14:03 #startmeeting Weekly Main Inclusion Requests status 14:03 Meeting started Tue Jan 14 14:03:39 2020 UTC. The chair is cpaelzer. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 14:03 14:03 Available commands: action commands idea info link nick 14:03 can you repeat for the log jamespage? 14:03 #link https://bugs.launchpad.net/ubuntu/+source/ovn/+bug/1859601 14:04 context is that we've split the ovn source package out of the openvswitch package (already in main). 14:04 as I said this morning, you already have my ack jamespage 14:05 the objective for 20.04 is to make OVN the default SDN for OpenStack on Ubuntu 14:05 as its source was in main before it has had reviews and such already 14:05 so I've seeded the required binaries into one of the server supported seeds 14:06 joeubuntu: you need to be the third vote to decide, do you agree that in such a case we can fast-path ack it to be promoted? 14:06 +1 14:07 jamespage: ok I'm setting the bug state 14:07 ta 14:07 for the reasons cpaelzer and jamespage laid out. 14:07 jamespage: only it is listed as "unsubscribed" 14:07 jamespage: could you do the pkg-subscription for openstack team whie I change state 14:07 I sorted that out this morning - might not have refreshed yet 14:07 ah ok thanks 14:08 jamespage: the update is done, you just need to get an AAs attention as usual 14:09 other cases for discussion from you jamespage? 14:09 ta - no rush I'll chase one down if it does not happne 14:09 no - I think ceph-iscsi and nfs-ganesha are both with joeubuntu and the security team for review 14:09 ok then topic switch - I have an own bug(s) to discuss 14:09 * jamespage listens 14:10 https://bugs.launchpad.net/ubuntu/+source/containerd/+bug/1819761 and https://bugs.launchpad.net/ubuntu/+source/runc/+bug/1817336 14:10 Launchpad bug 1819761 in containerd (Ubuntu) "[MIR] containerd" [Undecided,New] 14:10 Launchpad bug 1817336 in runc (Ubuntu) "[MIR] runc" [Undecided,New] 14:10 last discussed on the engineering sprint 14:10 let me summarize the background 14:10 1. plenty of golang dependencies 14:10 2. comes with the deps vendorized 14:10 3. we were tasked to anaylze the dependencies if some of them are packaged 14:11 4. this is a special case as runc/containerd will NOT stay at their version, they will get regular bumps to new major versions 14:11 like we did with docker.io in the past 14:11 and so their dependencies will change over time as well right 14:11 ? 14:11 yes jamespage, they will update 14:11 cpaelzer: no, technically not a member anymore 14:12 cyphermox: :-/ but still o/ 14:12 therefore the decision in Paris was that we'd want dependencies that have extra packages alrady in main (e.g. due to LXD work in the past) to use these packages 14:12 cpaelzer: MIR is Canonical-specific, AFAIK; since I'm not Canonical anymore, I wouldn't necessarily be representing the best interests, but I'll be happy to give you my opinion on software when asked :) 14:13 if on a bump the vendored dependencies are needed on a newer level we'd swicth from package to vendored dependency 14:13 while the other vendored dependencies (since they'd move anyway and have no initial better review-path) can stay vendored 14:13 now the analysis is complete and was added to the bugs 14:13 TL;DR: none of the dependencies are in main already 14:14 concern - both juju and lxd are no longer shipped in the archive so are they archive maintainers for the packages that have been split out 14:14 ? 14:14 they/there 14:14 yes they are still 14:14 I suspect the LXD team where probably doing some of them 14:14 as usual with a MIR who brought it in owns it 14:14 but they might have dropped to universe since then 14:15 are they also in Debian? 14:15 most, but not all are packaged in Debian/Ubuntu 14:15 just considering that if they are Ubuntu only and end up with no reverse depends they might pop up on a potential to remove list 14:16 TBH I struggle to hate vendoring - on the one side I appreciate that its not the distro model and muddies the view of the world 14:16 on the other hand - its what upstream have tested with... 14:16 that said we've taken a similar approach in ceph 14:16 as long as we can easily track the versions in the vendored code I am not against it. 14:16 (and that's not go) 14:16 jamespage: joeubuntu: what I'd want from you know is considering the above; the Paris discussion if you remember; and the bugs; then with me discuss/ack that we'll go forward on this case with the MIR and security review keeping dependencies vendored 14:17 kanashiro: ^^ if you want to chime in as well 14:17 I think we should proceed as planned in Paris . 14:17 kanashiro: ahasenack were the ones working on prepping the MIRs and the package 14:17 joeubuntu: ok, that would mean since nothing is in main we keep all vendored then 14:18 tbh I think that's the reality of where the packages would be in 6-9 months anyway 14:18 so at least if we've reviewed whats vendored we have a know good startline... 14:18 sorry, typing too much ... 14:18 We've got the ability to track it on our end, so I'm cool with it. 14:19 ok, I'd MIR ack the bugs then and "all that is left" is the security review then (a lot I know) 14:19 it already is assigned to security and on their trello 14:19 doko: anything to add since you are now with us? 14:19 we can give you a few min to inhale all the backlog here :-) 14:20 As an update we are down to 9 MIRs from a high of 14 in December. 14:20 not from my side 14:20 do you mean a way to track the diff of vendorized deps between versions in both packages? 14:21 kanashiro: joeubuntu and all of seucrity will need to track if containerd has dependency A at version 1.2.3 14:21 kanashiro , if that is for me, we have the ability to track the version of vendorized code in all vendored packages 14:22 in case 1.2.3 becomes known to be affected to know that we also need to change the package that vendorized it 14:22 I can't recall the details of how... but the tools exist. 14:22 kanashiro: have you added the result of the package anylasis to both bugs already runc and containerd? 14:22 anylasis->analysis 14:23 cpaelzer, I just need to add a comment to containerd bug, but I have it in my notes 14:23 note we will have potentially two versions of vendor code in the archive: the package in universe, and the vendored code inside runc/containerd 14:23 I only see it on the runc bug 14:23 kanashiro: ok please add it there 14:23 runc ibug is ok 14:23 ahasenack: yeah but that (two versions) is the natural consequence of vendorizing 14:23 yep 14:23 ok I conclude that we are ok and will update the bug then 14:24 lets go on with the tail end of the "normal" MIR meeting 14:24 * kanashiro is copying and pasting his notes regarding containerd now 14:24 #topic Review of previous action items 14:24 I'm not aware of recent action items 14:24 anything for you others? 14:25 ok - I take that as a no, then ... 14:25 #topic New MIRs 14:25 #link https://bugs.launchpad.net/ubuntu/?field.searchtext=&orderby=-date_last_updated&field.status%3Alist=NEW&assignee_option=none&field.assignee=&field.subscriber=ubuntu-mir 14:25 there is https://bugs.launchpad.net/ubuntu/+source/gamemode/+bug/1853830 14:25 Launchpad bug 1853830 in gamemode (Ubuntu) "[MIR] gamemode" [Undecided,New] 14:25 from the Desktop team or Wimpress 14:26 anyone having extra context on that one? 14:27 didrocks: I know you are not here - but you'll read thids ping. Could you clarify if this is up for review or waiting on anything else? 14:27 It is there quite some time for not being mentioned before, therefore a pre-check by someone close to desktop could be useful 14:27 next list 14:28 #topic Incomplete bugs / questions 14:28 #link https://bugs.launchpad.net/ubuntu/?field.searchtext=&orderby=-date_last_updated&field.status%3Alist=INCOMPLETE_WITH_RESPONSE&field.status%3Alist=INCOMPLETE_WITHOUT_RESPONSE&field.subscriber=ubuntu-mir 14:29 https://bugs.launchpad.net/ubuntu/+source/usbguard/+bug/1816548 14:29 Launchpad bug 1816548 in usbguard (Ubuntu) "[MIR] usbguard" [Undecided,Incomplete] 14:29 joeubuntu: did you (=security) and any desktop member get further on this one? 14:29 last was an update by jdstrand missing still some fixes 14:30 seb128 isn't here, didrocks could you ask internally if Desktop wants to re-emphasize or drop the efforts on usbguard? 14:31 cpaelzer It is still in progress by jdstrand 14:31 should it move from incomplete to new then? 14:31 as incomplete indicates waiting on the reporter 14:32 Let me check with jdstrand when he comes back. 14:32 ok, you have the powers to set tat state then if he agrees 14:32 yup 14:32 the last I see worth for discussion is https://bugs.launchpad.net/ubuntu/+source/openjpeg2/+bug/711061 14:33 Launchpad bug 711061 in openjpeg2 (Ubuntu) "[MIR] openjpeg2" [High,Incomplete] 14:33 had security review a) https://bugs.launchpad.net/ubuntu/+source/openjpeg2/+bug/711061/comments/62 14:33 which let doko set it to incomplete 14:33 then a lot of things happened 14:33 and we got https://bugs.launchpad.net/ubuntu/+source/openjpeg2/+bug/711061/comments/71 14:33 which is security review (b) 14:34 doko: joeubuntu: Is that ready for promotion now? 14:34 as (b) was an ack 14:34 Yes it was acked. promote 14:34 doko: or are you waiting on https://bugs.launchpad.net/ubuntu/+source/openjpeg2/+bug/711061/comments/70 ? 14:34 Launchpad bug 711061 in openjpeg2 (Ubuntu) "[MIR] openjpeg2" [High,Incomplete] 14:35 cpaelzer hold on, 14:35 holing ... :-) 14:35 +d 14:36 I may be wrong... 14:36 I also don't see desktop-packages subscribed 14:37 I thought it was done, let me check with the team and get back via email. 14:37 joeubuntu: you could check with the team and then do a small update on the bug maybe? 14:37 directly on https://bugs.launchpad.net/ubuntu/+source/openjpeg2/+bug/711061 14:37 Launchpad bug 711061 in openjpeg2 (Ubuntu) "[MIR] openjpeg2" [High,Incomplete] 14:37 will do 14:38 thanks 14:38 so I guess we unblcoked all we heard about 14:38 #topic Any other business? 14:38 we had the two big discussion in the beginning 14:38 anything else? 14:39 ... no? 14:39 3 14:39 2 14:39 1 14:39 than ks everyone 14:39 today was a rather big one, so thanks-twice! 14:39 #endmeeting