2020/09/30/#snappy.txt

[06:01] <mborzecki> morning
[06:55] <pstolowski> morning
[07:00] <mborzecki> pstolowski: hey
[07:07] <mvo> good morning pstolowski and mborzecki
[07:07] <mborzecki> mvo:  hey
[07:10] <pstolowski> o/
[07:50] <pedronis> mvo: hi, why is #9399 not merged?
[07:51] <mvo> pedronis: let me check
[07:51] <pedronis> I might have confused you
[07:51] <zyga> o/
[07:51] <mvo> pedronis: yeah, it did not get a +1 from you so I did not merge it
[07:52] <mvo> pedronis: if it's good I will merge it now, but the boat for 2.47 is missed :/
[07:52] <pedronis> mvo: I pushed to it
[07:52] <mvo> pedronis: uh, sorry, then I was just too cautious
[07:52] <pedronis> it sounded you wanted it in 2.47 ?
[07:53] <mvo> pedronis: yes, it's a bit unfortuante
[07:53] <pedronis> sorry, I pushed to it and then ignored it, I thought you would just merge it or ask me again
[07:53] <mvo> pedronis: I mostly want it for groovy, I would do a 2.47+20.10.1 with just this merge in the deb
[07:53] <pedronis> also I hoped my follow up would get a review
[07:53] <mvo> pedronis: sorry for that, but it's not the end of the world, let me squash merge it now
[07:55] <mborzecki> pedronis: hi, i've updated #9427, sadly with the s/ChangePreserveBefore/ChangeIgnore/ and the other tweaks it grew a bit
[07:55] <mvo> pedronis: merged and cherry picked
[07:55] <pedronis> mborzecki: yea, it's a bit big now :/
[07:55] <mborzecki> pedronis: and there's a separate change that uses bootloader.ForGadget() in makebootable, which i can propose separately if that's easier to review
[07:56] <mborzecki> pedronis: fwiw, i can propose the ChangePreserveBefore -> ChangeIgnore rename separately too
[07:57] <mvo> pedronis: I will look at the followup today. quick question about the snapshot setid work from pawel, would you prefer to review all of those prs or do you think two regular reviews are sufficient? was wondering if that the import is something we could finish before the sprint (or do you think this is utterly unrealistic?)
[07:57] <pedronis> mvo: they are on my list to look at them quickly today
[07:58] <mvo> pedronis: cool, thanks
[08:03] <pedronis> mborzecki: if we can make it a bit smaller it would be better I think
[08:04] <mborzecki> pedronis: ok
[08:05] <pedronis> mvo: anyway my PR #9424 needs a little bit more work
[08:14] <mvo> ok
[08:16] <zyga> mvo thanks for the review, that really speeds things up
[08:16] <zyga> I'll merge and iterate as there is more :)
[08:16] <zyga> I will look at both pawel's and your comments
[08:17] <mvo> zyga: they where style comments on my side
[08:17] <mvo> zyga: how is the snapd->snap userd part comin along?
[08:17] <mvo> zyga: would be amazing to demo something
[08:18] <mvo> zyga: shame we don't have lightning talks
[08:18] <zyga> not much progress yesterday but that's my focus now
[08:18] <zyga> is that due this week or next week?
[08:20] <mvo> zyga: next week is the sprint
[08:20] <mvo> zyga: the better story we have then, the better
[08:23] <zyga> mvo, I think I will have a demo for the standup tomorrow
[08:30] <mvo> cool
[08:37] <pedronis> pstolowski: mvo: I reviewed #9417
[08:43] <pstolowski> ty
[08:45] <pedronis> pstolowski: and #9434
[08:47] <pstolowski> great, ty
[09:39] <mborzecki> pedronis: split out the gadget bits, `3 files changed, 447 insertions(+), 57 deletions(-)` that should make it more manageable
[10:04] <pedronis> mborzecki: thx
[10:11] <mborzecki> zyga: do you recall whether the apparmor profile allows to run ping? https://forum.snapcraft.io/t/cant-run-the-simplest-ping-script/20275
[10:11] <mborzecki> iirc ping used to be (is still?) setuid
[10:25] <pstolowski> pedronis: i've solved XXX in #9417 (by passing setid to Open) but it affects a few other places, ok if this is a followup?
[10:29] <mvo> pstolowski: fwiw, I'm in favor of a follwup
[10:29] <mvo> pstolowski: then I can merge 9417 into the import PR and make it slightly smaller
[10:29] <pstolowski> ok
[10:39] <mvo> pstolowski: but there is a shellcheck issue in this PR that is real and will need attention
[10:40] <pstolowski> mvo: yes i just fixed it and checking
[10:40] <mvo> pstolowski: just run "./spread-shellcheck tests/main/"
[10:40] <mvo> pstolowski: nice
[10:41] <zyga> mborzecki looking
[10:41] <zyga> mvo I don't remember
[10:41] <zyga> mborzecki, nope, no ping
[10:41] <zyga> mvo notifications work now
[10:42] <zyga> mvo should brainstorm for 3 minutes the wording
[10:42] <zyga> but I just did a "run chrome, switch and refresh"
[10:42] <zyga> so if you want it in edge tomorrow I need reviews :)
[10:42] <zyga> mvo GNOME experience is, sadly, terrible
[10:43] <zyga> but I will check why and try to fix it
[10:43] <zyga> (the notification is on screen far too short to read comfortably)
[10:43] <zyga> it gets stashed in the roster though, so that's oka
[10:43] <zyga> I will try the plethora of hints available
[10:43] <zyga> but it really works ")
[10:46] <mvo> zyga: nice!
[10:46]  * mvo hugs zyga
[10:48] <zyga> https://github.com/snapcore/snapd/pull/9438 is short and simple
[10:48] <zyga> improves how to pass expiration timeout
[10:51] <pedronis> pstolowski: as you prefer, that PR is not big either way
[11:06] <jamesh> zyga: sorry for not giving a review of your notifications branch.  I notice you're implementing a lot of features not used by modern shells.  That's one of the reasons I was looking at both the fdo and gtk APIs together rather than focusing on fdo
[11:28] <pstolowski> mhm, i've seen "cannot sign assertion: bad GPG produced signature: .." more often recently
[11:38] <zyga> jamesh: I'm aware of that, we can trim it down later, most of the "features" are really just constants
[11:39] <zyga> all of the APIs there are working in current gnome (fortunately)
[11:40] <zyga> jamesh I wrote some code using this now, from snapd, via userd, so that's going to be a good chance to discuss what we want to show exactly
[11:44] <zyga> mvo: I need a review of small 9422
[11:44] <zyga> perhaps someone could do it before samuele can do the architecture review
[11:47] <pedronis> zyga: I reviewed #8573
[11:47] <zyga> thank you, let me look
[11:49] <zyga> great feedback, thank you
[11:49] <zyga> I will pick that up tomorrow, today is the notifications side day
[11:51] <pedronis> zyga: ok, I'll ty to also look at #7700 today, that should be simpler?
[11:51] <pedronis> *try
[11:56] <jdstrand> mborzecki: ogra is right, you need network-control because ping is setuid so need 'capability setuid'. there are other higher up the stack was to check networking that don't need that cap if people don't want network-control
[11:57] <pedronis> pstolowski: I did a quick pass on #9442
[11:57] <mborzecki> jdstrand: thanks for confirming
[11:57] <pstolowski> pedronis: ty
[11:58] <dariball> hi, ubuntucore docs (https://core.docs.ubuntu.com/en/stacks/disk/udisks2/docs/installation) mention using "udisks2" for automount support, but the snap is not found: https://snapcraft.io/search?q=udisks2 ... any hint on what I am doing wrong ?
[11:58] <jdstrand> ways*
[11:59] <dariball> or is there a replacement for "udisks2" which is not yet part of the docs?
[12:00] <zyga> degville ^
[12:00] <jdstrand> network-observe would also do
[12:00] <jdstrand> I responded to the topic
[12:04] <degville> dariball: sorry - the udisk2 snap has been deprecated. We should definitely add that to the Ubuntu Core docs. I'm not sure of an alternative, but this may provide a clue: https://forum.snapcraft.io/t/udisks2-snap-removed-alternatives/12467/
[12:07] <dariball> degville: ok, good to know, thank you
[12:57] <pedronis> degville: pstolowski: I updated based on feedback https://github.com/snapcore/snapd/pull/9424#issuecomment-701357792
[12:59] <degville> pedronis: thank you! on first glance, it looks really good.
[14:01] <mvo> pstolowski: I merged master into 9442 and will review now
[14:07] <pstolowski> ty!
[14:31] <pstolowski> quick errand, bbiab
[14:40] <mvo> pstolowski: I did merge but the reviewing did not quite work out due to meetings, will try again a bit later
[14:48] <pstolowski> re
[14:48] <pstolowski> sure np
[14:50] <pedronis> zyga: the work you are doing will partly replace #7700, right ?
[14:55] <parker__> Hi everyone, does anyone here know how I might install a `.deb` as a part?
[14:58] <parker__> http://pastie.org/p/5TdqAudVPiPfRR1GMRI7aU -- I'm trying to install faudio before I install wine
[14:58] <zyga> pedronis: in a way, yes
[14:58] <zyga> pedronis: I think the "hold and wait" aspect will not change but the UI will
[15:00] <pedronis> pstolowski: I approved it but I noticed this in the other PR: https://github.com/snapcore/snapd/pull/9434/files#r497580398
[15:02] <pedronis> pstolowski: actually, maybe I'm just confused
[15:04] <pstolowski> pedronis: TestLastSnapshotIDErrorOnDirName hits tis
[15:04] <pstolowski> *this
[15:05] <pedronis> pstolowski: sorry, I was confused, see my comment in the PR
[15:07] <pstolowski> np
[15:10]  * cachio lunch
[16:01] <ijohnson> pedronis: I have more fun raciness I have just discovered to discuss in our meeting
[16:02] <ijohnson> pedronis: autorefresh.Ensure() grabs a lock, then releases the lock while doing a bulk refresh assertion and console-conf-start gets the lock doesn't see a change, then unlocks and Ensure() comes back and creates a new change
[16:02] <pedronis> ijohnson: well, I asked to look into that
[16:02] <pedronis> I remembered there might be something like that going on
[16:02] <ijohnson> pedronis: and I did, but apparently I didn't look closely enough
[16:03] <ijohnson> anyways I'm looking at remedies now
[16:21] <parker__> Hi all, got disconnected so unsure if this was answered
[16:23] <parker__> I'm trying to install wine5.18 which dependents on the library `faudio`. I've added both the 64 and i386 debs as parts but they don't seem to be getting installed http://pastie.org/p/5TdqAudVPiPfRR1GMRI7aU
[16:24] <zyga> mvo: https://twitter.com/zygoon/status/1311340971175673861
[16:24] <zyga> I will  tweak the text some more, remove hacks and open a PR
[16:25] <zyga> but I need to do an errand with my wife now, we'll be back in two hours, traffic permitting
[16:33] <ijohnson> very nice zyga !
[16:46] <cachio> zyga, when you are back could you please take a look to #9414 and #9425
[17:43]  * cachio -> kinesiologist
[18:53] <zyga> ijohnson https://github.com/snapcore/snapd/pull/9446
[18:54] <zyga> cachio tomorrow morning
[18:54] <ijohnson> \o/
[18:54] <ijohnson> nice I will try and get you a review to try and land this asap
[18:54] <zyga> it's behind a feature flag
[18:54] <zyga> so we can change anything without impact
[18:54]  * ijohnson still needs to finish a review for pawel's service pr first though
[18:54] <zyga> so yeah, LGTM if we can :)
[18:54] <zyga> I suspect it will be quite a shed for bicycles though ;)
[18:54] <ijohnson>  haha
[18:55] <zyga> because it's short and man, wording ;-)
[19:42] <parker__> Another interesting problem I've been unsable to solve `The 'wine' part is missing libraries that are not included in the snap or base. They can be satisfied by adding the following entries to the existing stage-packages for this part:
[19:43] <parker__> This then lists many packages which I have added into the stage-packages list, but upon remaking the snap the error message remains the same
[19:44] <ijohnson> parker__: is the part that uses those stage-packages also define a "prime" or a "stage" key? if so you may be accidentally removing the files from those stage-packages
[19:46] <parker__> ijohnson: not that I can tell http://pastie.org/p/1FQ7eewGAX2kDPoxgGTzK2
[19:46] <ijohnson> parker__: what is this line doing:      `rm -rf ${SNAPCRAFT_PART_INSTALL}/usr`
[19:47] <parker__> oh shoot...
[19:47] <parker__> that would definately cause a problem
[19:48] <parker__> cheers
[19:48] <ijohnson> np
[19:50] <parker__> +1 seems to have worked! hah