DLN Community Project - Quality Control Platform

Meeting is in 40 minutes.

Below this post will be instructions to join chat until an hour in,

It’s not intended to be a public walk-in, it’s a distraction-free place for people showing interest in the project to collaborate. If you can’t make the meeting the proposals will go straight to the forum so no one misses out.

Remember to stay humble and give priority to devs/sysadmins who can pull this thing together.

Hope to see you there :slight_smile:

I think Mumble was acting up for me when I joined, sorry about that!

Just wanted to say was great meeting you all and excited seeing everyone is so pragmatic :smiley: Looks like this should be a good little endeavor!

@Ulfnic I’ll get the notes from the meeting cleaned up, as a newbie in the community, just let me know where best to share/upload them.

@kobberholm looking forward to seeing what you come up with in the mean-time :+1:

2 Likes

I’m planning to do a meeting directory post so there’s a central hub where people can link back and forth between the various topics and their proposals. That way it’s not a monolithic mega thread of doom.

I’ll send you my email or you can PM them to me. This’d save me a lot of time and I can combine them with the ones I have plus anyone else contributing.

Thank you @goldfish92

Was impressed by the quality of everyone who joined, what a start.

Is everyone happy with the meeting day/time?

Keeping things EU/US compatible it needs to be Saturday or Sunday roughly around that time but it can be moved a few hours.

Worked quite nicely for me. Did run into my 7Days with friends time though at the end… almost missed blood moon! :joy: Maybe setting an agenda before hand next time would be useful?

So,
I know there was some discussion around securing the service(s) in yesterdays meeting and I know we didn’t agree to any concrete plans but ;-)…
Instead of twiddling my thumbs until the next meeting I’m happy to try and make a proof-of-concept between now and then. Might be interesting and even if it is throw away hopefully it will be fun and help discover some things.

Proposed architecture attached. With a User Account and Authentication (UAA) server and an API Gateway implemented we would be able to integrate any front-end or back-end services with accounts and permissions using standard protocols which shouldn’t restrict our framework choices going forwards.

I welcome comments and feedback :slight_smile:

Auth Architecture

P.S. Can’t wait for the GitLab for sharing these kinds of things more easily :wink:

3 Likes

1 hour earlier would be better for me and it won’t be as late for the EU side. Sat vs Sun doesn’t matter either way.

Proposed time: 2020-09-05T17:00:00Z

Anyone have conflicts?

2 Likes

NEXT MEETING:
2020-09-05T17:00:00Z (40 minutes)

40 minutes of pure action items in quick succession for what’s been established and discussion on how we proceed, critical items in the first 20 minutes. I’ll keep the time and announce the stages. Feel free to join for the first 20 or the full 40.

After Meeting Hangout:
General discussion, future ideas, ect.

Location && Instructions to Join:
Same place. For anyone new please add your information to the sign up form and PM me for details.

edit: changed to new link

Open Sign Up: https://devops.destinationlinux.network/index.php/s/mMcJDzpbRECoT9W

Itinerary:
It’ll be informed mostly by how forum discussion plays out, edit in what you’d like to see in the below itinerary link, post here or on the individual discussion threads and i’ll centralize it closer to the meeting time.

Open Itinerary: CryptPad

Big thank you to Michael for covering the QC platform on This Week In Linux:

Full meeting notes are here for project discussion:

CryptPad

All thanks to @goldfish92, they’re high quality and he didn’t even have a meeting recording, saved me a lot of time. That’s some kind of special skill.

I was going to split everything up into individual threads but after listening through and doing the first two I realized i’d be nuking the forum with QCP proposals. I think it should be meeting notes for contributors to read and discuss with 1-2 threads for topics that full community discussion would really help.

For discussion topics check out goldfish’s work and I put out an open topics document for next meeting above.

If i’d pick a proposal that really needs highlighting its:

Choosing a backend framework:

The person with the most expertise/time for our backend framework is @kobberholm who’s proposing PHP Laravel and even possibly producing a proof of concept this week which is pretty stellar.

No one so far has expressed a similar skillet though meaning most of the stack would be riding on a single person. I’m hoping there’s someone who can join kobber with a similar skillset because presently no one will be able to understand the backend.

A good alternative is NodeJS but no one’s near kobber’s expertise/time mix. If we had another person strong in PHP Laravel or NodeJS I think we could call an affirmative on the choice because there’s enough of us to compliment a NodeJS backend dev.

Calling For:

A PHP Laravel or NodeJS backend developer with good database skills.

For PHP Laravel they’d be working with @kobberholm
For NodeJS they’d need to take initiative on all things backend and bring our other Javascript dev(s) in on development.

1 Like

Thank you to @ITGuyEric for getting word out about the project on the Sudo Show:

1 Like

Meeting is tomorrow: 2020-09-05T17:00:00Z (20min critical items + 20 minutes stay on for discussion)

Please check out the itinerary and add what you feel would be important to cover.

Open Itinerary: CryptPad

The pace will be a lot quicker. Be thinking about your call to action for what we should do right now.

Of course!

Once the project hits a big milestone, we’ll have to have you on the show to tell us about it.

If that happens i’ll have the journey of the wonderful people who pulled this together to share.

Thanks again @ITGuyEric

1 Like

NEXT MEETING:

Same day/time.
2020-09-12T17:00:00Z (40 minutes)

Open Itinerary: https://devops.destinationlinux.network/index.php/s/TxJboiesLqjiNyz

Location && Instructions to Join:
Same place unless otherwise mentioned below (there needs to be testing of NextCloud’s VOIP). For anyone new please add your information to the sign up form and PM me for details.

Open Sign Up: https://devops.destinationlinux.network/index.php/s/e5QnZ9Wo76JaKEz

Once again @goldfish92 knocked it out of the park with the meeting minutes while simultaneously taking full part in the discussion.

We’re entering the initial GitLab stages (public shortly), accounts for our NextCloud are out to everyone and we’re discussing finer tooling and near term objectives.

If you’re thinking about joining we’re wide open.

As always meeting notes are open for discussion by everyone. Feel free to read through and address or re-address any topic mentioned. Comment here or create your own discussion thread, love to hear your thoughts.

https://devops.destinationlinux.network/index.php/s/xiB5KKdQ5LSgAKZ

1 Like

As per moving communications to NextCloud, it seems we may need to install the Talk app to enable this. More admin work I’m afraid…

@esbeeb @letter I somehow missed your very important discussion about who the testers are and what they are “allowed” to test.

My current way of thinking is that anyone can test anything, so even a single developer making a web page for his family, can give them a link to his test page and tell them to perform the test tasks and fill out the form.

This is different from who/what we RECOMMEND for testing.
Just as Canonical can register their organization and create projects for testing, I am planning a tester organization feature.
This will be for curated test campaigns, where a group of people is encouraged to test the same software during the same period and performing the same tasks.

Who is running these tester groups and what they will test is up to the community, but some possible examples could be:

  • Newbie Friendly Linux Distros
    For people new to Linux, where they perform easy tasks that they would likely already know how to do on Windows or Mac.
  • Graphics and Media Creation
    Graphics designers and photographers review photo management and editing software.
  • Pixel Peepers United
    Theme reviews.
  • Sight Impaired Reviews
    Not everyone sees equally well and software needs to be tested for that.
  • Big Daddy Linux Weekly Distro Test
    @BigDaddyLinux could organize his distro review in a group

These groups can make it a social activity and provide guidance for new testers.

While you can search and organize groups and campaigns in any way you like, I imagine a default recommendation based on activity and the satisfaction of previous reviewers of the group.

Just to explain that last bit:
When you complete a test, there will be a quick review process, where you can rate how well the test was handled by the manager. This score will be used to rank new campaigns run by that organization in the future. This will be the case both for tester organizations and developer organizations.

The key point is that in order to get users engaged, they need to be able to group up with like-minded testers and test things they actually care and know about.
Maybe there will even be a feature where a developer organization can request review by a specific tester organization, but that for now, that request will simply be a tag the developer puts on his campaign and that searchers can filter their results based on.

1 Like

I’ve been expressing concerns over gatekeeping because if anyone can put up a test it’ll quickly encourage people to generate testing fluff to pull traffic to their platform. Some will be obvious but it’d be easy to make tests appear legitimate. This could effectively bury that single developer you’re talking about using the platform for the fist time.

Love the ranking proposal you mentioned along with organization accounts, excellent way to start mitigating that. I’d also recommend…

  • Posting limits gauged by
    • Non-organization account verification levels
    • Positive feedback
    • Additional factors…
  • Have testing creation forms encourage high quality tests over speed of deployment to make it harder for fluff posters to appear genuine.
  • As any system can be gamed and rankings can be motivated by external factors like a company making a bad PR move unrelated to the test… Prominence should be a mix of as many social and non-social metrics as we can include.
    • Things that show “proof-of-work” on behalf of the test creator is a good example.
    • How many questions test creators have left hanging.
    • Valuing social feedback by the prominence of the tester to some degree can also help.

Love the idea about tester organizations.

Thankfully this’ll be a wonderful problem to have because it’ll require a finished production run and a popular platform. :slight_smile: Good to think ahead either way.

1 Like