23:59:52 <nitzmahone> #startmeeting Ansible Azure Working Group
23:59:52 <zodbot> Meeting started Wed Mar 27 23:59:52 2019 UTC.
23:59:52 <zodbot> This meeting is logged and archived in a public location.
23:59:52 <zodbot> The chair is nitzmahone. Information about MeetBot at http://wiki.debian.org/MeetBot.
23:59:52 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
23:59:52 <zodbot> The meeting name has been set to 'ansible_azure_working_group'
00:00:10 <nitzmahone> #chair yungezz yuwei jborean93
00:00:10 <zodbot> Current chairs: jborean93 nitzmahone yungezz yuwei
00:00:29 <jborean93> hey
00:00:42 <yuwei> hi
00:00:45 <yungezz> I have 2 topics today, prs and facts/info thing
00:00:56 <yungezz> Hi Jordan
00:01:49 <yungezz> For prs, one of mine: https://github.com/ansible/ansible/pull/53161
00:02:32 <nitzmahone> Looks like it has a merge conflict
00:02:43 <yuwei> servicebus and workspace PR https://github.com/ansible/ansible/pull/53731  https://github.com/ansible/ansible/pull/51320
00:03:00 <yungezz> I will fix the conflict
00:03:25 <yungezz> Not sure does zim has Pr also
00:03:32 <yungezz> Let me check
00:05:38 <zikalino8265> hello :-)
00:05:54 <nitzmahone> good $time_in_your_region
00:06:15 <yungezz> Hi zim, do you have pending review prs?
00:07:03 <zikalino8265> just these i mentioned from 3rd party contributor
00:07:22 <zikalino8265> https://github.com/ansible/ansible/pull/53622
00:07:37 <zikalino8265> he mentioned he will merge my changes into his a day ago
00:07:42 <zikalino8265> but i think still not there
00:08:00 <nitzmahone> community freeze is next week, so still some time
00:08:27 <jborean93> I've got a PR that will need to be reviewed by you before it's merged
00:08:27 <jborean93> https://github.com/ansible/ansible/pull/54443
00:09:01 <jborean93> The requirements file has conflicting versions and this upgrades 3 of them so the `msrestazure` has a valid version
00:09:15 <jborean93> Unfortunately I've had to make changes to the modules but hopefully I've picked up most of them
00:09:19 <nitzmahone> yuwei: servicebus changes look good, fix merge conflict then good to merge
00:09:24 <zikalino8265> for me looks good, are ll the latest?
00:09:51 <zikalino8265> jborean93 i will check the details and merge
00:10:14 <jborean93> `azure-mgmt-batch` is not the latest as it was 2 major versions behind and I didn't want to make any big leaps
00:10:18 <jborean93> thanks
00:10:25 <nitzmahone> Yeah, not sure how we can keep that from happening in the future
00:10:43 <zikalino8265> i was also planning to look into versions before freeze so i think you did most of the work :-)
00:10:46 <nitzmahone> Other than maybe pip install the requirements file to a fresh venv and watch for errors
00:10:52 <jborean93> Will be looking at making sure we test the requirements file for these conflicts
00:11:00 <zikalino8265> i will upgrade later if anything is still behind
00:11:05 <jborean93> sounds good
00:11:29 <nitzmahone> IIRC there were a few that were still using prerelease versions because we'd have to significantly update the modules for the post-release versions
00:11:38 <jborean93> I don't know if `azure-mgmt-web` has a bug but it no longer raises `CloudError` if a resource does not exist, couldn't find any docs on this
00:11:49 <nitzmahone> But I'm guessing we're going to run into this kind of thing more and more for the ones that are lagging behind.
00:15:37 <yungezz> Any other Pr? If no more prs, then facts/info thing
00:15:51 <nitzmahone> nothing from me
00:18:36 <zikalino8265> a kind of silent :-) so that facts/info thing...
00:18:47 <yungezz> For facts to info renaming, we don’t want to change it in 2.8 as listed in the Pr, but some of new modules was merged by Bcoca already. We’d like to revert this to avoid inconsistency to users
00:18:55 <nitzmahone> agreed
00:19:13 <zikalino8265> on reverting?
00:19:22 <yungezz> Matt could you help to talk with bcoca on the history why we removed ansible_facts?
00:19:26 <nitzmahone> yeah, fine with me to keep consistency for 2.8
00:19:58 <yungezz> I think we need get his approval to do reverting right?
00:19:59 <nitzmahone> Yes, I'll make sure there's no hassle from this end
00:20:00 <zikalino8265> well, bcoca seems to not agree, just a bit worried
00:20:20 <yungezz> Thanks Matt!
00:20:20 <zikalino8265> and i still have 5 modules to merge
00:21:07 <nitzmahone> We've had this pattern for a long time in the absence of a non-_info suffix, so no reason to change it until 2.9 when we have something agreed-upon to change it to
00:21:07 <yungezz> So when every thing ok, pls help to approve the reverting Pr
00:21:08 <zikalino8265> will there be any special announcements about _info in the future?
00:21:36 <nitzmahone> Yeah, we'll update the developer guide and probably add some automated checks for new modules that ensure _facts only return _facts and that _info does not
00:21:38 <zikalino8265> i think that's what's very important for us.
00:22:00 <yungezz> That will be great, update developer guides
00:22:06 <zikalino8265> i think i like "that ensure _facts only return _facts and that _info does not" very much
00:22:08 <jborean93> it was only recently decided to use `_info` for "facts" modules that don't return `ansible_facts`, so tests and docs should be forthcoming
00:22:34 <nitzmahone> It's been floating around as an idea for awhile, but we recently decided for sure that's the plan
00:22:55 <nitzmahone> (and as part of 2.9 will figure out whatever the transition process looks like w/ aliases etc)
00:23:04 <yungezz> Got it. So we little confused at beginning
00:23:39 <nitzmahone> Which could possibly include even dynamically masking ansible_facts returns if an _info module returns them (we know how it was called, so the aliased ones would probably just work)
00:23:47 <zikalino8265> while talking about naming.....
00:24:05 <zikalino8265> shoud we decide on usage of _underscores_ in the middle of module names?
00:24:14 <nitzmahone> Yeah, sorry about that- I think I was pretty clear that we should hold off doing anything until 2.9, but sounds like not everyone else on the Ansible side was on the same page
00:24:35 <nitzmahone> We could
00:24:40 <yungezz> Ok, thank you Matt
00:25:29 <nitzmahone> So would you want to do underscores everywhere?
00:25:50 <nitzmahone> (if we had the corporate DeLorean and could go back in time to change it?)
00:26:04 <zikalino8265> yes, yuwei already started adding _ in some modules, but i am not sure yet
00:26:05 <yungezz> :)
00:26:22 <zikalino8265> we have some inconsistencies
00:26:26 * nitzmahone looks through current modules for "the worst offender" if we did it that way
00:26:43 <zikalino8265> for instance
00:26:50 <zikalino8265> azure_rm_virtualmachine_extension
00:26:54 <nitzmahone> azure_rm_application_security_group_facts
00:27:03 <zikalino8265> and azure_rm_virtualmachineextension_facts
00:27:13 <zikalino8265> i created _facts to match other modules
00:27:27 <zikalino8265> but it doesn't match azure_rm_virtualmachine_extension
00:27:44 <zikalino8265> i was thinking of renaming azure_rm_virtualmachine_extension to match
00:28:19 <yungezz> We can remove extra underscore
00:28:27 <yungezz> To keep consistency
00:28:32 <nitzmahone> If we want to stay consistent with the majority, it should be azure_rm_(thingymabobwithnounderscoresanywhere)(_facts|_info)?
00:28:41 <yungezz> Agree
00:28:49 <zikalino8265> yes, that's what i want to do. to test rname/obsolete process
00:29:00 <nitzmahone> OK
00:29:26 <zikalino8265> so do you think we should also change **azure_rm_aks_version**
00:29:29 <nitzmahone> So don't do it yet, since we don't know exactly what we're going to do there yet, but once we figure it out, including that one in the pile should be fine
00:29:49 <nitzmahone> it's inconsistent, so probably
00:29:50 <yuwei> so I need to change the servicebus one
00:29:56 <zikalino8265> you mean about nunderscores or renaming?
00:30:21 <zikalino8265> and i think azure_rm_aks_version is also facts, so it should be azure_rm_aksversion_facts
00:30:28 <yungezz> Underscore for service bus
00:31:23 <yuwei> like azure_rm_servicebusqueue
00:31:29 <nitzmahone> yeah
00:32:00 <nitzmahone> If we're going to clean the others up to no underscores except main Azure prefix and _fact/_info, they probably should be renamed before we merge them
00:32:10 <nitzmahone> (servicebus)
00:32:18 <yuwei> Yes
00:32:52 <zikalino8265> ok, good, one more consistency issue settled down. makes me happy :-)
00:33:40 <yungezz> No more topic from me
00:34:00 <nitzmahone> Cool, nothing else here either. Thanks all!
00:34:03 <nitzmahone> closing in 5..
00:34:06 <nitzmahone> 4..
00:34:09 <nitzmahone> 3..
00:34:12 <nitzmahone> 2..
00:34:15 <nitzmahone> 1..
00:34:17 <yungezz> Thanks everyone!
00:34:18 <nitzmahone> #endmeeting