[05:54] PR snapd#9429 closed: o/daemon: validation sets api and basic spread test

[07:50] good morning [08:00] morning [09:03] PR snapcraft#3392 opened: plugins v2: add support for out of source tree builds [09:20] PR snapd#9796 closed: devicestate: log checkEncryption errors via logger.Noticef [09:35] PR snapd#9797 closed: secboot: add workaround for snapcore/core-initrd issue #13

[10:01] pedronis: hey, i've updated #9513 [10:01] PR #9513: snapshotstate: detect duplicated snapshot imports [10:02] pstolowski: thx [10:28] hey guys [10:36] PR snapd#9799 opened: secboot: add extra paranoia when waiting for that fde-reveal-key [10:37] pedronis: are you working on any changes to api_snapshots_test.go to make it use apiBaseSuite or can I do it? I noticed invalid endpoints there (/v2/snapshot/import) which of course do not affect the test but are confusing [10:45] pstolowski: no, not working on it, you can change it if it makes sense. ATM my work in progress there is about api.go itself [10:46] ack [11:51] PR snapd#9800 opened: tests: use apiBaseSuite for snapshots tests, fix import endpoint path [12:51] pstolowski: thanks for #9800, I made some remarks, let me know if they make sense or you we should chat more [12:52] PR #9800: tests: use apiBaseSuite for snapshots tests, fix import endpoint path [12:54] thanks, looking [12:58] pedronis: i think we actually have what you suggested there, daemonWithOverlordMock(c) in apiBase (I missed that), or I misunderstood? [12:59] pstolowski: yes, but we do something about store, but the store shouldn't be involved here, so maybe the old code was also strange [12:59] and just copied from somewhere else? [13:00] pedronis: ah i see [13:14] pedronis: i've renamed the old helper to s.daemonWithOverlordMockAndStore(c) (and with that updated a bunch of api tests to use that) and added a new one s.daemonWithOverlordMock(c) that is used now by snapshots tests, does this sound ok? [13:15] the new helper doesn't create store [13:23] sounds reasonable [13:31] pedronis: pushed [13:33] hey pstolowski and pedronis - anything I should look at? [13:35] mvo: I made some changes to https://github.com/snapcore/snapd/pull/9789 prompted by Ian, maybe you want to look, I'm not 100% happy with them [13:35] PR #9789: many: separate consistency/content validation into gadget.Validate|Content [13:38] pedronis: ok [13:38] mvo: it's the last commit there [13:47] mvo: hey, you can look at my changes to snapshots PR [13:48] pstolowski: sure thing [13:53] pstolowski: I will also review those [14:48] ijohnson: I did a pass on #9798, thanks for it [14:48] PR #9798: o/{device,snap}state: enable devmode snaps with dangerous model assertions [14:48] thanks [14:51] pedronis: I responded can you take a look? [15:26] pedronis: core20/basic test is failing in release/2.48 because it cannot get a serial from the store for our custom model, do you remember what PR/commit I should search for that needs cherry-picking to 2.48? [15:26] mvo: one sec [15:28] mvo: this one I think: https://github.com/snapcore/snapd/pull/9713 [15:28] PR #9713: tests: sign new nested-18|20* models to allow for generic serials [15:28] pedronis: thank you! [15:42] mvo: pstolowski: https://github.com/snapcore/snapd/pull/9513#discussion_r543455554 [15:42] PR #9513: snapshotstate: detect duplicated snapshot imports [15:48] ty [15:48] mvo: pstolowski: also "// XXX: Something needs to cleanup incomplete imports. This is conceptually" has been addressed, right? [15:48] I see code in StartUp for that [15:50] pedronis: yes [15:51] we should remove the XXX in one of the PRs [15:51] ok [15:52] PR snapd#9786 closed: tests: backport netplan workarounds from #9785 (2.48) [15:52] PR snapd#9801 opened: tests: sign new nested-18|20* models to allow for generic serials (2.48) [15:54] pedronis: it sounds to me like the snapshot mgr local lock that you suggested should be a followup, this may touch a number of places? [15:55] pstolowski: yea, I fear we'll have to split import into prepare and import (like some things are split already in snapshotmgr.go if you look) [15:56] pstolowski: export might have the same problem? [15:58] pstolowski: or maybe the fix is easy, because we don't worry too much about blocking expiration [15:58] export and import are rare enough [15:58] pedronis: i'd expect export to fail (potentially in the middle of doing things) though? [16:00] no, we care about blocking expiration too much because is done from Ensure [16:00] :/ [16:15] pstolowski: mvo: now that I look at the code we have some issues for sure still pending [16:20] pedronis: should we discuss in a HO? [16:23] pstolowski: I wrote a note in the PR, let me know if need to chat more [16:23] it all be for after the break anyway at this point [16:23] *it might all be [16:25] pstolowski: https://github.com/snapcore/snapd/pull/9513#issuecomment-745401821 [16:25] PR #9513: snapshotstate: detect duplicated snapshot imports [16:27] pedronis: yes and i don't think we should block this PR, it seems like a change on its own [16:30] pedronis: did you had a chance to look at 9799? if it makes it into 2.48.2 that would be nice but I think it's not essetnial. I want to release later tonight [16:38] mvo: yes, but I forgot to comment on it, also maybe ijohnson should look at it [16:39] mvo: pedronis: sure I will look at it, but probably in my PM at this point, have a meeting and then lunch for me [16:59] PR snapcraft#3393 opened: requirements: uprev python-apt [16:59] PR snapcraft#3394 opened: requirements: rely on host's python3-apt [17:58] PR snapcraft#3393 closed: requirements: uprev python-apt [17:59] PR snapcraft#3384 closed: Add core20 to various /Dockerfile/s [18:04] PR snapcraft#3374 closed: [wip] ci: migrate spread tests to github actions [18:14] PR snapcraft#3365 closed: github actions: bootstrap CI workflow [18:15] PR snapcraft#3367 closed: requirements: pin pip to 20.1.1 for devel [18:53] PR snapd#9799 closed: secboot: add extra paranoia when waiting for that fde-reveal-key [18:55] PR snapcraft#3347 closed: Create an extra 'build' directory for plugins [19:00] PR snapcraft#3348 closed: ROS plugins use plugins' build directory [19:15] PR snapcraft#3381 closed: multipass build provider: check if instance exists before deleting [19:15] PR snapcraft#3389 closed: plugins v1: move build base test in v1 plugins to base class [19:16] hm, 9702 is the last pr tagged 2.48 but it has no security review [19:16] so tough luck I guess [19:17] mvo: that pr is not high priortiy so that's fine I think [19:17] it's more of a "nice to have some day" [19:18] PR snapd#9801 closed: tests: sign new nested-18|20* models to allow for generic serials (2.48) [19:40] PR snapcraft#3107 closed: fix yarn version url [19:55] PR snapcraft#3144 closed: docker: setup a multiarch build for snapcraft docker images [20:00] PR snapcraft#3243 closed: Add skip-keys to catkin plugin [20:00] PR snapcraft#3262 closed: specifications: configurable apt mirror [20:04] ijohnson: do you need me to review something? [20:05] PR snapcraft#3283 closed: specifications: unified provider [20:05] PR snapcraft#3300 closed: [wip] remote lxd / cross-arch support [20:07] pedronis: not right now, I just pushed the tests for my devmode pr [20:08] but that can be reviewed tomorrow [20:08] oh I forgot the nested label [20:10] PR snapcraft#3309 closed: Rough draft of changes needed to separate python versions between gno… [20:25] PR snapcraft#3284 closed: build providers: rename clean() -> clean_parts() to clarify scope [20:25] PR snapcraft#3291 closed: build providers: make mount_project() interface private [20:28] PR snapd#9802 opened: release: 2.48.2 [20:36] ijohnson: I made this self-comment in my open PR: https://github.com/snapcore/snapd/pull/9789#discussion_r543662855 [20:36] PR #9789: many: separate consistency/content validation into gadget.Validate|Content [20:40] * ijohnson looks [20:41] pedronis hmm I'll need to think on it, I'll try to review that pr again this afternoon [20:42] ijohnson: it's the principle that if the argument carry too much behavior change maybe one needs more function variant, it can combersome but it's usually the right call [20:43] I see yeah that makes sense [20:48] ijohnson, pedronis I call it a day (and a year :) I talked to the taipei guys about 2.48.2, if there are any issues just ping me on telegram I will try to help. [21:08] pedronis: I like your newest proposal the best [21:08] ijohnson: ok, I'm working on it [21:08] nice thanks for that [21:15] PR snapcraft#3380 closed: snap packaging: fix broken path check when copying local icon [21:37] ijohnson: pushed [21:38] pedronis: ok, will take a look [21:40] the diff should also be smaller as a consequence [21:40] PR snapcraft#3356 closed: extensions: correctly spell the GTK_USE_PORTAL variable [21:40] PR snapcraft#3362 closed: project loader grammar: do not transform build-packages === mup_ is now known as mup [21:55] PR snapcraft#3387 closed: ci: migrate snap publishing to github actions [23:25] ijohnson, hey [23:25] ijohnson, around? [23:27] ijohnson, netplan test failing like https://paste.ubuntu.com/p/yyVDwTcd5p/ [23:27] during beta validation