Hi, Here's the summary of the IRC meeting.
--- COMMUNITY MEETING Place: #openvpn-meeting on irc.freenode.net Date: Thu 24th September 2020 Time: 20:00 CEST (18:00 UTC) Planned meeting topics for this meeting were here: <https://community.openvpn.net/openvpn/wiki/Topics-2020-09-24> Your local meeting time is easy to check from services such as <http://www.timeanddate.com/worldclock> SUMMARY becm, cron2, dazo, lev and mattock participated in this meeting. --- Talked about OpenVPN 2.5. Noted that MSI is getting into a nice shape, though we have some fixes pending for it. We also have a new tap-windows6 release in the pipe. Agreed that we want to do an RC2 release for 2.5, mainly because there is a fair amount of changes related to Windows. Set the release date for 2.5-rc2 to "mid next week". This gives mattock some time to build a new tap-windows6 driver as well as get his other urgent stuff out of the way. --- Noted that Trac tickets related to OpenVPN Connect should be assigned to denys.babets. He will take them from there. -- Full chatlog attached
(21:00:08) mattock: hello (21:01:41) dazo: hey! (21:01:45) dazo ha scelto come argomento: Agenda at https://community.openvpn.net/openvpn/wiki/Topics-2020-09-24 (21:01:46) lev__: guten nacht (21:01:59) dazo: lev__: that's when the day starts! (21:02:20) lev__: then guten abend (21:02:33) dazo: :-P (21:03:50) dazo: Btw ... how's the MSI stuff going? (21:04:12) lev__: I think it is in a good shape (21:04:41) dazo: Great! (21:05:19) lev__: there is one issue which could be fixed, though, but I wouldn't call it showstopper (support for modify/repair) (21:05:39) dazo: so that's more a feature than a bug? (21:05:57) cron2: now! (21:06:01) mattock: hi (21:06:25) lev__: both (21:06:45) mattock: there is also some activity on tap-windows6 (21:06:51) lev__: unprivileged user may break VPN connection of administrator (21:07:13) lev__: but we have the same problem with Connect (21:07:22) mattock: btw. 2.0 had 18 release candidates (21:07:31) mattock: we can do a few before breaking the record (21:07:35) mattock: https://build.openvpn.net/downloads/releases/ (21:07:37) vpnHelper: Title: Index of /downloads/releases/ (at build.openvpn.net) (21:09:57) cron2: working on it :-) (21:10:07) cron2: mattock: have you seen that we need a new tap driver? (21:11:52) mattock: yes, and selva said "wait" (21:11:57) mattock: did not check later today (21:13:09) lev__: cron2: what is the thing with new tap driver ? (21:14:31) cron2: changing the MAC address via control panel was doubly broken (21:14:43) cron2: .inf file was broken, setting the wrong key - that was fixed yesterday (21:14:54) cron2: driver was using the wrong variable for ARP and ND - that was fixed today (21:15:04) cron2: so, now things should be fixed for good (21:16:45) lev__: cool, we should update Connect client as well (21:19:16) mattock: what about the next rc? (21:19:19) mattock: what, when? (21:19:56) cron2: you tell me :-) - it should come with the new tap driver. (21:20:01) cron2: I can do the RC2 any time (21:20:29) lev__: I would like to add dhcp-option DOMAIN support for wintun (21:20:31) dazo: Do we need another RC? (21:20:52) dazo: $ git shortlog v2.5_rc1..release/2.5 (21:20:52) dazo: Simon Rozman via Openvpn-devel (4): (21:20:52) dazo: netsh: Specify interfaces by index rather than name (21:20:52) dazo: netsh: Clear existing IPv6 DNS servers before configuring new ones (21:20:52) dazo: netsh: Delete WINS servers on TUN close (21:20:52) dazo: openvpnmsica: Simplify find_adapters() to void return (21:21:09) dazo: These are bugfixes ... doesn't look that scary even (21:21:32) cron2: dazo: not on the unix side, but windows with the new driver and "netsh config fixes" might need another round of testing (21:21:53) cron2: also, we are aiming to beat 2.1, which made 2.1_rc22 (not to forget having an rc21b!) (21:22:07) dazo: hahahaha .... oh dear! (21:22:09) mattock: 2.0 actually (21:22:16) mattock: 2.1.4 was the last 2.1 (21:22:18) dazo: no, 2.1_rc22 was a thing (21:22:31) cron2: mattock: release candidates come before 2.1.0 :-) (21:22:33) dazo: 2.1.4 was the last release (21:22:50) cron2: commit 1852709cd5093995f97ba4860d1a6083c6df6d6c (tag: v2.1_rc22) (21:23:10) cron2: anyway (21:23:13) mattock: yes of course, but 2.0-rc18 was the biggest achievement in RCs so far (21:23:23) mattock: but yeah (21:23:30) mattock: I propose "mid next week" for rc2 (21:23:31) cron2: mattock: uh - no? 2.1_rc22 is more than 2.0_rc18? (21:23:35) dazo: mattock: you're awake? 22 comes after 18 ;-) (21:23:44) mattock: where do you guys see 2.1_Rc22? (21:23:49) mattock: not on build.openvpn.net? (21:23:55) dazo: git tag -l | grep v2.1_rc (21:23:58) mattock: ok (21:24:00) cron2: in the commit quoted above (21:24:03) mattock: I withdraw my objections (21:24:10) mattock: that's pretty incredible (21:24:13) cron2: not sure what release had the most versions actually built and published :-) (21:24:45) cron2: anyway, mid next week, it is (21:24:47) mattock: anyhow, mid next week because tap-windows6 will consume half-a-day plus I just received "urgent" things for tomorrow (21:25:04) cron2: there is the time fix v3, plus another simon msica fix that is under discussion (21:27:00) cron2: plaisthos is fighting some convoluted stuff which is more likely to end up in 2.5.1 (21:27:41) cron2: dazo: the 2.0 releases are not tagged in git (21:28:26) cron2: 2.2 was amazingly short-lived, only up to 2.2.3 (21:28:29) dazo: cron2: that's right ... because our git history starts at the BETA21 SVN branch .... the 2.0 and prior releases are probably in the historical-cvs git repo (21:28:48) cron2: you have that old stuff? cool (21:28:48) dazo: we dragged our feets with 2.4 :-P (21:29:00) cron2: v2.3.18 (21:29:01) dazo: that history is not as fine-grained as the 2.1 history (21:29:06) cron2: v2.4.9 (21:29:09) cron2: yeah, indeed :) (21:29:22) dazo: https://github.com/OpenVPN/openvpn2-historical-cvs (21:29:24) vpnHelper: Title: GitHub - OpenVPN/openvpn2-historical-cvs: Just a read-only repository containing the historical OpenVPN 2.0 releases, pulled out of the now non-existing CVS repository. Each commit is essentially a new version, so not a fine grained overview, but gives an impression of the development process. (at github.com) (21:29:41) cron2: nice (21:29:43) mattock: if keep on halving the last number we'll soon move to the Google Chrome versioning scheme (21:30:03) cron2: let's just make the next community release real good, and make it "v4" (21:30:10) mattock: \o/ (21:30:12) cron2: so it's clear "it's more advanced than 3!" (21:30:18) mattock: yeah! (21:30:21) dazo: hahahaha (21:30:56) dazo: v4 should be an openvpn re-write in Rust :-P (21:31:53) cron2: so... anything else on 2.5 or tap? (21:31:59) mattock: not from my end (21:32:40) cron2: there's an ongoing discussion about the lack of DNS domain setting in wintun, but I'm not sure if it will lead to an actual patch (21:32:50) cron2: (for 2.5.0, that is) (21:33:05) becm: tap0901.sys-removal on uninstall (21:33:09) ***becm duck (21:33:21) cron2: becm: what does that mean, for "not the windows guy" people? (21:33:26) lev__: I can send the patch tomorrow, the code is fairly small (and has been in openvpn for 6 years) (21:33:40) lev__: *openvpn3 (21:35:24) lev__: after uninstalling client, driver still presents in system, to actually remove it one needs to use pnputil (21:35:47) cron2: is that with msi and msm packaging? (21:35:49) cron2: or nsis? (21:36:14) lev__: all of them (21:37:41) lev__: I don't know if this is a bug or feature (21:38:25) mattock: you're speaking of the drivers in the DriverStore remaining behind? (21:38:28) becm: since the install/unistall is (nearly) identical to wintun, I'd opt for bug. (21:39:04) becm: that one gets removed just fine. (21:39:41) mattock: ok (21:40:03) cron2: that would have been my next question :) (21:40:36) becm: it's not realy (regular) user-visible and negligable on a 16GiB Windows-Installation. :) (21:40:57) lev__: IIRC wintun has some sort of reference counting, where each user (like wireguard/openvpn) increments it by one, and when it decreases to zero driver gets removed (21:42:11) mattock: that does ring a bell (21:42:41) cron2: I've heard this mention as a generic plus for "MSM", so I expected this (without knowing anything) to be "automatic" (21:43:02) cron2: do we delete tap interfaces on deinstall? (21:43:14) lev__: OTOH this bug has a nasty consequence - if you happen to install a driver with a wrong signature, upgrade will not resolve it since windows would assume that driver already presents. We had this issue on Johan's laptop during last hackathon (21:44:08) lev__: we do delete interfaces (21:45:06) cron2: THat should be good enough then... but, I have no idea what MSM and MSI really do (21:46:34) lev__: becm: would you mind to create an issue to tap-windows6 about this (driver doesn't get deleted) (21:47:58) becm: lev__: can we rule out the double-registration (root\tap0901, tap0901) as the reason? (21:49:19) lev__: we could, but maybe not right now :) (21:50:18) lev__: mattock: do you have download numbers for 2.5 beta4/rc1 ? (21:50:27) mattock: unfortunately no (21:50:30) mattock: well (21:50:32) mattock: possibly (21:50:41) mattock: if I remembered where I need to login to get them (21:50:43) mattock: :D (21:51:00) mattock: I'll try something (21:51:23) dazo: Anything else? (21:52:17) cron2: not from me (21:52:32) mattock: no luck (21:52:40) mattock: nothing from me (21:53:10) dazo: One FYI on Trac tickets related to OpenVPN Inc products .... A new user should be signed up here now, denys.babets, who will be the ticket owner for the Connect and non-OSS related things (21:53:59) dazo: So anything Connect, Access Server or OpenVPN Cloud related can be reassigned to Denys (21:54:08) cron2: *thumbsup* (21:54:11) mattock: do we want to keep those tickets in Trac? (21:54:26) mattock: or just assigned them so that Denys can close them and move them to a more proper place? (21:54:37) dazo: He will follow-up on these tickets, sort out what can be sorted out in Trac and otherwise redirect to more appropriate places (21:54:43) mattock: ok (21:54:44) cron2: I have no opinion, but they need to be handled one way or the other (21:54:46) mattock: sounds good (21:54:48) cron2: yes (21:55:02) dazo: (would be silly to redirect to to our support platform just to say: You need a new config from your service provider) (21:56:28) mattock: ok, I'll write summary now (21:56:44) mattock: I propose "good night all" (21:58:02) dazo: g'night! (21:58:08) cron2: yeah. good night :)
pEpkey.asc
Description: application/pgp-keys
_______________________________________________ Openvpn-devel mailing list Openvpn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openvpn-devel
