**** BEGIN LOGGING AT Wed Jun 02 02:59:57 2021 Jun 02 12:47:28 I'm yet to see an official announcement on the yocto mailing lists, until then I stay here :) Jun 02 12:56:43 kanavin: We are moving FWIW Jun 02 13:48:30 is there a definition of all (or most) of the data definitions which can ba d.getVar'ed in "d"? Jun 02 13:48:51 all the variables, that is Jun 02 17:26:13 I got kicked from the Yocto matrix server. Jun 02 17:26:14 Noooo Jun 02 17:26:29 Appservice kicked me for "not being registered" like it was IRC, or somehting. Jun 02 17:27:24 I have a question about OE. Can I use _remove on CORE_IMAGE_BASE_INSTALL. Jun 02 17:32:52 matthewcroughan_, the matrix channel is now bridged to libera.chat. and the channel there is set to require registration Jun 02 17:33:18 Heh, that's a bit conovluted :P Jun 02 17:33:26 you can talk to @NickServ:libera.chat to register you nick there. Jun 02 17:33:30 So what you mean is that the matrix channel was deleted, in favor of using IRC. Jun 02 17:33:37 And that I can connect to it, using the libera bridge? Jun 02 17:33:42 no Jun 02 17:34:18 the matrix channel still exists. but the bridge makes sure that all users are also on the IRC side. if they can't join, they'll be kicked Jun 02 17:34:20 So you're saying that I need to verify with the libera bridge, to get into the Yocto matrix channel/ Jun 02 17:34:24 yep Jun 02 17:34:26 That's really really weird, I don't know anybody who does that. Jun 02 17:34:32 I also don't like it, so I won't do that :D Jun 02 17:34:42 the bridge is still work in progress :/ Jun 02 17:34:42 I will join using the bridge, but I will not go through that route. Jun 02 17:35:08 Have you got any other examples of people using bridges for verifcation prior to entering the Matrix channel? Jun 02 17:35:08 you can also join #yocto on libera.chat via irc directly Jun 02 17:35:44 shoragan: If I use the bridge, on matrix, that's essentally direct, by comparison to what you've suggested I do. Jun 02 17:35:46 matthewcroughan_, it's probably temporary until the bridge users can be exempted by their cloack Jun 02 17:36:08 Why do matrix-native users have to verify with some IRC thing, prior to joining the matrix channel? Jun 02 17:36:36 For example, in order to join this Yocto IRC, it is not required that people first verify with a Matrix server/bridge. Jun 02 17:36:46 If that was the case, it wouldn't make a lot of sense, right? Jun 02 17:36:52 the "issue" is the -r mode on the irc channel on libera.caht Jun 02 17:38:00 so the irc channel requires registration. and with the bridge, that (currently) also applies to used using matrix as their "irc" client Jun 02 17:38:02 Ah I see, and I guess the bridge actually creates an instance of a user per user right? Jun 02 17:38:06 Of which, each one must be registered? Jun 02 17:38:32 yes, *currently*, until we can exempt the matrix users Jun 02 17:46:39 I think my name is too long to register lol Jun 02 17:51:05 You've been connected to the IRC network 'irc.libera.chat' as matthewcroughan[. User modes +R have been set. Jun 02 17:51:14 See how the [m] is cut off? God damnit Jun 02 17:56:12 shoragan: i've been fighting with the same issue myself.. i think it's a big downside of the bridge. it forces matrix users which might not even know that IRC exist off the channel.. Jun 02 17:56:28 it kind of does the opposite of what we wanted, no? Jun 02 18:00:15 so a matrix only user now needs to register an IRC nick to join #yocto. and an existing IRC user needs to register a new nick with irc.libera.chat to join #yocto from matrix (especially if the same user needs to join other IRC channel directly without matrix)... Jun 02 18:00:59 ndec: can't we -r the channel? I know it helps with spam but would seem a short term better solution Jun 02 18:01:18 that would be a solution, indeed. Jun 02 18:05:05 RP: matthewcroughan_ : i've done -r, so you should be able to rejoin #yocto from matrix. can you please try? Jun 02 18:07:01 ndec, you can use /mode +be $~a *!*@2001:470:69fc:105::/64 instead of -r. that bans all unauthenticated users but exempts the matrix bridge. Jun 02 18:07:42 in general, i've not seen much span on libera.chat in the smaller channels Jun 02 18:07:49 that would let all matrix users go through? without having to register their nick? Jun 02 18:08:20 yes, it should. they need to have a registered nick on the matrix side anyway Jun 02 18:08:47 s/nick/account' Jun 02 18:08:50 you mean a matrix user account? Jun 02 18:08:52 ah.. Jun 02 18:09:47 and you can ban them on them matrix side pretty effectively Jun 02 18:10:06 (if that ever becomes necessary) Jun 02 19:58:20 What is the channel name on libera? Jun 02 19:58:29 Yocto or yoctoproject? Jun 02 19:59:26 same as here Jun 02 19:59:57 Right. How do we use the matrix one? Jun 02 20:00:03 Or do we use it at all? Jun 02 20:07:22 not sure what you mean.. just connect an IRC client to the other server and connect to the channel Jun 02 20:07:40 only reason many of us are still in this channel is to help direct people to the other server, if needed Jun 02 20:09:02 "other server". Jun 02 20:09:19 Afraid if you say it out loud three times you'll turn to stone? :-P Jun 02 20:16:39 hello all, I am having a challenge on the mainline - has the compiler been upgraded there recently (this is x86_64 qemu)? I am having a build break in s2n which doesn't occur on zeus/dunfell/gatesgarth.  A couple people internally have inferred that there is some bad/incorrect platform datatype definition which is a bit concerning: Jun 02 20:16:39 https://pastebin.com/DNLnBgUf Jun 02 20:18:40 rpcme49: we are using the latest gcc 11 Jun 02 20:20:03 pretty much every new gcc comes with new warning checks enabled ; add -Werror and new build failures result. Jun 02 20:20:19 okay thanks - I saw it was new - the semantic version is like maybe 10 or 12 days ago?  I will download gcc separately at v 11 and repro on just plain ol ubuntu 18.04 and get it back to the s2n devs Jun 02 20:21:02 rpcme49: we're trying to stay with their stable updates as in theory that should let us pull in fixes Jun 02 20:21:33 this is really good I think:)  it will also keep us on our toes! Jun 02 20:40:57 abelloni: are you heading over the libera? https://autobuilder.yoctoproject.org/typhoon/#/builders/101/builds/2384 looks to be webkit causing load :( Jun 02 21:08:30 RP: yes, I will Jun 02 21:11:01 abelloni: we have better logs in the above build failure in the builddir but it just shows the process is running at nice -5 and still not managing to start qemu in 120s :( Jun 02 21:11:29 it is running webkit and llvm builds, this is looking to be a pattern Jun 02 21:25:40 alimon: around? I think we're seeing a problem with timeouts in ptest-runner :( Jun 02 21:28:22 alimon: Should the signal(SIGALRM, timeout_child_handler); be in the parent only and not the child? Jun 02 21:39:22 alimon: actually, I think its the alarm and different pthreads... Jun 02 22:19:57 alimon: sorry, I now think this test really did timeout. The output is rather confusing :/ Jun 02 23:30:38 I'm starting to wonder if we should really call an ERROR when a setscene task failed, the "real" one runs instead and succeeds so the build actually was successful Jun 02 23:52:15 alejandrohs: silently ignoring it just means people pay no attention when there is likely something that needs fixing Jun 02 23:53:20 alejandrohs: its like ignoring disk bad sector errors as it was a write and it found somewhere else to write the data so it was "fine" Jun 02 23:58:14 yeah I guess youre right, if its something transient however it triggers a false alarm Jun 03 01:07:22 please read an important IRC announcement: https://lists.yoctoproject.org/g/yocto/message/53737 Jun 03 02:23:06 https://www.reddit.com/r/servers/comments/ar3k6y/trying_to_install_windows_to_an_intel_s5520ur/ Jun 03 02:23:14 whoops, wrong chat! Jun 03 02:23:28 was making fun of the UEFI implementation on this board (if interested) ;D **** ENDING LOGGING AT Thu Jun 03 02:59:56 2021