I2P dev meeting, July 6, 2021 @ 20:00 UTC

Quick recap

  • Present:

eyedeekay, zzz, zlatinb

Полный журнал IRC

(04:01:20 PM) eyedeekay: Hi everyone, it's time for the monthly community meeting, but as I forgot to make the announcement I will not be surprised if no one is here. In case anyone else is here, I'm open to having the meeting now. If no one else is here, I'll announce a new one on zzz.i2p so we can re-schedule
(04:01:37 PM) zzz: hi
(04:01:47 PM) eyedeekay: Hi zzz
(04:02:10 PM) eyedeekay: zlatinb, anybody else here?
(04:03:39 PM) eyedeekay: Ok that's my bad. Well zzz, I've got a short 2-item agenda for us if you've got time:
(04:03:39 PM) eyedeekay: 2) Next Version Number
(04:03:39 PM) eyedeekay: 3) Jpackage Updates
(04:03:39 PM) eyedeekay: But I was hoping to have zlatinb for 3)
(04:04:06 PM) zzz: 4) remaining 0.9.50 release items
(04:04:31 PM) eyedeekay: Ack
(04:05:10 PM) eyedeekay: 2) Next version number
(04:06:27 PM) eyedeekay: I'm less and less reluctant about 1.0.0 now
(04:07:31 PM) eyedeekay: zlatinb had some ideas about where performance could be improved, and what we both agreed on was that we needed something more onboardable for 1.0.0, i.e. the jpackage things
(04:08:05 PM) zzz: I think an arbitrary jump to something like 1.5.0 or 2.5.0 or 5.1 avoids the 1.0.0 angst
(04:10:04 PM) mode (+v zlatinb) by ChanServ
(04:11:37 PM) eyedeekay: 1.5.0 seems right somehow? or maybe 1.51
(04:12:22 PM) zzz: small numbers seem better
(04:12:48 PM) eyedeekay: Yeah you're right
(04:13:36 PM) eyedeekay: 1.5.0 works for me if it works for you
(04:15:13 PM) zzz: I'll put it in a post on my forum and see what the reaction is
(04:15:30 PM) eyedeekay: Sounds good
(04:16:05 PM) eyedeekay: 3) Jpackage Updates
(04:17:24 PM) eyedeekay: On my end I've got a WIP jpackage+Windows Installer+Firefox profile bundle which should be self-updating as of this morning. It's still untested and a draft PR, I'm sure I'll find something broken about it tonight, but so far, so good
(04:17:35 PM) mode (+v anonymousmaybe) by ChanServ
(04:18:55 PM) eyedeekay: It works by starting the installer just before the router shuts down, sleeping until the router shuts down, then allowing the installer to re-start the router when it completes
(04:19:25 PM) zlatinb: not much on my end, I'm still a little stuck on figuring out how to an end-to-end test that starts from checking news.xml, fetching the update.dmg, performing the update, restarting the router
(04:19:41 PM) zlatinb: but the concept is the same as on windows
(04:22:35 PM) zlatinb: just quite a bit of infrastructure to set up I guess
(04:23:01 PM) eyedeekay: I don't have much else to add, except that I'm going to try to figure out how to test it against a test news server tonight, which should help figure out the infrastructure
(04:24:13 PM) eyedeekay: 4) Remaining 0.9.50 release items
(04:24:27 PM) eyedeekay: Oops, pasted that too soon
(04:24:37 PM) eyedeekay: Anything else on 3)?
(04:25:02 PM) eyedeekay: 4) Remaining 0.9.50 release items
(04:25:10 PM) zzz: still no debian/ubuntu, who's in charge of bugging mhatta?
(04:25:40 PM) eyedeekay: I've been bugging him as much as I can, opened a PR on bote to get his attention, not sure what's up there. No response
(04:26:11 PM) eyedeekay: Maybe I'm just not looking in the right place anymore
(04:26:25 PM) zzz: it's now been 7 months since he cut a release
(04:27:40 PM) zzz: anyway, I believe that's the only remaining item
(04:28:12 PM) eyedeekay: I heard Debian accepts anonymous maintainers if they have a portfolio and a GPG key now, I could reach out and apply? I hate to make myself even more of a bus factor, but at least I pretty much know how to go from i2p.i2p->deb
(04:30:17 PM) zzz: the problem is that I think he has several changes he never upstreamed back to us, so those differences would have to be resolved
(04:31:53 PM) eyedeekay: If he does then I think they'd have to be reflected in the debian/patches, maybe I can find a way
(04:31:53 PM) zzz: thats all I got, put it on the list for next month
(04:32:00 PM) eyedeekay: Will do
(04:32:16 PM) zlatinb: for this item I want to ask about the streaming buffer overflow
(04:32:43 PM) zlatinb: is that something we want addressed for the next release?
(04:32:50 PM) zzz: huh?
(04:32:57 PM) zlatinb: s/overflow/choke/
(04:33:14 PM) zzz: what item?
(04:33:23 PM) zlatinb: oh sorry, thought we were discussing 0.9.51
(04:33:27 PM) zlatinb: nvm
(04:33:32 PM) zzz: but no, not a pressing problem, more of a test issue, low priority
(04:34:01 PM) zzz: we were discussing .50 debs
(04:34:32 PM) eyedeekay: I've got time, if zzz's got time I'm happy to make that item 5)
(04:34:44 PM) zlatinb: yes please
(04:34:49 PM) eyedeekay: Go for it
(04:35:11 PM) zlatinb: I think it happens in the live net on short tunnels, not 0 but 1-hop
(04:35:41 PM) zlatinb: at least I've seen suspicious behavior in muwire when configured with 1-hop tunnels on both nodes
(04:36:24 PM) eyedeekay: I have a bunch of 1-hop services, is there anything I can look for in the logs to help you confirm it?
(04:37:12 PM) zlatinb: at this early stage it can be debugged in a testnet, logging is too verbose for a live server
(04:37:58 PM) zlatinb: I would like to spend some time on that and if there is a problem and a fix for it aim to put it in the next release
(04:39:21 PM) zzz: to answer your question, it's a known issue for many years, presumed very rare in live net, the impacts are transient and possibly unfixable... so it's worth investigating (and I asked for help doing that), but for those reasons I wouldn't put it as a must-fix for the next release
(04:39:56 PM) zlatinb: I think the recent speed improvements have made it less rare
(04:40:31 PM) zzz: sure. maybe, maybe not
(04:41:17 PM) zlatinb: ok I'll investigate and see what comes out
(04:41:35 PM) eyedeekay: It'll be interesting to see what you find
(04:42:10 PM) eyedeekay: Anything else for the meeting?
(04:43:09 PM) eyedeekay: All right then that's it for today
(04:43:18 PM) eyedeekay: Thanks zlatinb and zzz for being and bearing with me, I'll post the logs shortly and make sure I get the announcement up on zzz.i2p this time
(04:43:24 PM) eyedeekay: being *here