I2P dev meeting, January 4, 2006

Quick recap

  • Present:

bar, cervantes, Complication, gott, JosephLeBlanc, jrandom, tethra,

IRC 完整日志

15:22 < jrandom> 0) hi
15:22 < jrandom> 1) Net status and
15:22 < jrandom> 2) Load testing results and peer profiling
15:22 <@cervantes> jrandom: arguably the slowest most horrific punishment since they banned stoning
15:22 < jrandom> 3) 2005 review / 2006 preview / ???
15:22 < jrandom> 0) hi
15:22 < gott>   falafel
15:22 < gott>        n : small croquette of mashed chick peas or fava beans seasoned
15:22 < gott>            with sesame seeds
15:22 <@cervantes> hullo
15:22  * jrandom waves after recovering from the falafel pelting
15:22 < jrandom> weekly status notes posted up at http://dev.i2p.net/pipermail/i2p/2006-January/001246.html
15:23 < jrandom> as I'm 10m late, I'm sure y'all have already read those notes and have comments ready
15:23 < jrandom> *cough*
15:23 < jrandom> ok, lets jump on in to 1) net status and
15:24 <@cervantes> ie. it works well, except protocol is slowing it down
15:24 < jrandom> I don't have much to add beyond whats in the mail - we had cooking for a while before the release, and its gone pretty well from what I can tell
15:25 < jrandom> heh yeah, I'm not sure if infoshop has posted anything since the last rss import
15:25 < JosephLeBlanc> Hmm, well I have just a couple of comments.
15:25 <+Complication> With regard to status, though I've mentioned before... after build -7 got into CVS, my Celeron 300 seems to behave like an actual computer. It like, transfers data with a resemblance of stability.
15:26 <+Complication> Lets me extensively browse eepsites, and only rarely kicks me from IRC.
15:26 < jrandom> word Complication 
15:26 < jrandom> whats up JosephLeBlanc 
15:26 <@cervantes>  /kick complication
15:26 <@cervantes> doh
15:26 < JosephLeBlanc> I am running the latest CVS and, afaics, most things are in order.
15:27 < JosephLeBlanc> However, I was wondering if that jbigi athlon problem was fixed which I talked to you about some time ago.
15:28 < jrandom> jbigi for amd64 on window isn't yet in jbigi.jar, though I hope to get it into
15:29 <+Complication> I *think* (but can't confirm) that it's focusing more tightly on peers it has seen success with... and this approach *may* keeps those paths from collapsing more effectively (it's behind a somewhat too agressive NAT).
15:29 < JosephLeBlanc> Well, the logs are returning: 'NOTICE: Resource name [jbigi] was not found' 
15:30 < jrandom> JosephLeBlanc: there is a line after that regarding jbigi - what does it say?
15:31 < JosephLeBlanc> It reads, "INFO: Optimized native BigInteger library 'libjbigi-linux-athlon.so' loaded from resource
15:31 < jrandom> ok great
15:32 < JosephLeBlanc> Just wanted to give you a heads up about that NOTICE line.
15:32 < jrandom> that means it first tries to pull the resource "jbigi", but it doesn't exist (which is normal - the jbigi resource is for very rare situations)
15:32 < jrandom> it then tries to pull the OS/architecture specific resource "libjbigi-linux-athlon.so" and succeeds
15:33 < JosephLeBlanc> Ah, okay. Then, it seems that whatever bug I was experiencing earlier has been fixed in -7
15:33 < jrandom> w3wt
15:33 < JosephLeBlanc> Thanks a billion, bud.
15:34 < jrandom> np
15:34 < jrandom> Complication: aye, I think you're right, and some of the strategies for 0.6.2 will build on that concept further
15:35 < jrandom> ok, anyone else have something for 1) net status /
15:37 < jrandom> if not, lets move on to 2) Load testing results and peer profiling
15:39 < jrandom> ok, lots of stuff in the email, does anyone have any questions on it?
15:40 <+bar> how big was the resonable improvement?
15:41 <+Complication> Late remark about (just tested with my laptop) - jbigi was loaded correctly there too.
15:42 <+Complication> Previously, this machine (Mandriva 2005) was failing to load the correct one.
15:42 <+Complication> Due to the jcpuid problem.
15:42 < jrandom> I'd rather not quote a number in the meeting, as it'll affect people's expectations.  measureable improvement, but nowhere near wire speed.
15:43 < jrandom> (and the load test is a bit contrived)
15:43 <+bar> alrighty np :)
15:44 < jrandom> cool complication.  Yeah, I finally bundled up scintilla's C jcpuid port :)
15:45 < jrandom> ok, I don't have much to add on 2) beyond whats in the mail.  More info on the resulting speed profiling will come out once its shipped in
15:47 < jrandom> if there's nothing else on that, jumping on to 3) 2005 review / 2006 preview / ???
15:49 <+bar> i agree with what's in the mail, 2005 was a fantastic year and i can't see 2006 becoming any worse, i2p wise
15:49 < jrandom> we've come a long way in the last year, and y'all have done lots of great work pushing us forward.  this coming year looks to be the big one for us, moving out of the geeky backrooms and into the venues where it matters.  
15:50  * tethra came in towards the wrong end of 2005, but it was still pretty damn cool. *thumbs up*
15:51 <+Complication> Hoping for the best.
15:51 < jrandom> I don't have much more to add, so if there's other stuff that people want to bring up for the meeting, the floor is open
15:52 <+Complication> Speaking of which, if someone wanted to translate some docs into Spanish (see Discussion forum), whom would it be best for them to coordinate with?
15:52 <+Complication> e.g. which parts are liable to change heavily, or such matters
15:53 <+Complication> Also, in which format would it be best to have various docs?
15:54 < jrandom> the tech intro doc @ dev.i2p.net/cgi-bin/cvsweb.cgi/i2p/router/doc/techintro.html?rev=HEAD is fairly stable
15:54 <+Complication> Included with the router... on the website... both?
15:54 < jrandom> (there will be updates on 0.6.2 though)
15:54 < jrandom> well, I'm hugely in favor of inline documentation
15:54 < jrandom> but that should probably wait until the new router console is in place
15:55 < jrandom> website intro docs would be good as well, but that should probably wait until the new website is in place
15:55 <+Complication> Aha... so it would be best to not touch that yet...
15:56 <+Complication> ...instead preferring docs like the above.
15:58 < jrandom> probably.  ok, is there anything else people want to bring up for the meeting?
15:59 < jrandom> if not...
15:59 < tethra> did we discuss the eepget UI idea yet?
15:59 < gott> If the meeting is still on, please fix the trouble with accents in IRCclient
15:59 < tethra> ie, before i got here
15:59 < jrandom> nope, wanna implement it tethra?  :)
15:59 < jrandom> gott: patches welcome
16:00 < gott> If not, please fix the trouble with accents in IRCclient.
16:00 < gott> jrandom: By the time, I fix it, it will be fixed by somebody else ;-)
16:01 < tethra> jrandom: i'm not entirely sure how to go about it, as i'm not much a coder. if you feel like pointing me at any tutorials in writing .war type apps and i'll be glad to have a shot, though.
16:01 < gott> I reserve the right to be defeatist in deed and action.
16:01 < tethra> haha
16:01 <+Complication> :)
16:02 < jrandom> tethra: the oreily servlets books are pretty good
16:02 < jrandom> ok, if there's nothing else...
16:02  * jrandom winds up
16:02  * jrandom *baf*s the meeting closed