I2P dev meeting, November 1, 2005

Quick recap

  • Present:

brutus, cervantes, Complication, defnax, dust, jme___, jrandom, polecat, redzara,

IRC 完整日志

15:04 < jrandom> 0) hi
15:04 < jrandom> 1) and net status
15:04 < jrandom> 2) boostraps, predecessors, global passive adversaries, and CBR
15:05 < jrandom> 3) i2phex
15:05 < jrandom> 4) voi2p app
15:05 < jrandom> 5) syndie and sucker
15:05 < jrandom> 6) ???
15:05 < jrandom> 0) hi
15:05  * jrandom waves
15:05 < jrandom> weekly status notes posted up @ http://dev.i2p.net/pipermail/i2p/2005-November/001186.html
15:05 < jrandom> (lets see if this cat will let me use both hands to type...)
15:06 < jrandom> ooh, looks like we're a few minutes early (damn clock skew), but maybe this'll make up for being a few minutes late before ;)
15:07 < jrandom> anyway, jumping into 1) and net status
15:08 < jrandom> I don't have much to add beyond whats in the status notes
15:08  * cervantes is waiting till the correct time to say hi
15:08 < jrandom> heh
15:09 < jrandom> you've got 19 seconds, acording to timeanddate.com :)
15:09 <@cervantes> hi
15:09 < jrandom> ;)  ok, anyway, anyone have any comments/concerns on  from what i can see, its gone prety well
15:10 <+Complication> Counted 747.6 routers today :P
15:10 < jrandom> yeah, we've had a higher churn than usual lately
15:10 < jrandom> still getting a bunch of referrers in from that digg / gotroot article
15:10 <+Complication> And one trick to "knowing" more peers is simply restarting less often :)
15:10 < jrandom> heh true enough
15:10 <@cervantes> *cough*sourceforge*cough*
15:11 <+polecat> I've had some trouble with number of participating tunnels dropping abruptly now and again.  Might be that darn NAT thing.
15:11  * jrandom cringes.  have you had many referrers from sf cervantes?
15:11 <+Complication> cervantes: you mean *that* SourceForge page? :eek:
15:11  * cervantes doesn't log referrers
15:11 < jrandom> hmm polecat, could be a problem with your nat, but dropping the # of participating tunnels isn't really a bad thing - it /should/ do that 
15:11 < jrandom> ah ok cervantes 
15:12 <+polecat> Really?  I thought lots of participating tunnels was good.
15:12 <+Complication> polecat: mine used to drop them rapidly when it exceeded practically usable bandwidth
15:12 <@cervantes> I make a point of only logging the minimum required to debug forum issues ;-)
15:12 <@cervantes> since folk seem touchy on the subject
15:12 <@cervantes> I've noticed...
15:13 < jrandom> polecat: sure, but the # should drop if your machine gets loaded or otherwise acts funky
15:13 < jrandom> reasonable enough cervantes 
15:13  * jrandom logs everyone's mother's maiden name, to remind people not to trust anyone ;)
15:14 < jrandom> (or do I?  you'll never know ;)
15:15 < jrandom> polecat: is your nat just randomly restarting, or losing its ip address, or something else?
15:15 <@cervantes> yeah I might change my mind on that issue....it's way too much fun seeing where everyone follows links from :P
15:16 < jrandom> its how i found the got-root and digg articles :)
15:16 < dust> i've noticed better network throughput lately, or it it just me imagining things again?
15:17 < jrandom> it should be better, especially for short lived connections (e.g. http responses)
15:18 < jrandom> otoh, its not as much of an improvement as i had hoped, so there's still work to be done on that front
15:18 < dust> e.g. i2phex pretty much takes any limit i give it if i give it enough parallel transfers
15:18 < jrandom> nice
15:20 < dust> per-tunnel seems limited <~10k/s
15:20 < dust> or per-transfer
15:20 <+polecat> Okay, my machine does get loaded now and again.
15:21 <@cervantes> have any of the folk on bandwidth restricted connections noticed improvement?
15:22 < jrandom> hmm right, 10KBps per stream seems ballpark to what i'm seeing as well
15:22 < jrandom> cervantes: i think we scared 'em all away (but if anyone with a modem or Really Shitty Connection wants to give it a try and report back, it'd be appreciated :)
15:23 < jrandom> ok, if there's nothing else on 1), lets move on to 2) boostraps, predecessors, global passive adversaries, and CBR
15:23 < jrandom> there's been lots of discussion on this fron over the mailing list (october had more posts than any other month since i2p started!)
15:24 < defnax> did anywere looked eepsites.i2p?
15:24 < jrandom> beyond whats in the status notes, i'm not sure what i have to add at the moment.  anyone have any questions/comments/concerns?
15:24 <@cervantes> I think you've successfully generated full tunnel CBR by maintaining a constant level of i2plist mails
15:24 < jrandom> heh cervantes 
15:24 < jrandom> defnax: yeah, it looks neat, a nice database growing there
15:25 < jrandom> same with tino.i2p
15:25 < defnax> but i dont like it
15:25 <+polecat> Hey, I'm on a bandwidth restricted connection!  i2p gets 10K/s up and 32K/s down.  :)
15:26 < defnax> www.eepsites.com then can all normal users on internet search for i2p sites
15:26 < defnax> and mpaa or riaa can browse what sites avaible and 
15:26 < jrandom> so?
15:26 < jrandom> mpaa/riaa/etc can just run i2p if they want to see whats on i2p
15:26 < jrandom> w3wt polecat 
15:27 < jrandom> (jesus, some sick search queries @ eepsites.com...)
15:27 < defnax> thats not good for anonnymity
15:27 < defnax> then all users now where can find torrents on I2P eepsites
15:27 < jrandom> defnax: same with tino.i2p
15:27 <@cervantes> I like the faux google ads on eepsites.i2p.... but anyway, digression
15:27 < jrandom> defnax: thats not good for /secrecy/.  thats different from anonymity.
15:27 < jrandom> people hosting public eepsites should expect that anyone can view their eepsites
15:28 < jrandom> if they want to restrict who can access it, they should do so
15:28 < jrandom> yeah definitely cervantes :)
15:28 <+polecat> Anyone who wants a private eepsite, just don't give it a name in hosts.txt.  Problem solved!
15:28 < defnax> but normal internet users must dont know which eepsites is avaible!
15:28 < jrandom> polecat: thats not entirely sufficient
15:29 <+polecat> Really?
15:29 < jrandom> i'm sorry, perhaps I misunderstand defnax.  why shouldn't people know what eepsites are available?
15:29 < defnax> i know it then not listend in a search enigine when dont make eespite public
15:29 <+polecat> I thought it was a brute force sweep through the base64 keyspace to find eepsites...
15:29 < jrandom> polecat: right, someone can harvest the netDb
15:29 < defnax> i think this person dont need I2P
15:29 < jrandom> well, harvesting for leaseSets is substantially more work than harvesting for routers...
15:30 < jrandom> defnax: I'm sorry, I don't understand
15:30 < jrandom> eepsites.com is a public interface to a search engine across public eepsites.  nothing private is being revealed
15:30 <@cervantes> lol @ last 5 searches
15:30 <+Complication> Yep, the "last searches" box suggests someone (ironically someone who is probably non-anonymous) is a tiny bit sick.
15:30 <+Complication> Ah, whatever.
15:30 < defnax> i mean he dont need I2P ! he tells on public IP on which eespites are torrents or other things avaible!
15:31 < defnax> on I2P is ok , but not on normal internet
15:31 < jrandom> defnax: sure, that person running eepsites.com does not themselves require i2p.  you can find out their home address, phone number, etc.
15:31 < jrandom> but, on the other hand, same goes for forum.i2p.  
15:31 < jrandom> (and to some extent, www.i2p, though that doesn't give you /my/ info ;)
15:32 < jrandom> some sites are public.  thats fine.  thats neat.  
15:32 <+fox> <jme___> defnax, what attacks is possible thanks to this site, which isnt without this site ?
15:32 <@cervantes> Complication: snigger
15:32 < jrandom> they're offering a potentially useful service to people who may want to try out i2p before installing
15:33 < defnax> ok any news from I2Psnark? 
15:33 <+Complication> cervantes: yeah, nothing like good old irony :)
15:33 < defnax> i will become before 0.6.2 a webinterface/gui?
15:33 <@cervantes> defnax: there's been an i2p inproxy for several months
15:33 < jrandom> defnax: nope, but I suppose we should jump forward in the agenda before we hit 6) ???
15:33 < jrandom> ok, is there anything else on 2) boostraps, predecessors, global passive adversaries, and CBR?
15:34 < jrandom> or shall we move on to 3) I2Phex
15:34 < jrandom> [consider us moved]
15:35 < jrandom> ok, those not yet on should upgrade, as there's some important stuff in the release.  Those already on who want to help test out some not-yet-released improvements, there's further work in CVS, so if you try that out and run into troubles, please post on the forum
15:36 < jrandom> in other news, I've heard some good progress on the gwebcache front as well, but no word on its integration with i2phex yet
15:36 < jrandom> redzara: any word on the merge?
15:37 <+Complication> The post .34 cvs improvements seem to make the GUI *much* more responsive.
15:38 < jrandom> cool, yeah, I couldn't handle the <= ..34 responsiveness, but i'm not sure if the fixes are entirely regression-free, since i don't really understand all the code.  but it /seems/ ok :)
15:42 <+redzara> jrandom : sorry, but we have just change hour time in France to Winter Time and work is near finished for I2phex, I've just to track 2 or 3 bugs
15:43 < jrandom> ah cool!
15:43 < jrandom> no rush, just wondering
15:44 <+redzara> and maybe I've to get lastest I??phex code to see if GregorK's mod apply to lastest phex code ?!?
15:45 < jrandom> yeah, disabling the remote request functionality will be required, but it was a trivial two line fix (comment out MAGMA and URI requests).
15:45 < jrandom> same for the latest synchronization issue (remove unnecessary locks on network operations)
15:46 <+fox> <jme___> I??phex <- interesting how typo can reveal location too :)
15:46 < jrandom> not as much as "in France" does ;)
15:46 <+redzara> this is already done in my code
15:46 <@cervantes> hehe
15:46 < jrandom> (but thats another bug thats not yet fixed... the irc charset stuff)
15:46 < jrandom> ok cool redzara 
15:47 <+redzara> jme___ : I don't search to hide my location, you konw :-)
15:47 <+redzara> so nothing more to say about i2phex for me
15:47 <+fox> <jme___> redzara, ok :) 
15:48 < jrandom> ok great, thanks for the update
15:48 < jrandom> anyone else have anything on i2phex, or shall we move on to 4) voi2p app?
15:49 <+redzara> consider we are moving :)
15:49 < jrandom> for 4), I'm not sure if I've got anything to add beyond whats in the mail, and it seems aum has disconnected, so we'll probably have to wait for an update at a later time
15:49 < jrandom> (unless someone else has something to discuss for 4)?)
15:50 < jrandom> if not, consider us moved to 5) syndie and sucker
15:50 < jrandom> dust: wanna give us the word?
15:51 <@cervantes> so is syndie good at sucking now?
15:51 < jrandom> yes *cough*
15:51 < dust> heh
15:52 < dust> well, the note says pretty much it
15:52 < dust> there are still stuff to do
15:53 < dust> word on the testing and reporting of bugs
15:54 < jrandom> ok great, do you know offhand what the story with rome-0.8 is?  worth waiting for a pending release, or should we grab a cvs build and upgrade it later?
15:55 <+fox> <brutus> oi, how about the auto ping-pongin' ircProxy, any progrssthat?
15:55 < jrandom> no progress i know of
15:55 <+fox> <brutus> (ups, sry)
15:55 <+polecat> voi2p, make a mp3 of your voice and i2p bittorrent it.
15:56 < dust> no, dunno the eta on next rome
15:56 < dust> i couldn't get to the cvs
15:57 < dust> (i don't remember why)
15:57 < jrandom> ah 'k, well, we don't need it yet, it'd just be neat.  a later date then
15:58 < jrandom> ok, anyone have anyhing else on 5)? or shall we move on to 6) ???
15:59 < jrandom> [consider us moved]
15:59 <@cervantes> brutus: I don't beleive anything has been done concerning that
16:00 < dust> should it be done?
16:01 <+fox> <brutus> oki, yeah, i guess it's pretty low priority as well
16:01 <+polecat> I still want to know how we could make i2p, and anonymity techniques in general, more accessible in poor or dangerous places.
16:01 < jrandom> polecat: by getting someone with a dialup connection to help test ;)
16:01 <@cervantes> free rifle with every installation?
16:02 < jrandom> polecat: we're definitely working on that, but we have so, so much more to do first.
16:02 < jrandom> dust: the irc thing?  might be worthwhile, but sucker improvements are probably more important
16:02 < jrandom> (imho)
16:03 <@cervantes> (somewhat biases opinion ;-)
16:03 <@cervantes> *biased
16:03 < jrandom> true that, but i think my bias is Right :)
16:04  * cervantes notes the capital lettering ;-)
16:05  * Complication looks at the phone socket, and wonders if anything good can come of such <things> :D
16:05 <+Complication> Then again, if DSL travels over it, it cannot be inherently evil. :D
16:05 <+polecat> No... not Things!
16:05 <@cervantes> Complication: you can also use it to call people....
16:06 < jrandom> ok, anyone have anything else for 6) ???
16:07  * cervantes wasn't sure we had anything for ??? in the first place
16:07 < jrandom> in that case... 
16:07  * jrandom winds up
16:08  * jrandom *bafs* the meeting closed