I2P dev meeting, January 24, 2006

Quick recap

  • Present:

bar, Complication, jrandom, postman, stealth, tethra,

تمام وقایع IRC

15:25 < jrandom> 0) hi
15:25 < jrandom> 1) Net status
15:25 < jrandom> 2) New build process
15:26 < jrandom> 3) ???
15:26 < jrandom> 0) hi
15:26  * jrandom waves
15:26 < jrandom> weekly status notes up @ http://dev.i2p.net/pipermail/i2p/2006-January/001254.html
15:26 -!- Teal`c [tealc@irc2p] has joined #i2p
15:26 -!- gloin [gloin@irc2p] has quit [Connection reset by peer]
15:26 < bar> hi
15:26 < jrandom> lets jump on in to 1) Net status
15:26 -!- gloin [gloin@irc2p] has joined #i2p
15:27 < jrandom> I don't have much more to add beyond whats in the mail... anyone have any questions/comments/concerns?
15:27 <+Complication> Moving to CVS build -6 has been... challenging
15:28 < jrandom> aye, understandable
15:28 <+Complication> Net is probably doing fine. It's just my node whcih isn't.
15:28 <+Complication> =which
15:28 < bar> it's a rough road, but it's the right road. i'm 100% supportive of this move
15:29 < jrandom> tunnel building on 2+ hop tunnels is a pain, with nasty failure rates as has been reported
15:29 < jrandom> much of this is likely to be addressed with 0.6.2's new creation crypto, but I'm not convinced that all of it will be.
15:30 < jrandom> I do wonder whether we'll be able to get it reliable enough prior to that though.  But we'll try
15:31 <+Complication> If there's any stats I can provide (though you probably have more than enough of them at your own disposal) just ask
15:31 < jrandom> so, 1 hop tunnels are fairly reliable on the latest builds, but those who need 2+ hop tunnels should expect... bumps
15:31 < jrandom> thanks Complication
15:32 <+Complication> Most of my apps are 2+0..1
15:32 <+Complication> And the router itself too, if I remember correct
15:33 < jrandom> well, I could suggest staying at the release, but the release will be building short tunnels anyway if and when it encounters catastrophic failures
15:34 < jrandom> (s/short/1hop/)
15:34 <+Complication> Right, I could probably adjust it to 2+0
15:34 <+Complication> And have less spectacular effects
15:35 < jrandom> aye, though that'll still, in effect, turn to 2+/-1, but it'll try its best to stay at 2hops
15:36 <+Complication> With build -6 too?
15:36 -!- gloin [gloin@irc2p] has quit [Connection reset by peer]
15:36 < jrandom> no, the current release will fail hard rather than go to fallback tunnels
15:37 <+Complication> Or is there probability involved, which never quite goes zero?
15:37 < jrandom> the trouble there is that if it goes for 10 minutes without building the tunnels, it'll restart the router (due to the watchdog)
15:37 <+Complication> Saw it once :)
15:37 < jrandom> no, -5 or newer will use exactly the hop lengths allowed by the client (2+/-0 means only 2 hop tunnels.  never anything else)
15:39 < jrandom> ok, anyone have anything else for 1) Net status?
15:39 < jrandom> or, I suppose we're already discussing 2) New build process ;)
15:40 < jrandom> does anyone have anything else to discuss on 2) New build process?
15:40 <+Complication> Not much here, anymore :D
15:41 < jrandom> hehe ok, if not, lets shimmy on over to 3) ???
15:41 < jrandom> anyone have anything else they want to discuss?
15:42 < bar> may i ask, how many backwards incompatible changes are lined up now, and if some (all?) can be put into one release?
15:42 < bar> i mean, is there more than one backwards incompatible release planned, until 0.6.2?
15:42 < jrandom> bar: the hope is to do them all at once
15:42 < jrandom> (though there may be further ones down the line)
15:43 -!- Complication [Complicati@irc2p] has quit [Connection reset by peer]
15:43 -!- Complication2 [Complicati@irc2p] has joined #i2p
15:43 < bar> hmac bug, new crypto and restricted routes at once?
15:43 < bar> that's a tall order :)
15:43 < jrandom> restricted routes?
15:43 < jrandom> the hmac bug "fix" is changing one value ;)
15:44 < bar> ah :)
15:44 -!- Complication2 is now known as Complication
15:44 < bar> umm.. perhaps restricted routes was 2.0..
15:44 < jrandom> yeah, but restricted routes will be doable without losing backwards compatability
15:45 < jrandom> (in fact, it can be done with 0.6.2, if done carefully, to a degree)
15:45 < bar> ok, great
15:45 < jrandom> I'm also thinking of when to drop tcp... maybe in the next release
15:46 < jrandom> or maybe after, so we don't have /too much/ at once
15:49 < jrandom> ok, anyone have anything else for the meeting?
15:51 < jrandom> if not
15:51  * jrandom winds
15:51 < stealth> I have some question: I noticed that all eepsites are mapped  to the external internet e.g. http://tracker.postman.i2p.tin0.de/. Is that wanted ?
15:51 < jrandom> [saved]
15:51 < jrandom> sure, I think thats cool
15:51 < jrandom> anyone who publishes information should expect that their information is public 
15:52 -!- gloin [gloin@irc2p] has joined #i2p
15:52 < jrandom> I think tino has a way for people to topopt out as well
15:52 < tethra> that was short
15:53 < stealth> They are also indexed by google...
15:53 < jrandom> isnt that a good thing stealth?
15:53 < Complication> Did it not involve some convention similar to "robots.txt"
15:54 < jrandom> aye Complication 
15:54 < Complication> (might be best to ask tin0)
15:54 <@postman> damn, i am too late
15:54 <@postman> (again)
15:54 < jrandom> nah, hasn't ended yet postman :)
15:54 < Complication> He wrote about it in the forum, at some point
15:54 < Complication> Might be findable there
15:54 <@postman> ahh cool ( hello then) :)
15:55 < jrandom> yeah, its opt-out-able, but I don't understand the concept of opt-out for the i2p content (are people pushing some idea of 'copyright' - "don't copy my stuff or make it visible other places"?)
15:55 < jrandom> but, whatever, tino is being nicer than I would be regarding inproxies ;)
15:56 -!- Rawn [Rawn@irc2p] has quit [Connection reset by peer]
15:56 -!- gloin [gloin@irc2p] has quit [Connection reset by peer]
15:57 -!- Karellen [Karellen@irc2p] has quit [Connection reset by peer]
15:57 < Complication> Yes indeed, the assumption shouldn't follow that other providers of in-proxies will be equally nice
15:58 -!- Karellen [Karellen@irc2p] has joined #i2p
15:58 -!- Rawn [Rawn@irc2p] has joined #i2p
15:58 -!- mode/#i2p [+v Rawn] by chanserv
15:59 < Complication> Information intended to be secret... is best simply not published
15:59 < tethra> indeed :/
15:59 < stealth> Yes but it might turn too much publicity to i2p before evrything is really totally save. The problem seems to me that I2p has at the moment not enough nodes for a very good anonymity...
16:00 -!- Complication [Complicati@irc2p] has quit [Connection reset by peer]
16:00 < jrandom> our anonymity isn't dependent upon the size, and i2p has been googled plenty
16:01 < jrandom> (or, the base level of anonymity isn't dependent upon the size)
16:01 < jrandom> but, of course, no one who needs hard anonymity should use i2p now.
16:01 -!- digger3 [digger3@irc2p] has quit [Connection reset by peer]
16:01 -!- digger3 [digger3@irc2p] has joined #i2p
16:02 < bar> i wouldn't worry, 99% would just ignore the seemingly dead link that turns up on google... the other 1% is likely somewhat geeky and would want to know more
16:03 -!- gloin [gloin@irc2p] has joined #i2p
16:03 < bar> (well.. dead, that depends on tino's inproxy being up or not, of course)
16:05 < jrandom> ok, anyone have anything else for the meeting?
16:06 < jrandom> if not
16:06  * jrandom winds up
16:07  * jrandom *baf*s the meeting closed