20:00:05 0) Hi 20:00:05 1) 0.9.30 update (zzz) 20:00:05 2) UI branch status - (str4d) 20:00:05 3) I2P Summer Dev plans - (str4d) 20:00:05 4) EdDSA update - (str4d) 20:00:09 0) Hi 20:00:12 hi 20:00:28 Hello 20:00:33 hi 20:01:00 1) 0.9.30 update (zzz) 20:01:24 ok, things going well, testers are finding some problems which is great. Jetty 9 going pretty smoothly so far 20:01:35 Hi! o/ 20:01:54 I've identified the plugins that need updating on zzz.i2p... for best results those need to be rebuilt before the release 20:01:56 mostly yes 20:02:04 except for old jetty config files 20:02:24 glad that we are doing it now so we will be ready for stretch and zesty 20:02:36 yes 20:02:56 thanks to echelon and others for testing. Will probably mark a dev build as -rc sooner than usual to get more testing 20:03:12 ok 20:03:20 hi 20:03:34 I have set the checkin deadline for friday april 28, release first week of may 20:03:54 anything else on 1) ? 20:04:15 when will Java have GOST signatures? 20:04:41 2019 20:05:01 atoi we haven't agreed to the proposal yet, if we do, best guess late next year due to priorities 20:05:02 bu fine with 1, good we have postponed UII 20:05:54 but I've made some other proposals (136 and 137) to make the introduction of new sig types easier 20:06:00 anything else on 1) ? 20:06:02 can't wait to have GOST working 20:07:22 2) UI branch status - (str4d) 20:07:35 str4d, what's the latest on your ui branch? 20:07:53 UI branch has been relatively stable for the last few weeks 20:08:28 plan is to prop in early may for .31 ? 20:08:57 There's some theme updates I need to integrate and push addressing some feedback, but in terms of structure it's basically there 20:09:08 Yeah 20:09:13 ok great 20:09:19 anything else on 2) ? 20:09:24 can´t wait to test it^^ 20:10:00 If anyone wants to test but doesn't want the hassle of building, I'll throw up an i2pupdate.zip once these next theme changes are in 20:10:18 I wait for prop... 20:11:02 3) I2P Summer Dev plans - (str4d) 20:11:19 str4d what are the plans? 20:11:47 Okay, sadie and I threw around some ideas, and what shook out was that for this Summer Dev, we should focus on speed 20:12:37 Because a) it's a logical extension of last year (we made it easier for apps to use I2P, but now we need to make them *want* to), and b) it covers about half of our existing roadmap already 20:13:37 ok, where might we find these plans, and how do we find people to do them? 20:14:06 sadie took notes from my rambling, so I'll pester her for them :) 20:14:18 I'm basically thinking this means: 20:14:51 - Push forward the various proposals that affect speed (LS2, newer E2E encryption, massive multihoming) 20:15:24 - Get something running similar to Tor's bwauth that we can use to start collecting metrics specifically about network speed 20:16:27 I was thinking of writing some code to test I2P speeds with different programs (rsync, torrents, sftp etc) 20:16:34 - Expose tunnel selection through I2CP (a la psi's lua stuff) 20:16:51 I can share that if it will be useful/interesting 20:16:51 massive multihoming <3 20:17:00 sounds good. should we put this on the agenda for next month's meeting to get an update? 20:17:00 - ElGamal speedups for the short term 20:17:17 Yep. 20:17:35 I note that most of what's on the .30 roadmap, set only 3 months ago at CCC, is total fiction and will be pushed out. I'll be updating it shortly. 20:18:10 anything else on 3) ? 20:18:16 We have two months until it will officially start. I'll continue chatting with sadie to flesh out a plan, but what we *really* need is people to pick things they want to work on 20:18:32 manas already has a head-start there ;P 20:18:35 what kind of speedup? 20:18:48 orignal_, the table precomputation 20:18:48 :) 20:19:03 mine or something else? 20:19:09 if you and sadie document a framework and options that will help people pick things 20:19:22 +1 20:19:30 anything else on 3) ? 20:19:35 orignal_, Won't be yours, because that directly leverages OpenSSL which we don't have access to 20:19:52 I mean basis 20:20:03 not implementation 20:20:28 Oh, yes likely based on yours (although I'd want to understand the mathematics myself) 20:21:03 4) EdDSA update - (str4d) 20:21:10 good to know :) 20:21:13 str4d, what's going on with EdDSA? 20:21:39 I just pushed EdDSA-Java 0.2.0, over a year after the last release 20:22:28 what's a difference? 20:22:37 Includes various cleanups, the soon-to-be-standard encoding for PKI, fixed JCA naming, a Security Provider, and should actually be constant-time now thanks to a third party that had it audited 20:22:59 I've pulled the code into i2p.i2p.zzz.test2 20:23:26 any speed improvements? 20:23:30 But what it now needs is testing to check that everything still works fine 20:23:45 how much testing have you done? 20:23:58 orignal_, speed decrease in signing due to making it constant-time (but not that much) 20:24:06 zzz, the EdDSA code itself is well-tested 20:24:20 constrant time of signing or verifing? 20:24:32 I'm still a skeptic about constant-time, but if we don't do it everybody will complain 20:24:39 signing (verifying has never been constant-time, as there's no secret info) 20:25:01 str4d, you targeting 31 or 30? 20:25:02 What I *haven't* tested yet is how it interacts with the rest of I2P, specifically relating to the JCA naming fixes 20:25:39 as I posted on zzz.i2p I think it's a fool's errand to chase the naming guidelines from some RFC 20:25:47 Probably too close to 30 to pull in as we're using the code directly instead of the library 20:25:50 the performance botlleneck is verification 20:26:22 For my library, it made sense to fix things before 0.2.0 because there wasn't a Provider before, so no one was using the JCA names 20:26:26 str4d, the prop will also be merge hell due to the javadoc changes you made that I already fixed in .29, wont be fun 20:27:14 Okay, sounds like the plan is to merge i2p.i2p into .test2, fix the merge conflicts, then get people testing it 20:27:25 but we've always had a provider. your changes in test2 tended to prefer some github PR over what I already did to accomplish the same thing, so I'm a little skeptical about the whole thing 20:27:42 you could prop that way if you want, or just defer the pain until the prop the other way 20:28:23 Yes, but if we're honest, I doubt anyone was using it 20:28:31 as I said, I defer to you, it's originally your code, but I still want to stare at it 20:29:11 I have one router with a EdDSA family key to test. can't remember what format. Remember I made all these fixes almost 4 months ago, since then we've been waiting for you and your RFC friends :) 20:29:11 (who wasn't already directly importing i2p.jar, and would already have migrations to do when they upgrade) 20:29:32 anything else on 4) ? 20:30:08 Not from me :) 20:30:13 * zzz looks for the baffer 20:30:19 anything else for the meeting? 20:31:20 Tor's next tor-dev meeting in September-ish is likely to be in Montreal 20:31:33 when? 20:31:53 Neat 20:31:55 ------------------------------------^ 20:31:56 anything else for the meeting? 20:32:14 I'm planning on going 20:32:28 thinking of switching teams? 20:32:47 Just aiming to influence them ;) 20:33:14 save yourself the plane ticket money, that's not going to happen 20:33:24 Ya Montreal is nice :) 20:33:27 Also hoping that, as Summer Dev will have just finished, we'll have some nice speed improvements to talk about there 20:34:28 * zzz *BAFS*** the meeting closed