I2P dev meeting, January 2, 2007

Quick recap

  • Present:

bar, covracer, jrandom, z^z,

Full IRC Log

15:02 < jrandom> 0) hi
15:02 < jrandom> 1) Net status and plans
15:02 < jrandom> 2) Syndie 1.001a
15:02 < jrandom> 3) ???
15:02 < jrandom> 0) hi
15:02  * jrandom waves
15:02 < jrandom> weekly status notes posted up on http://dev.i2p.net/pipermail/i2p/2007-January/001325.html
15:03  * mrflibble waves to jrandom
15:03  * jrandom expects most are still nursing their hangovers, so we'll do this quietly
15:04 < jrandom> lets jump on in to 1) Net status and plans
15:05 < jrandom> as mentioned in the notes, there's been a lot covered, but we've got our work cut out for us this year
15:05 < jrandom> we'll want to discuss the various tradeoffs though and make sure the most appropriate ones are taken to support the specific functionality we're driving towards
15:06 < jrandom> but, we'll see how it goes as things progress
15:06 < jrandom> for the time being though, the net seems fairly steady-state, which is Good
15:07 < jrandom> anyone have anything they'd like to bring up re: net status and plans?
15:09 < jrandom> if not, lets hop on over to 2) Syndie 1.001a
15:09 < z^z> anything I can work on on the net - netdb or tunnels?
15:09  * jrandom hops back
15:10 < z^z> I know we have questions on netdb propagation and on cpu usage but I would need a good pointer to get started
15:11 < z^z> think about it anyway
15:11 < jrandom> z^z: netDb search when the # of known/reachable floodfill peers reaches 0 needs to, most likely, do a random iterated walk across known peers
15:12 < z^z> ok thx will poke around and ask questions later
15:12 < jrandom> perhaps a new flag on the netDb lookup message asking for "give me some floodfill peers"
15:12 < jrandom> kickass z^z!  that'd likely have a substantial impact for new users - let me know if you run into any trouble
15:13 < z^z> ha that will take me into new territory sounds like fun for the new year
15:13 < jrandom> :)
15:14 < bar> "do not delete floodfill peer router infos from netdb if there are too few of them" <-- does anyone remember if this one got into cvs or not?
15:15 < jrandom> nope
15:15 < jrandom> or, not that i recall...
15:16 < bar> okie
15:17 < jrandom> (a great spot for that would be KademliaFloodfillNetworkFacade::dropAfterLookupFailed)
15:18 < jrandom> er, KademliaNetworkDatabaseFacade, that is (floodfill extends it)
15:20 < jrandom> (there's also a few bits in the DatabaseLookupMessage that could be used to flag 'send me floodfill peers' - the 'tunnelSpecified' is a boolean, but transferred in a full byte)
15:21 < jrandom> ok, anything else on 1) Net status and plans?
15:23  * jrandom resumes the hopping to 2) Syndie 1.001a 
15:24 < jrandom> she's coming soon, maybe in a day or two.  lots of bugfixes and cleanup (thanks to everyone helping!), with more details in the announcement when its released
15:25 < jrandom> thats about it to mention on that, though (but if you're using the new syndie, you can follow up on the latest discussions there ;)
15:27 < jrandom> anyone have anything to bring up on syndie 1.001a, or shall we skip on over to 3) ???
15:27 < jrandom> anyone have anything else they'd like to discuss in the meeting?
15:28 <+fox> <covracer> are you still not in favor of an ebuild?
15:29 < jrandom> for syndie, or i2p?  
15:29 <+fox> <covracer> i2p
15:29 < jrandom> correct, i am still not in favor of an ebuild
15:29 < jrandom> (thank you for the offer/suggestion though!)
15:30 < jrandom> i2p's problems are not related to the size of the network, so increasing the size will not address them
15:30 < jrandom> instead, it will just force more people to deal with the problems and the upgrade path to address them
15:30 <+fox> <covracer> yeah
15:31 <+fox> <covracer> alex has done some good work though on an ebuild
15:31 <+fox> <covracer> it's in the java-experimental-migration overlay iirc
15:31 <+fox> * godmode0 is back (gone 01:57:51)
15:32 <+fox> <covracer> well at any rate it depends on lots improvements in gentoo's handling of java and jetty
15:32 <+fox> <covracer> and won't get into the main tree any time soon
15:33 < jrandom> cool (that alex's work is going well), and hopefully we'll get i2p to the point where pushing it to main will be a great thing :)
15:34 <+fox> <covracer> would a ebuild for syndie be welcomed or should it also be postponed?
15:34  * jrandom wonders how much shakeup there is going to be in java handling once the sun jvm & libs go gpl
15:35 < jrandom> syndie will hopefully be ready for full production use in a matter of months, with beta in maybe a month, so looking at an ebuild there would be great
15:36 < jrandom> when syndie goes production i'd like to make it as easy as possible for people to use - apt-get, emerge, rpm, etc
15:36 <+fox> <covracer> okay, I'll see if I can hack an ebuild together this week of vacation--I've got nothing better to do
15:36 < jrandom> kickass, thanks covracer!
15:37 <+fox> <covracer> easy installation is very important for wide adoption
15:37 < jrandom> (and let me know if you run into any bits that could be simplified 'upstream' - i'd like to make packaging as transparent as possible)
15:37 < jrandom> aye, definitely
15:38 <+fox> <covracer> alright, although I'm only vaguely aware of the best practices of ebuild writing, not being a dev myself or really all that active on the coding front
15:40 < jrandom> cool, you likely know more about ebuild writing than I though :)  good luck, and thanks
15:40 < jrandom> ok, anyone have anything else they'd like to bring up for the meeting?
15:40 < bar> well, i think an official post in cervantes' syndie forum and the old syndie wouldn't hurt, if/when you're looking for more testers for the new syndie
15:40 < bar> except for the last meeting log, i don't think there has been much mentioning of the alpha release, many i2p users simply haven't heard the news, methinks
15:41 < jrandom> good idea - i'll spam 'em when 1.001a is out
15:42 < bar> alritey :)
15:47 < jrandom> ok, if there isn't anything else for the meeting...
15:47  * jrandom winds up
15:47  * jrandom *baf*s the meeting closed