[00:00] kyrofa: look at the wiki link above [00:00] kyrofa: that's the accurate state of the packaging support [00:00] kyrofa: I think only ubuntu and debian are up-to-date and maintained [00:00] kyrofa: everyting else is not [00:06] zyga, that works, thanks [00:45] Bug #1660865 opened: [interface] hardware-observe should allow broader access to /proc/bus/pci [01:41] odysseywestra anyone get around to you? [01:41] sorry it took so long to get back [01:43] I took a look at your project, I think it will pretty easy to snap. Is there a step you are confused about? [01:55] Hi I would like to ask if I can install avahi-daemon on Ubuntu Core. [01:56] On Ubuntu Desktop, I can install by "sudo apt install avahi-daemon". [01:57] On Ubuntu Core, "snap find avahi" returned 0 snaps. [01:57] So, I installed the classic snap. [01:58] On the classic prompt, I tried to install avahi-daemon with "sudo apt install avahi-daemon", but it seems it failed. [01:58] I saw the following message at the end of installation. [01:58] invoke-rc.d: policy-rc.d denied execution of force-reload. invoke-rc.d: policy-rc.d denied execution of start. [01:59] So, it seems it failed to install avahi-daemon. [01:59] Does anybody know why I can't install this package on classic snap? [02:53] HumbleBeaver: Basically, how to put it together really. Maybe it would help If I could see what other have done? [03:01] Does anybody know how to install avahi-daemon on Ubuntu Core? [03:07] odysseywestra, thats basically how I got started, [03:07] the snapcraft.yaml file is stupid easy once you get over the initial "now what" about it. [03:10] odysseywestra http://paste.ubuntu.com/23903144/ <--this is one I got help with here today, by mhall119 and jstrand [03:12] odysseywestra http://paste.ubuntu.com/23903151/ <--heres one I'm still messing around with its probably closer to what you need. [03:12] Note the source is a git repository. [03:16] okay. I'll take a look at those, and go from there. I'll show you what i have in a bit. Thanks for giving me a starting point. [03:19] Bug #1660879 opened: snap refresh with more than one argument produces poor error message [04:02] ping [04:02] Anybody used avahi on Ubuntu Core? [05:43] does anyone know if there is a bug that prevents the classic flag from functioning in 'try' mode. I'm using version 2.20.1ubuntu1 on 16.04 and it seems to ignore the classic flag [05:45] same on 2.21 [06:18] PR snapcraft#1097 opened: lifecycle: print the command needed to clean the dirty part [07:45] olympionex: yex, it was already fixed in master AFAIK [07:46] olympionex: I think that you can switch to edge (snap refresh --edge core) to use nightly builds on your machine (even on classic :-) [07:46] olympionex: (classic distro) [07:46] olympionex: and this bug, I believe, is gone there [07:46] olympionex: for more on this please ask chipaca [07:46] olympionex: he should be around in a few hours [07:55] zyga: thanks very much! [08:55] PR snapd#2724 closed: overlord,tests: have enable/disable affect security profiles [08:58] PR snapd#2754 opened: tests: improve debug when the core transition test hangs [09:33] PR snapd#2727 closed: overlord/ifacestate: register all security backends with the repository [09:33] PR snapd#2755 opened: interfaces: port the mount backend to new APIs while retaining snippets [09:43] Do we have support for ubuntu 16.04 on dell gateway 5000 [09:43] can I upgrade from 15.04 to 16 ubuntu core [09:54] PR snapcraft#1094 closed: core: switch to using rpath for clasic confinement [09:57] PR snapcraft#1092 closed: meta: properly get the icon extension from splitted name [10:00] PR snapcraft#1086 closed: print snapcraft's version on startup when running with --debug [10:21] is it possible with snap to compile different versions of binaries to have an optimized version for different CPU generations? [10:22] timp: I don't know if the automatic snapcraft building can do that, but you can certainly hand-roll a snap to meet those requirements. Sounds like a lot of work for ? how much reward, though! [10:23] would I then include different binaries inside a single snap? Or can different snap packages be selected depending on the CPU? (not just the architecture, but also the generation) [10:24] I'm asking for a project that is meant for scientific computations. Doing eigen analysis and such. The reward there can be big, because the computations for full datasets can take hours [10:24] I'm not sure of the latter option, but you could certainly pack multiple binaries in a single snap and have a wrapper script pick the correct one [10:26] I don't know how easy it would be to automate, though, as I don't know if confinement would give you access to /proc/cpuinfo (or similar) to autodetect [10:27] timp: you can do that inside your snap [10:27] right. Sounds like a bit of a hassle though. [10:28] timp: one binary that uses appropriate function based on the cpu [10:28] timp: this is a gcc feature [10:28] timp: and it should be pretty easy to use [10:28] ah, that sounds cool. [10:28] I got the question here http://community.mrtrix.org/t/provide-mrtrix3-as-a-snap-package-for-linux/687/2 [10:28] timp: so that one binary will work good everywhere [10:28] timp: https://gcc.gnu.org/wiki/FunctionMultiVersioning [10:28] timp: try that, good luck :) [10:28] in particular they want to use the -march=native flag. [10:29] thanks for the info :) [10:29] timp: well, that particular flag is pretty pointless [10:29] timp: as the resulting snap will be optimized for whoever builds it [10:30] timp: for binary packages that's the totally wrong way to go [10:30] right, that's why they provide sources now instead of binary packages [10:30] timp: I think that with this gcc feature you can deliver performant binaries anywhere [10:32] I will propose that to them. [10:33] I would say that is good enough for binaries. And the source will anyway be available for people who want to compile it themselves. [10:39] PR snapcraft#1096 closed: schema,copy plugin: better errors when item has no value [10:52] Hi folks ! [10:52] Does anyone have experience running Snappy on a raspberry? :) [10:54] Hanonim: yes, classic Ubuntu or Ubuntu Core? [11:01] jamiebennett: Ubuntu Core [11:02] jamiebennett: I'm new to snaps so I'm at that stage where it's quite frustrating not to have apt [11:03] My first question is this one : is it possible to use wiringpi on snappy core ? I've searched a lot and the one thing i found show it's quite impossible (or really hard?) [11:04] you can surely use it inside a snap [11:04] snap interfaces exposes all GPIOs, you just need to connect the interface as needed and wiringpi should be able to manage it [11:05] ogra_: is there a proper tutorial somewhere ? [11:05] the interfaces are named after http://pinout.xyz/# [11:05] i sthink http://snapcraft.io should have everything teaching you how to roll a snap and use interfaces with it [11:06] (if you run into issues, just ask here :) ) [11:06] also https://docs.ubuntu.com/core/en/ ... [11:06] ogra_: ok, i better study it better. now, my goal is to run a java app using pi4j, do you think it is "snapable" ? raspbian can run my stuff [11:07] sure ... [11:07] you might need to tweak some paths etc to make it work in the snap context, but if it runs on raspbian it will run on core too [11:08] ok well, i'll come back if i'm stuck [11:08] thanks very much for your help ! [11:09] that is why we are here ;) [11:11] a somewhat "simpler" solution might be to install ubuntu server, but then there is not support for the GPIOs, is there ? [11:20] Hanonim, i dont know i must admit ... you have to try :) [11:20] * ogra_ hasnt used classic installs on Pi in a long time [11:22] ogra_: the issue i see is that wiringpi writes to /dev/mem which is, i beleive, read-only in snappy core. is it the kind of problems snapcraft resolves ? [11:24] snapcraft just helps you to roll a snap ... device access is managed through interfaces ... [11:24] (the second link i gave above describs the security system and interfaces) [11:29] okay, i'll study it and stop asking too much questions :) [11:35] i'm surprised there is almost nothing on how to use gpios/pi4j on snappy core, i expected it to be a hot topic [11:41] Pharaoh_Atem: ouch on gitlab [11:41] Pharaoh_Atem: do you have the repo of the policy somewhere? [12:13] PR snapd#2756 opened: snapctl: add config in client to disable auth and use it in snapctl [12:26] Bug #1660957 opened: Need a way to get aa_is_enabled() and aa_query_label() to function [12:31] zyga, the repo is fine [12:31] I can push it elsewhere, yes [12:33] zyga, do you need it now? [12:44] Son_Goku: it would be good to keep it safe [12:45] Son_Goku: just worried that gitlab will go belly up [12:45] they're in the middle of recovery right now: https://www.youtube.com/watch?v=nc0hPGerSd4 [12:46] Son_Goku: wow, nice that they do it live [12:47] Son_Goku: but pretty terrible with the defunct backups and rm -rf mistake :/ [12:47] that was pretty damn unlucky, yeah [12:47] https://docs.google.com/document/d/1GCK53YDcBWQveod9kfzW-VCxIABGiryG7_z_6jHdVik/pub [12:47] I read the report [12:48] the used space graph is depressing [12:48] rm -rf /internet [12:48] nohup ;) [12:48] yep === JanC_ is now known as JanC [13:33] mvo: woo! [13:33] i'm confused ! in this video [13:33] https://www.youtube.com/watch?v=VbDAvTEBxew [13:33] hey Son_Goku! in a meeting right now, but I will read backlog and reply async :) [13:34] it's not ubuntu snappy core, it's ubuntu server ? [13:34] he uses apt and everything [13:34] yeah, he's not using snappy core [13:34] that stuff doesn't work there [14:03] hello there, I have a problem pushing to the store staff with classic confinement. Can I get help here ? [14:04] https://myapps.developer.ubuntu.com/dev/click-apps/6815/ [14:23] PR snapd#2757 opened: tests: add regression spread test for #1660941 [14:33] This list of interfaces : [14:33] http://snapcraft.io/docs/reference/interfaces [14:33] shows "gpio" and "i2c" interfaces [14:33] on my raspberry pi 2 i only see "io-ports-control", which is not referenced [14:34] is there a link between gpio and i2c provided by the RPI and the io-ports-control interface ? [14:35] Hanonim, http://paste.ubuntu.com/23905508/ [14:35] this is only in the edge channel builds currently though ... but shoudl migrate to stable eventually [14:36] jdstrand: hello! hey, we got hit by malformed 'name': '00-test-snap-build' lint-snap-v2_name_valid [14:37] daily edge builds can be found here http://people.canonical.com/~ogra/snappy/all-snaps/daily/current/ [14:37] the interface names follow http://pinout.xyz/# [14:37] jdstrand: (but the store allowed the name registration and as far as some knowldgeable people think, it should be a valid snap name. Do you recall the status of converging on one snap name regex to rule them all? [14:37] ogra_: nice, too bad it's not fully stable. what about i2c ? [14:38] what does io-ports-control do ? i can't find info [14:38] in the works [14:38] i think morphis knows about io-ports ... likely nothing you want to use [14:38] iirc thats some legacy thing [14:41] ogra_: hmmm, for now i'm a bit disappointed with the current state of affair. i might keep using raspbian unfortunately [14:42] Hanonim, well, come back eventually :) it is constantly improving ... [14:42] (and has many advantages over an apt based system ... ) [14:43] ogra_: yes, the idea is really good, just too restrictive right now (at least for my projets) [14:44] restrictive but secure :) [14:44] jdstrand: I have the regex that snapd is supposedly using and it allows the name I posted, so maybe the tools need to update the regex they use? let me know and I can share it [14:44] your devices wont be conquered by a botnet :) [14:44] ... yet :) [14:44] and reliable ... it is nearly impossible to break it [14:45] roadmr, pessimist :P [14:45] ogra_: yup, and documentation will help. for instance, i can't even find one example on how to create a java snap. maybe i didn't look properly [14:46] Hanonim, https://github.com/ogra1/jtiledownloader ... thats a desktop app though [14:46] but headless apps wont be much different to package [14:47] roadmr: we should be using the same snapd regex as snapd now. Did snapd change? [14:47] jdstrand: maybe! the regex is this ^(?:[a-z0-9]+-?)*[a-z](?:-?[a-z0-9])*$ [14:47] thanks [14:48] jdstrand: (I don't follow snapd that closely, my report is mainly about what the package submitter observed with that name) [14:49] this is what the review tools use: ^[a-z](?:-?[a-z0-9])*$ [14:49] let me check snapd [14:50] jdstrand: yes, your regex doesn't like the name I reported, while the one I gave, which should be snapd's, does [14:50] snapd is using: var validSnapName = regexp.MustCompile("^(?:[a-z0-9]+-?)*[a-z](?:-?[a-z0-9])*$") [14:51] it looks like they changed [14:51] jdstrand: /o\ [14:51] I'll adjust the tools [14:51] thanks jdstrand! [14:52] yeah, it changed in 88665e9a [14:54] jdstrand: oh! yes, that's about when we had that discussion on names, I think. [14:54] (early sept 2016) [14:55] roadmr: this has been an ongoing conversation, cause I distinctly remember updating the tools to have precisely what snapd has. In fact, before 88665e9a, snapd used ^[a-z](?:-?[a-z0-9])*$ (the same as the tools) [14:55] regardless, fixing it [14:57] jdstrand: thanks! [14:57] jdstrand: sorry it took so long to find this discrepancy and alerting you in such an indirect way :( [14:58] it's fine. I'll have a fix in just a moment [14:58] many thanks :) [15:03] roadmr: can you pull r835? [15:03] jdstrand: sure, coming up [15:08] jdstrand, hey, (I think it's you I should ask), can you approve uploads #39, #40 and #41 of the ubuntu-terminal-app snap? it's [15:09] jdstrand: hey [15:10] jdstrand: going through the apparmor issue, trying to wrap my head around what is going on [15:10] jdstrand: no luck yet but I'm trying more thigs [15:10] jdstrand: I replied on the stpcpy branch, not sure if you replied back yet [15:17] zyga: I saw and I did. bummer on reproducing [15:18] zyga: I'm having the fun time of debugging an intermittent spread failure. I run it from here. fine. from travis, fail [15:19] at least got some nice debugging improvements for snap-confine out of it [15:19] snap-confine testsuite* [15:20] jdstrand: when you ran it locally, which spread version did you use? [15:20] jdstrand: fgimenez found some issues in how some of the tests are ran that would explain some of the failures we saw (e.g. no core snap) [15:21] jdstrand: there was a new release of spread lately, perhaps upgrade to see if this is consistent with what you see on linode [15:21] I may be out of date. thanks for the tip [15:21] jdstrand, zyga hopefully will be fixed very soon [15:24] jdstrand: HA [15:24] jdstrand: I just got a super trivial way to reproduce this :) [15:24] jdstrand: snap install snapd-hacker-toolbelt --devmode [15:24] zyga: nice! :) [15:24] jdstrand: snapd-hacker-toolbelt.busybox sh [15:25] nessita, am i not supposed to be able to see my conjure-up package in the snap store? [15:25] jdstrand: nsenter -m/proc/1/ns/mnt [15:25] jdstrand: this gives me the following error in syslog: [15:25] jdstrand: Feb 01 16:25:13 xenial-server audit[22275]: AVC apparmor="ALLOWED" operation="open" info="Failed name lookup - disconnected path" error=-13 profile="snap.snapd-hacker-toolbelt.busybox//null-/usr/bin/nsenter" name="" pid=22275 comm="nsenter" requested_mask="r" denied_mask="r" fsuid=0 ouid=0 [15:26] jdstrand: I'll adjust the test case to simplfy this [15:26] jjohansen: ^^ if this helps you to have a look as well [15:26] jdstrand, jjohansen: (above please prefix Feb 01 16:25:13 xenial-server audit[22275]: AVC apparmor="ALLOWED" operation="open" info="Failed name lookup - disconnected path" error=-13 profile="snap.snapd-hacker-toolbelt.busybox//null-/usr/bin/nsenter" name="" pid=22275 comm="nsenter" requested_mask="r" denied_mask="r" fsuid=0 ouid=0 [15:26] er [15:26] jdstrand, jjohansen: above please prefix snapd-hacker-toolbelt.busybox sh with sudo [15:27] stokachu, checking, one sec [15:28] jdstrand: this essentially takes my patch out of the loop [15:28] jdstrand: as it is simply nsenter that fails [15:28] jdstrand: and the snap is installed in devmode so apparmor should be in complain mode [15:28] jdstrand: and snap-confine is out of the way (after setting up the namespace initially) [15:28] jdstrand: and also its apparmor profile is out of the way [15:29] nessita, thanks, i just tested an upload and it seems to be going through [15:29] zyga: note that apparmor is really out of the way if it is in complain mode. please give your finding to jjohansen [15:29] (there could be a complain mode bug) [15:29] stokachu, how are you uploading? [15:29] because I don't see you being defined as collaborator, which is weird [15:29] nessita, snapcraft push ./conjure-up_2.1.0_amd64.snap --release edge [15:30] nessita, it's still uploading so it may complain [15:30] jdstrand: it is! (snap install with --devmode) [15:30] stokachu, let me know how that does please [15:30] jdstrand: yeah, I'll prep a mail [15:30] goes* [15:30] oh, need to run to pick up kids from school [15:30] zyga: I'm saying that just cause it is in devmode doesn't mean that an apparmor complain mode bug isn't causing trouble [15:32] nessita, http://paste.ubuntu.com/23905726/ did show an error [15:34] stokachu, asking for some advice, one sec [15:34] nessita, ok ty! [15:35] heya, zyga, kyrofa: any tips on how I would go about running the xdg-open shim from ubuntu-core? the way I see it a snap would have no way to get ahold of ubuntu-core stuff given http://snapcraft.io/docs/reference/env [15:38] sitter: I get to open just fine for konversation (using now); only thing missing is to set it as the default action [15:39] sitter: http://imgur.com/a/aArpk [15:40] sitter: there is a host side bug though which is that nothing is currently depending on snapd-xdg-open; I thikn zyga was looking into that [15:41] hm [15:41] sitter: aside from not being able to set it as a default action, only other thing I am missing is proper icons; I get `?` most likely due to the Icon entry having a value of a name that doesn't exist on my host. [15:41] sergiusens: indeed. I think it is only QDesktopService::openUrl that falls over [15:42] sergiusens: e.g. try help -> about konversation and click any link [15:42] sitter: yeah, that doesn't work [15:42] which supposedly comes back false here https://github.com/qt/qtbase/blob/5b1f18b0f1cd07f2dfcafe79b456338f908de531/src/platformsupport/services/genericunix/qgenericunixservices.cpp#L91 [15:43] yeah, we did develop snapd-xdg-open and a fake xdg-open in core snap side for that reason [15:43] dependency was still to be sorted out, zyga is owning this AFAIK [15:44] ah, I get it. usr/local/bin is missing from PATH ^^ [15:45] oh? I remember to have mentioned it on the ML and I thought after all our discussions that was fixed [15:46] certainly isn't when using `snap run --shell` nor is it documented at http://snapcraft.io/docs/reference/env [15:46] stokachu, so seems like what I promised about you being added automatically as a collabortor was not true, sorry. Could you please ping mvo and ask him to invite you as collaborator for that snap? [15:46] no, I guess that wasn't fixed after it was removed from PATH after all (and you're right) [15:46] stokachu, mvo should fill the form under https://myapps.developer.ubuntu.com/dev/click-apps/5479/collaboration/ [15:47] nessita: sure, I can add stokachu [15:47] ok [15:47] sitter: didrocks if I get a bug report on snapcraft I can easily add the PATH [15:47] mvo, oh thank you [15:47] mvo, do you need my email or anything? [15:47] okay, i've read quite a bunch of docs about ubuntu core and snapcraft and i don't see how i could use wiringpi (that needs to write to /dev/mem) [15:47] stokachu: yes, your ubuntu-one primary mail would be nice [15:47] sergiusens: yeah, I wonder if that's the right place, but it can help for now as it seems it doesn't have traction snapd-side :) [15:47] stokachu: just /msg it to me [15:48] sitter: mind opening the bug for sergiusens? [15:48] I can add it to desktop-launch as well for Qt apps for now [15:48] (but doesn't help on KDE's one as you are using your own launcher) [15:50] sergiusens: https://bugs.launchpad.net/snapcraft/+bug/1661023 [15:50] Bug #1661023: PATH does not include /usr/local/bin and /usr/local/sbin [15:50] didrocks: yeah, easy enough to work around for now [15:53] nessita, thanks for you're help, im good to go now [15:53] your* [15:59] about dialog links work like a charm with modified PATH. thanks for the help :) [15:59] stokachu, excellent, sorry for the "honest" lie yesterday [16:00] sitter: nice! btw, we are going to promote your KDE snaps in the "top snaps of the month" for January :-) [16:02] \o/ [16:06] PR snapd#2756 closed: snapctl: add config in client to disable auth and use it in snapctl [16:45] re [16:45] jdstrand: I agree, I think you are exactly right [16:45] didrocks: hey [16:46] didrocks: so the snapd-xdg-open bug [16:46] didrocks: AFAIR there was some discussion and it didn't culminate with a decision as to how the dependency should be injected [16:47] didrocks: I don't feel I own this bug actually, I think it belongs to foundations/desktop or if it belongs I don't have a plan to solve it anytime soon; adding the dependency is easy but it would be complex on server/headless installs [17:07] Morning everyone, Is anyone here using the snap version of Telegram? [17:15] I do HumbleBeaver [17:15] which one? telegram-latest or telegram-sergiusens [17:15] stokachu: Both actually [17:16] oh are they both required to be installed? [17:16] * stokachu doesn't know which one to use [17:16] I'm using telegram-sergiusens and no, you don't need them both [17:16] No I had one installed and when it started to mess up, I uninstalled it and tried the other [17:17] I'm thinking it might have something to do with an extension I added when I installed the kconnect ppa. [17:17] jjohansen: hey, can you please ping me when you are around. I have some good progress on the bug we've been talking about [17:20] pachulo have you had any issue with it not allowing attachments? [17:20] If not its something on my system and I'll do some digging. [17:35] only issue ive seen with this telegram app is the tray icon is missing [17:40] thanks stokachu, I've been fighting with an issue on my system and I'm starting to see a pattern. [17:41] Why it affect the snaps I'm not sure [17:41] HumbleBeaver: the tray icon issue? [17:41] HumbleBeaver: because it's a broken protocol, it's been researched by a few folks AFAIK. The actual icon is saved to /tmp and /tmp is private for each snap. [17:42] there are a few other ways to show an icon there so it depends on the toolkit/app [17:42] so that requires a user talking to upstream to get that fixed? [17:43] we should have /tmp-IPC backend for cases like this [17:43] stokachu: it's a complex issue, I think [17:43] stokachu: part of legacy protocols, old agrements, not sure [17:43] ok, wasn't sure if it was something we could workaround in the snapcraft [17:43] with organize or something [17:43] stokachu: there's a bug that tracks this with a lot of details [17:43] stokachu: doubtful [17:43] stokachu: more likely in snapd [17:43] ah ok [17:44] stokachu: or at a toolkit level [17:44] stokachu: but then telegram bundles qt so you'd have to fix it and advertise the fix [17:44] would a classic snap work ootb for now? [17:44] yes [17:44] no [17:45] we found some issues with the design of classic snaps [17:45] it is a nitpick detail but a lot of users would probably not like having a default icon there [17:45] we know how to fix them but the work has not started [17:45] zyga, ah ok [17:45] I think I'd stick for strict/devmode for now [17:47] zynga, no the file attachment bug. I'm starting to think that installed something that has caused snap packages and others to behave strangely [17:48] I've asked a few others and they say it works fine. I can also drag an drop file into telegram without issue [17:49] jjohansen, jdstrand: https://bugs.launchpad.net/apparmor/+bug/1656121/comments/13 [17:49] Bug #1656121: unexpected errno=13 and disconnected path when trying to open /proc/1/ns/mnt from a unshared mount namespace [17:50] HumbleBeaver: I don't know anything about file attachment bugs but I could say that since the snap works in a different namespace it doesn't see the same filesystem as your desktop [17:51] jdstrand: the last thing on that comment is key I think [17:51] jdstrand: note that we don't see the disconnected profile anymore [17:52] er, path [17:52] jjohansen: if you have a branch I could test with a locally built kernel I'd like to keep digging [17:52] PR snapd#2758 opened: overlord/devicestate: implement policy about gadget and kernel matching the model [17:53] jjohansen: this is my top priority now [17:59] ogra_, how would one enable SPI on the joule? [17:59] kyrofa, oh, no idea [17:59] never seen a joule [17:59] kyrofa: is /dev/spi* around? [18:00] ogra_, turns out that's the hw being used in that askubuntu question I pinged you about yesterday :P [18:00] yeah [18:00] zyga, no idea [18:00] ogra_, any idea who made the gadget for that? [18:00] i remember but i have no idea about joule ... you have to ask someone knowing the HW [18:00] kyrofa: do you have a joule? [18:01] kyrofa, JohnAgosta might be able to point you to someone [18:01] zyga, I'm afraid not [18:01] kyrofa: meh, too bad then [18:01] JohnAgosta, any idea who made the Joule gadget? [18:02] kyrofa, are you referencing the gadget snap in the Joule image? [18:03] https://developer.ubuntu.com/core/get-started/intel-joule [18:10] PR snapcraft#1078 closed: tests: do not rely on project_dir [18:18] JohnAgosta, sorry, internet flaked on me. Yes, that's what I'm referring to [18:20] kyrofa, that was created by either of Robert Liu or Tim Chen. They are on Chinese New year break [18:21] JohnAgosta, do you know of anyone around who might know how to enable SPI? [18:21] kyrofa, please make sure you are looking at the correct one...this was all updated to beta4 images in the past week [18:22] kyrofa, my brain is flaking out -- what is "SPI"? [18:22] JohnAgosta, Serial Peripheral Interface [18:22] JohnAgosta, hardware interface protocol [18:23] Like i2c [18:23] JohnAgosta, at least in e.g. the rpi, it needs to be enabled in the gadget [18:24] JohnAgosta, we have an askubuntu question about it seemingly not being enabled on the joule, and no one knows how to enable it :P [18:24]