20:00:01 <jborean93> #startmeeting Ansible Windows Working Group
20:00:01 <zodbot> Meeting started Tue Dec  4 20:00:01 2018 UTC.
20:00:01 <zodbot> This meeting is logged and archived in a public location.
20:00:01 <zodbot> The chair is jborean93. Information about MeetBot at http://wiki.debian.org/MeetBot.
20:00:01 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
20:00:01 <zodbot> The meeting name has been set to 'ansible_windows_working_group'
20:00:05 <jborean93> hey all
20:00:11 <jhawkesworth_> hey
20:00:20 <jborean93> #chair jhawkesworth_
20:00:20 <zodbot> Current chairs: jborean93 jhawkesworth_
20:00:34 <it-praktyk> hey
20:00:37 <jborean93> Matt has an appointment right now so it will just be me
20:00:38 <jborean93> hey
20:00:44 <jborean93> #chair it-praktyk
20:00:44 <zodbot> Current chairs: it-praktyk jborean93 jhawkesworth_
20:00:49 <jhawkesworth_> oh ok
20:01:06 <jborean93> give it a few more minutes, see if dag will be joining
20:01:27 <jhawkesworth_> sure
20:03:50 * jhawkesworth_ cheekily adds something to agenda
20:03:57 * dag noticed it
20:04:06 <jborean93> #chair dag
20:04:06 <zodbot> Current chairs: dag it-praktyk jborean93 jhawkesworth_
20:04:08 <jhawkesworth_> https://github.com/ansible/community/issues/294
20:04:16 <jhawkesworth_> in case anyone needs the link
20:04:20 * gundalow waves
20:04:28 <jborean93> hey
20:04:30 <jhawkesworth_> hey gundalow
20:04:30 <jborean93> #chair gundalow
20:04:30 <zodbot> Current chairs: dag gundalow it-praktyk jborean93 jhawkesworth_
20:04:37 <jborean93> cool, let's get started
20:04:47 <jborean93> #topic https://github.com/ansible/community/issues/294#issuecomment-442096828
20:04:58 <jborean93> thanks it-praktyk for joining us
20:06:00 <jborean93> appreciate you giving us feedback on the review process, it helps us to identify things we've done wrong and we can improve in the future. I'm sorry about how things have worked out so far.
20:06:13 <jborean93> Is there anything you wish to talk about on this topic?
20:06:57 <it-praktyk> I pushed the last commit to that module https://github.com/ansible/ansible/pull/48828/commits/2d687ec6a691d8d71358ae447e0fdc828976a767
20:08:09 <it-praktyk> I would like to see some kind of development rules
20:09:36 <dag> it-praktyk: I am wondering why you no longer want to maintain that module ?
20:10:08 <dag> if I take a step back, the module has improved
20:10:21 <dag> without your input it wouldn't have existed in the first place
20:10:34 <jhawkesworth_> There are some instructions here: https://docs.ansible.com/ansible/latest/community/development_process.html but may be they don't cover the kind of rules you are thinking of?
20:10:59 <dag> if someone would have offered you the PR that would have resulted in this resulting module, wouldn't that be any different ?
20:11:28 <dag> I understand some of the frustration, do not get me wrong
20:11:41 <gundalow> it-praktyk: Hi, I look after the Contributor Experience for Ansible, would would you like to see in the "development rules"
20:14:08 <it-praktyk> I don't know exactly what these rules should be. I'm new here, I learned some things recently about technical aspects of development
20:14:10 <jhawkesworth_> There is also a code of conduct here https://docs.ansible.com/ansible/latest/community/code_of_conduct.html#code-of-conduct
20:14:25 <it-praktyk> but also that some new thins are incoming and they are undocumented
20:14:31 <gundalow> it-praktyk: sure, wondering what type of things, rather than the details
20:14:38 <gundalow> Happy to document them :)
20:16:34 <it-praktyk> I'm glad that you agreed (?) that the situation was not OK. And I hope that you/we can make healthier for the future.
20:17:40 <it-praktyk> Sorry for my grammar. I hope that you understood. If not please ask.
20:17:55 <dag> So I think we have to be careful to replace lots of code in a contributor's PR
20:18:27 <dag> but in this case I don't know what the alternative would be, accepting the PR, and create new PR's to improve/fix large parts ?
20:18:50 <it-praktyk> guiding and reviewing is not equal replacing
20:18:53 <dag> that would be a lot more effort
20:18:57 <dag> it-praktyk: I agree
20:19:15 <dag> it-praktyk: the goal of the review process is definitely to learn new things and grow in the process
20:19:40 <it-praktyk> to be honest from my (emotional) point of view it would be better
20:19:47 <dag> it-praktyk: and I think that some effort was spend in the process to achieve this
20:20:15 <dag> I understand very well
20:20:35 <it-praktyk> we are contributing because we would like to achieve some satisfaction, and replacing almost all my work at that stage was not neutral for me
20:21:13 <it-praktyk> so please correct copyrights for the win_psmodule (not with my name) and merge it
20:21:27 <it-praktyk> I hope that everyone learned something
20:21:53 <it-praktyk> Hmmm, I still repeat 'hope' ;-)
20:21:53 <dag> it-praktyk: even in this case, you can still claim authorship
20:22:11 <dag> it-praktyk: the resulting module was built on top of your work
20:22:18 <it-praktyk> but I don't have to as a part of my 'protest'
20:22:26 <dag> I understand
20:23:07 <dag> in itself, the additional commit would still give you credit for the PR
20:23:16 <dag> so your contribution is valued
20:23:31 <dag> I have had my PR being closed and replaced by someone else's PR :-)
20:23:53 <it-praktyk> https://github.com/ansible/ansible/pull/46516 is waiting for review
20:24:25 <it-praktyk> if does is something what I can improve (I know at least one things) please let me know
20:24:30 <it-praktyk> I'll try
20:25:02 <dag> gundalow: do we have sufficient feedback to improve the guidelines ?
20:25:03 <it-praktyk> btw, what is it Ansible.Basic mentioned https://github.com/ansible/community/issues/294#issuecomment-444238409
20:25:36 <jborean93> it's a new wrapper for interfacing with Ansible to replace `Ansible.ModuleUtils.Legacy`
20:25:53 <jborean93> it's to align some of the standards and checks that are available in the Python space provided by basic.py
20:26:06 <gundalow> dag: I think so, just not sure where it would go yet
20:26:10 <jborean93> was only recently added in devel for the 2.8 release
20:26:18 <gundalow> it-praktyk: Thank you for keeping us honest, I do welcome the feedback
20:26:53 <dag> gundalow: I understand, and I don't want to forbid adding commits to existing PRs (especially for smaller PRs or cosmetic changes)
20:27:30 <it-praktyk> I understand a case of cosmetic changes
20:27:37 <dag> gundalow: Thanks for taking care of this, I bet this happens more often than we know :-)
20:28:09 <gundalow> dag: yup, that's possible
20:28:35 <jhawkesworth_> We talked a bit about adding commits to PRs at Contributor Summit earlier in the year.  Personally I am not comfortable with doing this to others' PRs but I accept sometimes it is necessary
20:29:03 <jhawkesworth_> I am happy for others to modify my PRs
20:30:00 <dag> it-praktyk: we often get drive-by PRs (through the website or GitHub edits) and having back-and-forth review is often less productive (and more frustrating to the contributor) than simply merging it with the needed modifications
20:30:25 <dag> so it really is a thin thread to walk on, do we help the contributor, or not
20:30:56 <jhawkesworth_> also sometimes there is a pending release so not allways time to check with contributor
20:31:05 <dag> might also relate to culture, in fact
20:31:13 <gundalow> My personal view is making commits then merging (ie updating `version_added`, or fixing docs typo) is fine to update then merge
20:31:37 <gundalow> urgh, words
20:31:54 <jhawkesworth_> I think it may relate to culture too
20:31:59 <dag> gundalow: indeed, but you have to wait for CI to validate, so there's a window of opportunity for miscommunication ;-)
20:32:24 <dag> anyway, shall we move to jhawkesworth's agenda point ? :)
20:32:30 <it-praktyk> yes
20:33:14 <jhawkesworth_> I'm not aware of any open issues with Ansible.Basic now.
20:33:18 <jborean93> #topic https://github.com/ansible/community/issues/294#issuecomment-444238409
20:33:37 <jborean93> I'm probably -1 on porting them all right now
20:34:00 <jhawkesworth_> my fear is 2.8-devel is not going to be getting much use yet
20:34:01 <jborean93> It's a massive change and while we've fixed the bugs that have appeared I am wary of moving them all across without it being tested further
20:34:22 <jborean93> the other issue is that it is slightly slower than Legacy due to the compilation of the C# code
20:34:39 <gundalow> So is your thought to do the remainder after `stable-2.8` is branched?
20:34:48 <gundalow> Do we want to get more people testing `devel`?
20:34:58 <dag> jborean93: moving the modules at this point would be an excellent test for unknown issues
20:35:08 <dag> jborean93: especially at this stage in the release cycle
20:35:52 <jborean93> it would, the performance issues is still a concern. I know nitzmahone was wanting to look at persisted connections to try and overcome any hit we may get but that definitely won't be in for 2.8
20:37:30 <jborean93> just as a side note for people who want to learn more about Ansible.Basic, https://docs.ansible.com/ansible/devel/dev_guide/developing_modules_general_windows.html#windows-new-module-development
20:37:41 <dag> gundalow: I convinced it-praktyk to stay on as the maintainer of the module, but it will cost you an Ansible t-shirt :-D
20:37:56 <jhawkesworth_> worth it!
20:38:05 <dag> agreed! :-D
20:39:00 <dag> jborean93: would anyone notice ? Is it an issue that scales, or just a fixed overhead ? (guess the latter)
20:39:05 <jhawkesworth_> I guess if this is going to stretch over 2 releases potentially, it might be worth picking which modules we'd like to move before and after 2.8
20:39:24 <jhawkesworth_> I struggle to get any meaningful benchmarks when testing ansible changes.  I don't know why but I see very varied runtimes for the same playbook.
20:39:24 <jborean93> it's about .2-3 seconds slower. This can definitely add up over time when running lots of tasks
20:39:36 <jhawkesworth_> i guess you benched locally?
20:39:47 <jborean93> jhawkesworth_: the best way to test it out end to end is to use basic auth over http
20:39:49 <dag> jborean93: and PSRP doesn't compensate for it ?
20:40:03 <jborean93> still a bit variable but it mostly stays consistent
20:40:23 <jborean93> dag: it may, I need to test out the timings based on the recent changes to the exec wrapper process in the latest release
20:40:27 <jborean93> that may have slowed it down :(
20:41:31 <jhawkesworth_> I appreciate the persisted connection stuff is going to take as long as it takes to do, but wondering what is planned ahead of it
20:41:43 * jhawkesworth_ goes to find roadmap so I can answer the above myself
20:41:50 <it-praktyk> dag: A bribe is not needed
20:43:13 <jborean93> I'm ok with moving them across to Ansible.Basic but still concerned about the performance. I don't want to make this decision without nitzmahone's input which would probably need to wait until next week
20:43:23 <jhawkesworth_> sure.
20:43:50 <jhawkesworth_> I'm thinking some of the lesser used modules would be safe to move over.
20:44:26 <jhawkesworth_> and some which maybe are more likely to have missing/wrong args when used, although that's hard to guess
20:44:42 <jborean93> or modules that have a variable time, like win_package or so
20:45:07 <jhawkesworth_> yeah.  but maybe avoid most used modules like win_file, win_copy
20:45:12 <jborean93> if you wish to move a module you use or want to try out I won't say no to the PR. I'm just wary of opening the floodgates, especially if it doesn't have tests
20:45:14 <jborean93> yep
20:46:02 <dag> it-praktyk: it's not a bribe ;-)
20:46:11 <jhawkesworth_> Good point.  If it doesn't have tests (and it can) then maybe a tests PR ought to come first, if people have the time
20:46:37 <jborean93> sounds like a good plan
20:46:44 <dag> indeed, if there are no tests (or incomplete tests) we should not do it
20:46:57 <jhawkesworth_> sounds like we are in agreement
20:47:34 <it-praktyk> do we have a list of modules without tests?
20:48:15 <it-praktyk> IMHO, it could be helpful to publish and ask a community for help
20:48:42 <dag> it-praktyk: we did have it in the Wiki: https://github.com/ansible/community/wiki/Windows%3A-progress-tracker
20:48:47 <jhawkesworth_> it-praktyk: I think we have such a list on the wiki, let me see if it is up to date
20:49:30 <gundalow> https://github.com/ansible/community/wiki/Windows%3A-progress-tracker#missing-integration-tests
20:49:39 <dag> jhawkesworth_: it's fairly up to date, I think I reviewed it a few months ago
20:49:50 <jhawkesworth_> thanks dag
20:50:12 * jborean93 thought the list was larger than that
20:50:24 <it-praktyk> Can it be linked at https://github.com/ansible/community/tree/master/group-windows ?
20:50:43 <it-praktyk> I like how OpenStack communicates: https://governance.openstack.org/tc/reference/help-most-needed.html
20:50:57 <dag> jborean93: no, I removed everything we fixed (mostly your rewrites of various modules :-))
20:53:44 <jborean93> it-praktyk: I think the current page is mostly based on the work that dag has done. Not sure if there is a standard to following for the working group pages but sounds like there should be
20:54:23 <jhawkesworth_> iirc dag set that up and other WG pages then got created in a similar way.
20:54:26 <jhawkesworth_> I could be wrong
20:54:51 <dag> it-praktyk: You are free to make improvements on the Wiki :)
20:55:09 <dag> I just started to collect this information, but we never publicized it more than just listing it
20:55:12 <jhawkesworth_> I'll ad the link now
20:55:17 <jhawkesworth_> s/ad/add
20:55:24 <dag> so yes, we should definitely learn from other WGs
20:56:36 <jhawkesworth_> added direct link anyway
20:56:45 <dag> (on a side-note, someone at Ansible Automates Antwerp asked me about Ansible support for BitLocker, he was interested in contributing it)
20:57:14 <jhawkesworth_> not something I use but sounds like it would be useful to people
20:57:42 <jborean93> I just worked on LAPS and created a role, I'm sure BitLocker has deep hooks into AD just like that. Happy to see what they have
20:57:47 <gundalow> Everybody has Wiki edit powers
20:58:01 <jborean93> everyone or just contributors?
20:58:59 <dag> I think everybody
20:59:08 <dag> or at least that's how it used to be
20:59:30 <jborean93> cool, I wasn't sure if it was fully open or just editable to people who have contributed but the former sounds good
20:59:30 <gundalow> everybody
20:59:42 <jhawkesworth_> nice
21:00:19 <jborean93> anything else we wish to discuss? Getting near the end
21:00:47 <jhawkesworth_> nothing from me
21:01:44 <jborean93> ok will close the meeting
21:01:48 <jborean93> #endmeeting