[06:25] good morning [08:15] mo'in [08:15] are we there yet? [09:08] hi - can someone please tell me what happens to the app data in /var/lib/apps// if i update the package with snappy to a newer version? will the directory be renamed from e.g. 0.1 to 0.2 or will all app data get lost? [09:39] Good morning all; happy Monday, and happy No Dirty Dishes Day! 😃 [09:40] Good morning everyone [09:40] any chance someone can review snap 2662? [09:51] D_Cent: hi [09:52] D_Cent: the services in the snap are stopped, the data is copied to the new version, the new version is made current, and its services started [10:05] JamesTait: http://i.imgur.com/E1trXyD.gif [10:05] JamesTait: and good morning :) [10:06] Chipaca, that's mesmerising. [10:06] Chipaca, and good morning, too. ☺ [10:06] JamesTait: i too could watch people work for hours [10:07] Chipaca, I see now that I've been washing the dishes all wrong. [10:09] * ogra_ would like to see the other end of the machine ... and the guy who juggles them til they are dry [10:13] ogra_: here you go: http://www.awardsdaily.com/wp-content/uploads/2012/10/cloud.jpg [10:14] lol [10:48] “the UI for Ubuntu Core is provided via web apps written for webdm” [10:48] that's interesting [10:48] not connected to any reality i'm aware of, but interesting nonetheless [10:49] well, if snappy config is hooked into webdm thats not completely untrue [10:50] darn, and i just sent an email saying it was [10:50] untrue, that is [10:51] well, it isnt like he imagines ... so thats fine [10:51] feel free to expand on my rather curt email :) [11:23] * ogra_ glares at http://127.0.0.1:4200/snap/8nzc1x4iim2xj1g2ul64.chipaca/ in his webdm [11:23] creative naming :) [11:23] yeah, let's call it that [11:23] ogra_: hopefully the description makes it all clear [11:24] hmm, seems my webdm instance doesnt show any descriptions at all ... now that you mention it ... [11:25] the details page just has the icon [11:26] ogra_: ah, i guess it's the title, not the description [11:26] ogra_: but: https://uappexplorer.com/app/8nzc1x4iim2xj1g2ul64.chipaca [11:26] ogra_: the description is also enlightening [11:28] heh [11:28] Chipaca: thanks for your answer previously :) [11:29] D_Cent: that's what we do. ogra_ with his good looks, me with my on-point answers. [11:30] * Chipaca goes find lunch [11:33] lunch? are you in Europe? [11:34] lool: that depends on who you ask [11:34] lool: i'm in the uk [11:35] oh ok [11:35] lool: have been for >4 years now [11:36] Chipaca: for some reason I associate you with a later timezone :-) [11:37] lool: for my first 3+ years in canonical, i was :) [11:58] does anyone know how I can use a wifi dongle with snappy? i.e. connected to a usb port on a board? [12:05] rickspencer3: there is a hacky way documented here: http://www.marinus.nu/2015/02/enabling-wifi-on-snappy-ubuntu-core.html [12:05] we dont have a neat net management framework yet [12:08] actually wpasupplicant should be in the rootfs now [12:08] it is [12:08] so you just need to bottom part [12:08] yeah so you dont need to install anything :) [12:09] we need to snap network-manager and/or connman etc. [12:20] lool, how hard would it be for someone to add a configuration page to webdm for wpasupplicant? [12:20] seems like it could go into the store [12:20] the user plugs their board into the local wired network [12:21] navigates to ubuntu@webdm.local:nnnn [12:21] sets up wifi [12:21] oops [12:21] rickspencer3, it probably couldn't easily go into the store, as it would require special access (to change the wifi config) [12:21] it's not the target user experience [12:21] forgot the step of granting permissions [12:21] but that would work for now [12:21] beuno, could it not go into the store, but the user has to grant permissions to the usb? [12:21] so I think it would be rolled into webdm, that already has special access [12:21] this seems to be a fine ubuntu-core config [12:22] lool, agreed, but I was thinking shorter term [12:22] and then we'd miss a generic way to change configs in webdm [12:22] rickspencer3, that's an interesting question. I'm not sure where the boundaries are there [12:22] as an ubuntu-core config, it would be suitable from command-line as well [12:22] on the phone, we fulfilled many use cases by adding apps to the store while in parallel working on the core functionality [12:22] * beuno nods [12:23] we could likely just allow an exception for that app until we do [12:31] BTW edison and other boards just connect to the nearest open wifi on boot [12:31] arguably insane, but convenient === alex-abreu is now known as alex-abreu|off [12:34] lool, that is insane :) [12:34] lol [12:34] lool, I think it would be great if you could do all of these: [12:34] 1. drop a configuration file somewhere ... gadget snap? [12:35] 2. use the snappy cli to define the ap and password [12:35] 3. use a configuration gui via webdm (or standalone in the short term) [12:35] rickspencer3: you can make a snap that has a config handler that takes a nice yaml with ap and password and then goes off, creates the wpasupplicant.conf, starts it and uses dhclient [12:36] wedm will later surface those config yamls [12:36] well this could be in ubuntu-core for now [12:36] first raw, later with generic input form by yaml scheme, even later through making webdm UI extensible [12:38] what about iot devices that have only wifi? You need them to start in access point mode if they can't connect to a known network. (like the chromecast does) [12:38] lool: it could, but i kind off feel hesistant to drop a super quick hack into core [12:38] we dont know what we want to do on network management in core, so starting with a framework/app is a good idea for such quick things imo [12:40] alecu: yes, thats a pretty standard first boot/setup practice we would love to see :) [12:41] alecu: it's been on our TODO for ages; just never getting a change to implement this [12:41] we had good luck in the phone making certain functions apps first [12:41] and them bringing them into the core of the system as they were ready [12:41] for example, the app updater was an app for a while while we were perfecting system settings [12:41] then we moved that code into system settings when both were ready [12:42] asac: hi, what do you mean when you say make the webdm UI extensible? [12:42] beowulf: webdm exposing config of installed apps [12:42] afaict :) [12:43] right [12:43] just vague ideas [12:43] Chipaca: aye, i understand that part, and i understand (i think) the next part which is making yaml into form controls (right?), i'm just not sure what the extensible part was about? [12:44] that it woudl be cool if something could extend the webdm ui with its own config forms [12:44] beowulf: i read that as management handwaving [12:44] fwiw, that's kind of in reverse order of difficulty :) [12:44] but how that could be done is not really understood... we know how forms with schema would work for sure [12:44] beowulf: i.e. [ engineering magic happens here ] [12:46] Chipaca: i have a branch with a code editor embedded in the settings tab, so for a very basic first step if you wanted to send me yaml i could send it back edited... [12:46] *waves hands* [12:46] * Chipaca stuffs yaml in a box and sends it off to beowulf [12:46] hehe [12:47] Chipaca: what is the app config format or database that webdm will be touching, and how will apps find out about changes in those configs? dconf or something like that? [12:48] alecu, https://developer.ubuntu.com/en/snappy/guides/config-command/ [12:48] alecu: snappy config [12:48] alecu: ebdic-encoded hdf files [12:49] Chipaca: lol [12:49] ogra_: asac: thanks! [12:49] boo :) [12:49] good morning [12:49] sergiusens: good morning! [12:49] sergiusens: long time no see :) [12:50] I vanished to deal with a full week of bureaucracy [12:50] Chipaca: it's been a while ;-) [12:50] sergiusens: wat? paperwork on your holidays?!? [12:50] sergiusens: that's probably against geneva, or something [12:51] * Chipaca does not like paperwork [12:51] Chipaca: yeah, well, it had to be done at some point [12:51] * sergiusens dislikes paperwork too [13:28] morning [13:28] sergiusens: spent the week paying taxes? [13:29] rsalveti: yeah, delightful [13:29] haha, yeah, had to do that a few weeks ago === erkules_ is now known as erkules === vrruiz is now known as rvr [13:54] hmm, trying to re-build one of my old snaps fails :/ [13:55] http://paste.ubuntu.com/11206759/ [13:55] - snappy-systemd_hook_unknown_key_snap-inspector-term.snappy-systemd [13:55] Unknown field 'ports' [14:00] ogra_: I think the ports moved around https://developer.ubuntu.com/en/snappy/guides/packaging-format-apps/ [14:01] but old location ought to work I guess [14:01] lool, well, my package.yaml looks pretty much the same [14:01] http://paste.ubuntu.com/11206839/ [14:02] i dont have any internal port defined though [14:05] * ogra_ doesnt see any difference beyond this vs the example package.yaml there [14:15] lool: ogra_ ports in that page is outdated, I logged a bug weeks ago [14:15] lool: ogra_ oh, ports in there is fine now :-) [14:22] err, wut ? [14:22] is it outdated or is it fine ? :) [14:36] jdstrand, any idea why i would get "Unknown field 'ports'" with http://paste.ubuntu.com/11207212/ ? [14:38] ogra_: the tools need to be updated for ports [14:38] jdstrand, yay ... [14:38] ogra_: feel free to submit it and ping me [14:38] * ogra_ wasted the last hour on searching the issue in his own stuff :) [14:39] ogra_: :( [14:39] no, i'm fine ... i could have pinged you earlier :) [14:40] ogra_: so long as it follows docs/meta.md, you should be fine [14:40] where is that ? in the reviewers tools ? [14:44] ogra_: in lp:snappy [14:44] k [15:00] Chipaca: can I have regrets? [15:00] sergiusens: NO RUGRATS [15:00] Chipaca: I have a future of that coming soon.. rugrats :-P [15:01] sergiusens: :D [15:01] sergiusens: regrets for what? [15:04] sergiusens: were these regrets wrt the branches? logger? working for snappy? :-p [15:05] Chipaca: logger, let's stay on the standup if you wnt for 3 minutes [15:06] sergiusens: sure === vrruiz_ is now known as rvr [15:16] sergiusens: hi! with kyrofa we started building an experimental "snappy scope", based on webdm, as you suggested [15:17] alecu: and... :-) [15:17] sergiusens: there are a few questions we'd like to ask you [15:17] alecu: shoot [15:17] sergiusens: hmm... would you have some time for a hangout? [15:18] alecu: sure, 14hr ART? [15:18] sergiusens: sounds good, I'll set up a calendar event. [15:18] thanks! [15:20] jdstrand: is there a "apparmor linter"? [15:20] there is [15:28] Chipaca: you mean to see if the resulting profile is ok? there are two things that can pe used: apparmor_parser -p /path/to/profile and appamror_parser -QTK /path/to/profile [15:29] Chipaca: the first is just a preprocessor and won't do things like verify apparmor variables are correct, etc. the 2nd is a full compile and the equivalent of apparmor_parser -r /path/to/profile except without loading into the kernel [15:29] both can be run as non-root [15:32] jdstrand: i ask because i'd like to be able to do something when apparmor fails, to better inform the developer of what exactly failed [15:32] however, before one gets too excited, the way we are using templates and caps means that we can't naively run the parser on the files in the snap, because a transformation happens on what is in the snap to what is on disk in /var/lib/apparmor/profiles [15:32] jdstrand: right now we're not being too helpful, making the thing more opaque than it needs [15:33] Chipaca: if apparmor fails, you can run apparmor_parser -QTK /path/to/each/profile/from/the/snap and report back [15:33] jdstrand: excellent [15:33] * Chipaca adds a card [15:33] because if apparmor is failing, the files should have underwent said transformation and be in /var/lib/apparmor/profiles [15:34] jdstrand: https://trello.com/c/pK0vSzNZ/380-improve-information-provided-when-apparmor-fails [15:34] if they aren't, then you happen to know that click-apparmor failed to do the transform (which should only happen when using security-override). however, I wouldn't worry to much about that since click-apparmor will be going away [15:35] cool [15:49] hah! i love it [15:49] we have ErrInvalidPackageYaml [15:49] ... we don't use it :) [16:09] hm… how do I find out which beaglebone black revision I have? [16:15] waaaiiit [16:15] https://code.launchpad.net/snappy/+activereviews [16:15] git branches started showing up there [16:38] Chipaca: let's move to git [16:39] sergiusens: i believe we have a couple of blockers still? [16:39] otherwise, +0 from me :) [16:41] Chipaca: just auto merge and the recipe, the recipe we can run on our end in a new tarmac implementation so we have one dput per commit [16:43] Chipaca: and for the merge we can use launchpadlib to poll, until we have webhooks, and git merge [16:44] sergiusens: my +0 stands. With it I mean "don't let me stop you". And by *that* I mean I'm not going to help you do it :) [16:56] sergiusens, rsalveti, webdm is completely unusable on the released kvm image (works fine after snappy update webdm), do we want to keep it like that ? [16:57] or do we plan to make point releases [16:59] sergiusens, any idea why webdm on a kvm instance shows me all the arm kernel snaps ? i see beagle, overo, panda etc (merked as "not installable" but imho we should just hide uninstallable stuff) [16:59] *marked [17:00] (or is that a beuno question) [17:00] the store does what its told [17:00] ogra_: we decided to do that with beuno [17:00] beowulf: [17:01] so if its showing armhf, you are telling it you are armhf (or not telling it your arch at all) [17:01] i'm not telling anything, i just started a kvm instance with the released image :) [17:02] (and up to date webdm) [17:03] ogra_: somewhere along the way *cough* sergiusens *cough* i was told to show but not allow to install oem packages [17:03] if that's wrong it can be changed easily [17:03] sergiusens: hangout? [17:03] alecu: in abit [17:03] alecu: running late in prev meet [17:04] beowulf, well, not sure who can make such a decision ... seems useless to show stuff to the user that he cant install [17:04] sergiusens: no prob [17:04] beowulf: ogra_ oh, right sabfdl said we should see them ;-) [17:04] do you knwo why ? [17:04] ogra_: i agree that the current ui is a bit awkward in this repect [17:05] well, if it is a sabdflified thing ... [17:06] ogra_: we need another point release/pre built image [17:15] yeah [17:26] lool, I installed the new raspberry pi 2 image, but ssh ubuntu@webdm.local is not responding? [17:26] does this imply that I screwed up makng the sd card? [17:27] did you try the IP instead ? [17:27] ogra_, I did not [17:27] webdm.local worked on the last image and on my beaglebone over the weekend [17:29] ogra_, does it matter that I did not format the sd card first or anything? [17:29] I assumed dd would just blast over everything [17:29] no, since you write to it with dd partitioning doesnt matter [17:30] I'll try again after lunch [17:30] but if you run multiple devices with avahi service an the same name (webdm.local) on the same network, i'm not sure what happens [17:30] make sure my bbb still works ;) [17:30] ogra_, I only run one at a time [17:30] k [17:30] well, then it should theoretically work :) [17:31] * rickspencer3 looks forward to his bbb case coming today [17:31] cases ... so overrated [17:31] :) [17:32] bbb worked just fine :( [17:32] I will try reflashing the new pi2 image later [18:45] nothal_: help === vrruiz is now known as rvr [19:00] rickspencer3: I think I dont have webdm installed in the image [19:00] it was broken back then, but now that it's fixed I should totally include it [19:07] lool, ok, so I have to just find the ip and install it? [19:13] rickspencer3, yes [19:14] make sure the date is correct [19:14] then just sudo snappy install webdm [19:15] beuno, will nmap find it on my lan? [19:15] rickspencer3: is avahi failing for you? [19:16] sergiusens, it looks like webdm is not installed on the pi2 image default [19:16] and I don't know how to find the pi2 on my network [19:16] and I lost the admin/password for my router :/ [19:16] rickspencer3, sure it will, otherwise just peak at your rou... ah :) [19:17] rickspencer3: oh, bummer; nmap should find por 22 on the host and mostly nothing else [19:17] rickspencer3: nmap can also enumerate OS (forgot the switch for that), but maybe the arch is part of that enumeration and it should be easy to find with that [19:18] rickspencer3, otherwise, .ssh/known_hosts likely rememebers the last IP it accessed [19:18] and that likely won't have changed [19:19] beuno, oh, I have to blow that away all the time because of switching between the bbb and the pi2 [19:19] they don't play nicely together :) [19:20] ah [19:20] so there's an interesting to think about [19:20] more than one snappy on the network! [19:22] \o/ found it [19:23] beuno, indeed :) [19:23] right now, it's good for Internet of Thing [19:23] Cooperative Multitasking [19:24] it's the floppy-swapping of the new millenium [19:25] sergiusens, that sounds smart, I just tried ssh ubuntu@ipaddresses starting at the highest and working my way down until I found it [19:28] rickspencer3: the trick I sometimes use (outside of checking DHCP logs) is to ping broadcast then look at the arp table [19:28] arp -an [19:28] ping -b 192.168.0.255 [19:28] well, I just installed webdm, so this should work now :) [19:29] oops [19:29] this should work now, right/ [19:29] ssh ubuntu@webdm.local [19:30] yes it should [19:31] rats [19:32] is there a difference betwee webdm and webDM ? [19:32] do I need to start webdm? [19:33] ug [19:33] I ran snappy update and it looks like it is updating webdm, even though I just installed webdm :( [19:34] rickspencer3, there is a network scanner app in the phone store ;) [19:34] that is... hard to pull off [19:34] as in, the store only knows about one version of an app [19:34] (to find your RPi2 IP) [19:34] so you couldn't really install an older version even if you wanted [19:34] unless sergiusens uploaded a new one as you were typing! [19:34] ogra_, nice idea :) [19:35] beuno, no, it updated to the same version [19:35] ew [19:35] I was on 0.6.1, it updated me to 0.6.1 [19:35] I don't know why it would update me, though [19:35] * rickspencer3 tries a restart === soee_ is now known as soee [19:40] beuno: rickspencer3 no I didn't update and it shouldn't update... :-/ [19:41] it also doesn't seem to be working [19:41] maybe webdm is busted on the pi2 image? [19:49] yeah, every time I run sudo snappy update, it reinstalled webdm, which then does not seem to work [19:49] the pi2 image is a bit busted, I guess :) [19:52] rickspencer3: where did you get your pi2 image from? [19:52] rickspencer3: maybe give me the output of journalctl -u webdm_snappyd_0.6.1.service [19:52] rickspencer3: and also, snappy info, snappy list and snappy list --updates [19:52] rickspencer3: try this, snappy remove webdm & snappy install webdm (sudo) [19:53] rickspencer3: I think I know what this is and it is quite unfortunate [19:54] sergiusens, anything I can do to fix it? [19:54] I have a bbb so I am not dead in the water [19:54] just wanted to try out the pi2 [19:54] also, I don't have a case for my bbb yet :) [19:54] should come today [19:54] rickspencer3: snappy remove webdm snappy install webdm may do the trick [19:57] oh [19:58] sergiusens, can I get you any debug info or anything? [19:59] rickspencer3: I'm not sure this went through as I was swapping networks so I'll copy paste just in case [19:59] k [19:59] 16:49 < sergiusens> rickspencer3: where did you get your pi2 image from? [19:59] 16:50 < sergiusens> rickspencer3: maybe give me the output of journalctl -u webdm_snappyd_0.6.1.service [19:59] 16:50 < sergiusens> rickspencer3: and also, snappy info, snappy list and snappy list --updates [19:59] and also please pastebin the result of find /apps [19:59] snappy install pastebinit.mvo ;-) [20:00] sergiusens, I got the pi2 image from the where ubuntu.com/snappy/start said to go [20:22] rickspencer3: I'll get the image and try after I get back from my excersise routine then [20:23] kk [20:23] sorry, I got distracted with something else [20:24] rickspencer3: no worries, so did I [20:24] :-) [20:24] hehe [20:54] \o/ my bbb case came in the mail [21:29] rickspencer3: welcome to the prestigious club of the non-naked bbbs! [21:29] rickspencer3: your membership invoice is in the mail [21:37] mterry: https://bitbucket.org/fzzbt/go-gedit-plugin/ is the "official" gedit plugin (in the sense that it's linked from https://code.google.com/p/go-wiki/wiki/IDEsAndTextEditorPlugins ) [21:41] Chipaca, ah interesting [21:55] sergiusens: i lied about committing something for you to look at wrt refactor tonight, i fear