overwrite people's eepsite)
14:37 < jrandom> ((0.4.1.2+ clean installs will of course include it though))
14:38 < cervantes> oh and did we discover the cause of DrWoo's missing eepsite keys?
14:38 < jrandom> actually, on that note, i just want to mention that everyone should upgrade whenever there is a new release, as if you don't, you might not have an upgrade procedure
14:38 < jrandom> no cervantes, nor a reproducable bug :/
14:39 < cervantes> ah good we can blame user error ;-)
14:39 < deer> <DrWoo> cervantes: almost certainly something klutzy I did
14:39 < cervantes> :o)
14:39  * jrandom blames the gremlins
14:40 < deer> <Jake> http://en.wikipedia.org/wiki/User:Kmweber/List_of_Everyone_Who_Has_Ever_Lived
14:40 < jrandom> ok, moving on to 5) ???
14:40 < jrandom> heh
14:40 < jrandom> well, yes, that certainly qualifies as "other"
14:40 < jrandom> anyone have anything they want to bring up?
14:41 < dm> I'd like to put forward, at this point, that I am pleased with the new outlook the I2P community is showing towards my suggestions.
14:41 < dm> Kind Regards
14:41 < cat-a-puss> oh oh pick me! I have the base code for a distrubuted search.
14:41 < deer> <demonic_1> yea why do i2p after running 30 + hours go up to 100% cpu
14:41 < dm> dm
14:41 < deer> <Jake> yes, i want to bring up the issue of encryption inheritence based on 4th order gamal fractal equations and how it would apply to i2p
14:41 < deer> <demonic_1> and most of it system?
14:41 < jrandom> ooh kickass cat-a-puss!
14:41 < cat-a-puss> I anounced it here the other day, nobody noticed
14:41 < deer> <baffled> only tangentially jake.
14:42 < cat-a-puss> anyway, could use come cvs space
14:42 < deer> <DrWoo> cat-a-puss: do you have an eepsite for that?
14:42 < jrandom> demonic_1: hmm, there have been some critical bugs in the last release or two.  are you on 0.4.1.1?
14:42 < cat-a-puss> and I can start testing in about 2 weeks
14:42 < cat-a-puss> DrWoo: nope
14:42 < deer> <Jake> baffled, HaH !
14:43 < deer> <demonic_1> 0.4.1.1-3
14:43 < jrandom> cat-a-puss: r0x0r, not a problem.  bounce me an email with the name of the module you'd like it called & your pgp key and we'll get something sorted
14:44 < cat-a-puss> jrandom: alright
14:44 < jrandom> cat-a-puss: what sort of searching does it do?
14:44 < jrandom> demonic_1: did it consume that much CPU prior to 0.4.1?
14:44 < cervantes> (proxies to MSN)
14:44 < deer> <mule_iip> demonic_1: and you get 1 meg of log every minute? sounds familiar.
14:45 < deer> <demonic_1> no
14:45 < jrandom> heh mule, yeah the bug you found was a nasty fast-busy
14:45 < cat-a-puss> jrandom: it's basic keywork search, you need to specify the words to index under, and it will store the URL
14:45 < jrandom> demonic is more likely being hit by one of the NPEs in the tcp.ConnectionBuilder
14:46 < deer> <baffled> Well, it's dindin time so I'll go hunt up some more slut sounds in preparation for the streaming updates and chat with you all anon.
14:46 < cat-a-puss> jrandom: It should eventualy scale well, and all that jazz, but right now, all the servers need to be connected and nobody can join or leave, and there is no way to insert content yet, but all that will get fixed
14:46 < jrandom> ah cool, does it work with a distributed db, or is it more of a search-against-spidered?
14:47 < jrandom> ok cool
14:47 < cervantes> later baffled
14:47 < jrandom> lol, ttyl baffled
14:47 < cervantes> baffled: how do we know they're slut sounds, and not you on the end of your microphone?
14:47 < protok0l> ALL RIGHT!
14:47 < protok0l> i2p works again
14:47 < jrandom> w3wt
14:48 < jrandom> what was wrong?
14:49 < jrandom> ok, anyone else have anything they want to bring up for the meeting?
14:49 < deer> <Jake> can announce i2p to slashdot after the new streaming protocol is implemented ?
14:49 < dm> preferably before
14:49 < dm> but after will do
14:49 < jrandom> !thwap^2
14:50 < protok0l> POSTMAN!
14:50 < jrandom> ok, if there's nothing else..
14:50  * jrandom winds up
14:51 < deer> * Jake kisses jrandom 
14:51  * jrandom *baf*s the meeting closed