[02:03] yeah... I want to know what happened to the ubuntu touch recovery [02:04] You based it on the worst known recovery and removed all the features, when you click backup/restore it's jokes that it tricked you and it's mot going to do a backup [02:05] this is horrible, whoever is responsible needs to be talked to very sternly [02:05] It's disgusting how cirppled and unusable the system went between vivid and xenial [02:07] Now I have to write something useful and relevant to the ubuntu touch community before we can even think about pushing people to xenial, yeah that's right people are still using vivid because you keep putting out images and source but you don't seem to care about the quality of it [02:08] your system partition binaries have dead symlinks in them, including mount, making it impossible to boot the system because you can't mount the partitions [02:09] worst project management I've seen in a reputable organization like this [02:09] it's something I would expect from some fork of a fork of mint [02:17] why are you here complaining about a dead product? [02:17] dobey, because they messed with what i was working on [02:18] and i have no idea what you're even talking about [02:18] then ask questions that will have a meaningful answer === robert_ancell_ is now known as robert_ancell [02:18] lots of people can't afford nice new phones, they deserve support [02:18] that seems to be some advice you should perhaps follow [02:20] how is going on some tirade about a recovery image and vivid going to solve your problem, if you even actually have one? [02:20] you ask me why i'm complaining about you turning a perfectly working product into one that doesn't work, and has a total lack of functionality [02:20] My problem is that I need to make a recovery that does things aside from looking like unity [02:20] i still have no idea what you're talking about, or why you're saying "you" or why you're doing so in this channel [02:21] You being the ubuntu devs [02:21] if you're not one of them, than you have nothing to worry about, but this is pure idiocy, and whoever made the decision needs to know they messed up [02:22] a lot of people depend on free opensource software and a lot of us put our faith in ubuntu [02:22] you are being rude [02:22] well, whoever made this decision messed up and I wanted them to know it [02:22] that's all [02:22] xorpad, but wasn't all the ubuntu touch stuff forked by the ubports [02:22] ? [02:23] Ubuntu keeps making snapshots and new versions, so not completely [02:23] That's just an automated process [02:23] it sounds like xorpad wants to change the recovery to something other than the ubuntu recovery for some reason, and expects it to behave like it would on an android phone, even though it's not android [02:23] Canonical doesn't support Ubuntu touch any more [02:24] blahdeblah, yeah that's pretty clear from the lack of caring about the builds [02:24] i'm surprised that stuff is still running, really [02:24] I am too [02:24] I thought all that stuff was shut off back in June? [02:24] Drop us a ticket at rt@ubuntu.com and we can probably make it stop. :-) [02:24] even when the phone was supported, some of the channels were automated builds for build testing only, and not to be used on real devices generally [02:25] Okay, whatever... i just won't put any faith in ubuntu anymore [02:25] but then, there seems to be quite a few automated things that were set up for phone/unity stuff, still running, that i would expect to be dead now as well [02:25] welp [02:26] whatever, whoever is making these, tell them I was here [02:26] and what I said [02:26] because decisions like this shouldn't even be considered by someone on a project like Ubuntu [02:26] It's a build server, and I can say on its behalf that it doesn't care about you or me or anything. :-) [02:26] i still don't know what you were actually trying to say [02:27] that the code was great and now it's not functional in so many ways [02:27] nobody has any idea what "decision" you're takling about [02:27] you've been very vague [02:28] the decision to replace part of the firmware that was purpose built and useful it's something that pretty much does nothing [02:28] xorpad, I can pretty much guarantee that it was bitrot that did that [02:28] still lost [02:28] most of the folks (if not all of them) who worked on the Touch project were laid off [02:29] dobey: the initialization and system recovery code broke in the automated builds [02:29] basically, bitrot for the Ubuntu Touch firmware images [02:29] it didn't break, it was replaced by useless code [02:29] ... hence, broken? [02:30] no it does exactly what it is programmed to do [02:30] it's very obvious it was done on purpose [02:30] this wasn't a bug in an automated system [02:30] it was someone telling ubuntu phone and devs that ubuntu doesn't care about their own stuff [02:31] still no idea what you're talking about [02:31] well, then you're not the person i should be talking to and you can ignore me [02:31] I've said what I had to say, I'm done, if you don't get it, fine [02:31] there is absolutely nobody you should be talking to, because there is nobody to talk to [02:32] none of that changes the fact that ubuntu is still not android [02:32] Someone programmed the replacement that doesn't do anything other than joke about how it doesn't work [02:32] and I'm not running android [02:32] I'm running ubuntu [02:32] Simply with a kernel module to handle android drivers so it can run on android devices [02:34] the same recovery image was used on ubuntu phones for roughly 3 years [02:35] And then it was replaced with one that has zero functionality [02:35] "and then" ? [02:35] ubuntu phone has been a dead project for 6 months [02:35] i am pretty certain nobody replaced the recovery image in that time [02:35] well, the official site is still providing snapshots [02:36] so someones still working on it [02:36] nope [02:36] the only people working on ubuntu phone images, are ubport [02:36] ubports [02:36] No, so people who aren't affiliated with ubuntu can post images and snapshots and products on ubuntu.com? [02:36] no [02:37] as was previously stated, some automated thing is apparently still running if daily images are being built there [02:37] then you're not telling the truth, or your not aware of it [02:37] this was not an automated thing [02:37] It was a complete replacement of a product wiht one that literally tells jokes about how it doesn't work [02:37] i seriously doubt that [02:38] well, that [02:38] is why i'm here complaining [02:38] because someone did it [02:39] if it wasn't supposed to happen cannonical should track down who's messing with stuff that goes up on their official ubuntu site [02:39] and what image in what channel are you even talking about? [02:39] can you perhaps NOT be vague in a single accusation you're going to throw out? [02:40] no, i'm done with people who have no idea what's going on [02:40] as you clearly don't [02:40] lol [02:40] i see you trolling [02:40] wtf [02:43] this is horrible, whoever is responsible needs to be talked to very sternly [02:43] highlarious [02:44] I doubt anyone knows anything about ubutouch anymore [02:44] at least anyone working at Canonical [02:44] uhm [02:44] even so [02:45] even the last "devel-proposed" image for mako on system-image.u.c is a year old [02:45] and well, for all the other devices too [02:46] so sounds like someone broke their own junk and then is looking to blame someone else [02:47] and last "stable" channel release was feb 2017 [02:49] and staging was apparently april 22 [02:49] wahtever [02:49] so weird === infinity0_ is now known as infinity0 [06:07] doko: hi, next hwe stack update to xenial needs a fix to binutils so llvm-5.0 doesn't ftbfs on arm64 [06:08] I'll file a bug [06:15] tjaalton, doko, hi, you can find built packages at https://launchpad.net/~ricotz/+archive/ubuntu/mozilla/+packages [06:16] probably doesn't fix powerpc ftbfs? [06:16] I can't test those [06:16] same for s390x [06:17] I'll push this binutils to x-staging [06:18] could you do trusty and zesty as well which will be needed too [06:18] I don't need trusty [06:18] firefox does [06:18] and don't care much about zesty at this point :) [06:19] don't be narrow minded here, those backports are not only for X ;) [06:20] x-staging is [06:20] right, but this whole backport-request isn't [06:20] my tracker bug is #1716203 [06:21] do you have another? [06:21] another bug report, no [06:21] ok [06:21] i'll ask for a s390x builder === maclin1 is now known as maclin [09:45] jbicha: stuck cron job following last weekend's network event; fixed, thanks [11:48] hum, would be nice if somebody was reviewing the artful queue, things have been sitting there since friday mostly :-/ [11:52] Laney, sil2100, slangasek - https://bileto.ubuntu.com/#/ticket/2981 does not have autopkgtest results, how do I request them to be triggered? (i mean all the reverse-depends....) [11:52] i expected for them to be available by now [11:56] xnox: they should have been available nearly immediately; I'm not sure what happened there and don't have time to dig today, unfortunately [12:07] xnox: I need to look into that later on, it's not the first time this happens - I'll try finding some time later today [12:26] xnox: I don't know how that triggering stuff works, think that my results didn't come back last time [12:58] xnox, have you seen bug 1721839 [12:58] bug 1721839 in systemd (Ubuntu) "Services asked for by UDEV do not get triggered" [Critical,New] https://launchpad.net/bugs/1721839 === tkamppeter_ is now known as tkamppeter [13:29] niedbalski: around? [13:47] Anyone here who can help on a systemd issue? Bug 1721839 [13:47] bug 1721839 in systemd (Ubuntu) "Services asked for by UDEV do not get triggered" [Critical,New] https://launchpad.net/bugs/1721839 [14:19] tkamppeter, so if the SYSTEMD_WANTS is correct in udevadm info output, it means you got the udev rules correct. [14:20] tkamppeter, next you need to figure out the .device unit name in systemd which matches that particular udev device and get the status of it [14:20] tkamppeter, it maybe that the .device, from systemd point of view, is not "started" and therefore said .device units' .wants are not started [14:21] tkamppeter, you can use something like $ systemctl list-units '*.device' | grep pci [14:22] to find your printer, and check the state of that unit. is it active & plugged? [14:24] possibly a device unit is configured for one of the other udevadm entries, the one that was not tagged systemd? [14:27] xnox, the printer was plugged and turned on at that time, as it appeared in the udevadm output. [14:28] xnox, I can also start the service manually with "sudo systemctl start XXX.service". [14:29] tkamppeter, i am not asking about the physical real world state. [14:29] tkamppeter, i don't want you to start services by hand [14:29] tkamppeter, i want to see the status of the systemd .device unit corresponding to the udevadm device [14:29] the two are separate [14:30] xnox, starting services by hand was only a test to find out in which stage it fails. [14:30] and confusingly the systemd .device unit state can be "unplugged offline" or "plugged online" and that matters whether wanted corresponding .service files are started or not [14:30] tkamppeter, what is the output and states of all of your systemd .device units? [14:31] $ systemctl list-units '*.device' | pastebinit [14:33] xnox, "systemctl list-units" has in its output: [14:33] ● udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service loaded failed failed Automatic USB/Bluetooth printer setup (-devices-pci0000:00-0000:00:14.0-usb2-2x2d2) [14:37] xnox, http://paste.ubuntu.com/25707724/ [14:38] cool (that output got trimmed, but it looks ok) [14:39] tkamppeter, now you need to figure out why that service failed, as it seems like it was attempted to be started (impossible to know now if automatic, or by you manually, as one needs to reset-failed to get back to clean state after the printer is unplugged) [14:39] tkamppeter, what is $ systemctl status udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service ? [14:39] or e.g. journalctl -u udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service [14:46] tkamppeter, note that the stuff after @ seems like an incomplete device unit, eg. it doesn't start with sys- (not sure if you need that or not) [14:47] and do note there are %i and %I substitution variables available which can pass the instance name verbantim or systemd-UN-escaped [15:06] xnox, What I did today is only plugging and turning on the printer, I did not try to start the service manually. On the other day when I started it manually it was correctly working after the manual start. [15:09] The udev-configure-printer UDEV rules file and systemd .service files are the same as in 17.04 and there they worked. [15:09] xnox ^^ [15:09] xnox, so does %i mean verbatim and %I systemmd-UN-escaped? Or vice-versa? [15:13] xnox, here is a complete list-units paste: http://paste.ubuntu.com/25707920/ [15:14] xnox, the problem is that automatic start does not work whereas manual start works. [15:14] All data which I pasted for you today is without doing any manual start attempt. [15:18] xnox, systemctl status udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service --> http://paste.ubuntu.com/25707960/ [15:21] xnox, journalctl -u udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service --> http://paste.ubuntu.com/25707972/ [15:41] Laney, is there some manual way to run britney? for it to list what tests/packages it would test if it were asked to migrate libseccomp? [15:41] such that I would be able to run those tests manually [15:42] i guess i can by-hand look at the reverse-deps [15:56] xnox, did you get my pastes? [15:56] tkamppeter, yes, and then you dropped out of irc. [15:56] OK, I am back now. [15:56] tkamppeter, it looks like the .service unit was started, did run, and did try to accells /sys/devices.... [15:57] tkamppeter, but that - for whatever reason - has failed. [15:57] maybe the printer/device was not yet "ready enough| [15:57] maybe the printer/device was not yet "ready enough" [15:57] thus you need to somehow debug that unit, to see why it is failing when systemd starts it, at the time it does. [16:17] xnox: not that I know of [16:25] xnox, I have found the cause of the problem now: The directory which needs to get accessed by udev-configure-printer is: /sys/devices/pci0000:00/0000:00:14.0/usb2/2-2, note that the second last character is a dash. UDEV/systemd(?) inserts for the %i in the second line of the udev-configure-printer@.service file: -devices-pci0000:00-0000:00:14.0-usb2-2x2d2, replacing each slash by a dash and the original dash by "x2d" (should it not be "\x2d"? Not su [16:25] re what gets inserted for the %I in the ExecStart line, but udev-configure-printer ends up with a /sys/devices/pci0000:00/0000:00:14.0/usb2/2x2d2 which it does not find as the "x2d" is not turned back into a dash. [16:26] tkamppeter, sigh [16:26] let me see if there is an option to unbreak that [16:26] xnox, so the problem is the escaping, either UDEV or systemd do it wrong and it is a regression as in 17.04 it worked. [16:27] tkamppeter, it might be a bug-fix actually [16:27] xnox, this will probably break all UDEV/systemd interaction with all USB devices, probably even with all PCI hardware. [16:28] xnox: reverse deps and reverse testsuite-triggers [16:28] and the package itself [16:33] tkamppeter, i believe when systemd-escape is called it should be called with "-p --template=...." options [16:34] tkamppeter, '-p' makes the mangling to be path like, but let me test things here. [16:34] tkamppeter, can you paste me the udev-configure-printer@.service unit? [16:37] xnox, here we go: http://paste.ubuntu.com/25708406/ [16:38] tkamppeter, that unit file is invalid [16:38] tkamppeter, it is not a shell [16:38] tkamppeter, you cannot execute multiple commands with ';' or use shell builtins [16:38] Here is the UDEV rules file /lib/udev/rules.d/70-printers.rules: http://paste.ubuntu.com/25708425/ [16:38] tkamppeter, why do you have sleep 1 there? why is not a separate line? [16:39] tkamppeter, you really should not have arbitrary sleeps =/ [16:39] but that's besides the point [16:39] xnox, this was to try out whether it was really due to the printer not ready. [16:39] xnox, the original does not have the sleep 1 [16:39] tkamppeter, note that in systemd ExecStart is not executed by shell, but by direct exec [16:39] you should have had something like [16:40] ExecStart=/bin/sleep 1 [16:40] ExecStart=/lib/udev/udev-configure-pirnter add %I [16:40] but that's for future reference [16:40] maybe even ExecStartPre=/bin/sleep 1 [16:42] xnox, here is the original .service file: http://paste.ubuntu.com/25708443/ [16:43] xnox, this is the one with which I observed the escaping problem. [16:44] \xnox, and as we have observed the escaping problem now, we have a possible cause, the thing of printer-not-ready was an assumption before discovering the escaping problem. [16:44] xnox ^^ [16:50] tkamppeter, yeah sure. [16:54] xnox, if I run the systemd-escape commad on the command line by hand, the x2d replacing the dash at the end of the path is actually a \x2d, but it seems that in the further processing the \ gets lost and the unescaping breaks [16:55] tkamppeter_, i'm struggling to create a full dance of udev rule et.al. === tkamppeter_ is now known as tkamppeter [16:58] tkamppeter_, could you please try changing the udev rule to call `systemd-escape -p --template=udev-configure-printer@.service %p` ? [16:59] tkamppeter, also i wonder if we need to quote the SYSTEMD_WANTS="'%c'" rather than just ="%c" [16:59] but it does smell like a regression [17:03] xnox, adding -p I already tried without success. [17:04] tkamppeter, sigh [17:04] xnox, am I right that after changing a UDEV rule file it is enough to get the change being used if I run "sudo systemctl restart udev"? [17:06] no [17:06] tkamppeter, $ udevadm control --reload [17:07] and one needs to reset the failed unit state as well [17:07] and then unplug/replug printer [17:09] How do I reset the failed unit state? [17:10] xnox ^^ [17:10] tkamppeter, read manpages?! =) $ systemctl reset-failed myunit.service [17:11] https://www.freedesktop.org/software/systemd/man/systemctl.html#reset-failed%20%5BPATTERN%E2%80%A6%5D [17:20] xnox, tried the reset-failed, tried to restart udev and systemd and nothing happens, no new entry in journalctl -u, last entry is from an hour ago: [17:20] Oct 09 13:12:10 till-x1carbon systemd[1]: udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2x2d2.service: Unit configuration has fatal error, unit will not be started. [17:22] that is odd, unplugging printer, reseting the unit, plugging printer in, should trigger the unit again [17:22] "fatal error" is that unit still with bad =sleep 1? [17:23] I have removed the sleep 1 and after that restarted systemd and udev and also replugged the printer. [17:23] xnox ^^ [17:47] xnox, udevadm monitor --environment gives now [17:47] SYSTEMD_WANTS='udev-configure-printer@-devices-pci0000:00-0000:00:14.0-usb2-2\x2d2.service' printer.target [17:48] xnox, so here the backslash is in. [17:49] tkamppeter, that's good, but paths should really be encoded using -p option, such that this unit name is @devices.... rather than @-devices [17:50] tkamppeter, as in systemd leading slash is discarded, unless it's the only thing hence, for example mount units for / /var /var/lib are encoded as -.mount var.mount var-lib.mount [17:51] tkamppeter, but it does smell like there maybe parsing regression in udevd and/or systemd codes [17:51] xnox, adding -p ... [17:51] tkamppeter, is there a way for me to have a fake usb printer in udevadm? [17:51] tkamppeter, can i fake add usb printer to like a VM? [17:53] xnox, now I get SYSTEMD_WANTS='udev-configure-printer@devices-pci0000:00-0000:00:14.0-usb2-2\x2d2.service' printer.target [17:57] xnox, journalctl -u 'udev-configure-printer@devices-pci0000:00-0000:00:14.0-usb2-2\x2d2.service' gives: [17:57] -- Logs begin at Mon 2017-10-09 09:07:44 -03, end at Mon 2017-10-09 14:52:12 -03. -- [17:57] -- No entries -- [17:58] xnox, and systemctl status 'udev-configure-printer@devices-pci0000:00-0000:00:14.0-usb2-2\x2d2.service': [17:59] ● udev-configure-printer@devices-pci0000:00-0000:00:14.0-usb2-2\x2d2.service - Automatic USB/Bluetooth printer setup (devices-pci0000:00-0000:00:14.0-usb2-2\x2d2) [17:59] Loaded: loaded (/lib/systemd/system/udev-configure-printer@.service; static; vendor preset: enabled) [17:59] Active: inactive (dead) [17:59] tkamppeter, inactive (dead) sounds good, no? it ran and quit? [17:59] or never ran..... [18:00] but note that the decoding is bad anyway [18:00] the subject should be %i and it's not good [18:00] actually not sure [18:01] tkamppeter, i do not have any time today to look into this properly, i will come back to the back report sometime this week. [18:01] tkamppeter, i have a few more urgent systemd things to look into [18:03] xnox, can you arrange it before Final Freeze on Thursday, so that Artful will have this fixed? [18:03] tkamppeter, no, i cannot commit to that. [18:04] tkamppeter, if you want to escalate, you can tag the bug rls-aa-incoming for core/foundations management to prioritise and assign it. [18:04] tkamppeter, at the moment we are focusing on critical bugs only, that are preventing first-boot, install, reboot. [18:05] Does it mean I set the tag "rls-aa-incoming"? Have I also set another tag and/or subscribe the bug to someone? [18:07] xnox ^^ [18:08] tkamppeter, rls-*-incoming is away for canonical, to escalate issues cross-teams at canonical. Bugs tageed with such a tag will end up on http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-aa-incoming-bug-tasks.html [18:09] tkamppeter, which all teams review during their sprint-board loading and commit to fixing for a timeframe [18:09] tkamppeter, this has been going on.... since like forever, at least since precise? if not before that [18:09] tkamppeter, there are such reports for every realease [18:09] .... [18:10] Added the tag now. [18:11] xnox, otherwise we have to make it an early-term (0-day, 1st week, 1st month, ...) SRU. [18:12] xnox, would be bad if the 17.10 will not have USB printer auto setup for all its life. [18:13] xnox, should we tell in the Release Notes that 17.10 is without automatic USB printer setup and that it is worked on it for an early SRU? [18:14] tkamppeter, do edit the release notes, the known issues section, with a link to this bug. if we fix it before release we can remove that note [18:24] xnox, Release Notes edited. [18:25] tkamppeter, thanks =/ and sorry [18:27] xnox, I could ask willcooke whether you could expense a USB printer so that you can test for fixing it as SRU shortly after release. WDYT? [18:27] tkamppeter, ha [18:27] tkamppeter, no way =) i think it should be doable to reproduce with "ls@.service" and udev rules that does SYSTEMD_WANTS for any device by calling systemd-escape --template=ls@.service %p [18:28] tkamppeter, since one should be able to ls all the dev paths [18:28] tkamppeter, and then i will need to plug in as many usb sticks as i can find [18:28] =) [18:29] xnox, OK, then we have a way to do it. Take some time after Final Freeze and let us get an early SRU. === lamont` is now known as lamont === JanC_ is now known as JanC === ahasenack is now known as andreas