CHAOSS Weekly Hangout, 18 Jul 2023

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: CHAOSS Weekly Community Call July 18 2023

Description

Meeting minutes are here: https://docs.google.com/document/d/1PMDWc6xMe0fNE7shxTK5_HE_ykRBG5w55_Zx5hvzsEY/edit?pli=1

Meeting summary is here: https://chaoss.discourse.group/t/chaoss-weekly-community-meeting-july-18-2023/213

A

Foreign everybody, it's July, 18th and you're here at the chaos weekly, Community call, so I hope everybody's doing really really well. Today, um it's good to be back. My voice is coming back a little bit, I kind of lost it after being a posse for a week and just talking at people literally for like a week straight because I don't really talk to people here when I live by myself. I just talk to you all, that's that's it so like, and you know, whenever I have zooms a selling interaction I get.

A

So let me uh share.

A

Where did the share screen? Go? Yes, oh that was my dog yeah. Sorry she's having an issue my.

B

Dogs are a mess today, my weather.

A

Dog's at the vet getting a procedure done in a.

B

Tumor removed she's packing up along.

A

Here, okay, anyway, we're professional, it's okay! So if you have not, if you have not told us your favorite way to eat potatoes, um this is the the correct answer. Actually, thank you Sophia. That's the right answer, yeah! That's from a movie!

A

If you, if you have not seen Lord of the Rings, then you should but you'll need like nine hours to just set aside of your life, because you, you know or more, if you want to do the pre The Hobbit, all that, but mostly just you know, the Lord of the Rings is the best I like mine, baked with garlic salts.

A

So there's a chat over.

C

Here: yeah, oh yeah, yeah, yeah, I'm, not gonna. Try.

D

To stick with that yeah.

A

um Okay, so we can hop right to it um if anybody comes, if somebody can keep dropping the minutes in the chat. That'd be great.

A

um The first item is just um one to give a quick wrap up on bossy and see uh how it was for everybody. Here's a picture that we took um sorry Dawn wherever you are I, know Don's off this month, but Don had her eyes closed, so we're hoping that, but there's another another better one floating around somewhere. We don't know for sure but um yeah. Hopefully there is um yeah, but it was great. We we saw so many new people that had not ever heard of chaos before so that was amazing.

A

um Of course, we saw some old friends that had heard of chaos and it seemed to be on a lot of people's minds um and then, of course, all of the all the folks here.

A

The chaotics that are in that picture were also giving talks and interacting so I feel, like chaos was kind of Everywhere by the end, I gave my um Justin Flory who's in that picture as well and I gave our talk on Sunday morning, which was the last day of the conference and um I asked the question to you know a pretty full room. I asked the question and who had heard of chaos before, and almost everyone had raised their hand by that point. In.

B

The conference.

A

So I feel like that was a success. To kind of do some Outreach do some education and bring some new folks in to the community, so that was really really good. um Sophia. Do you have anything to add to that I think you're, the only one on the call that was there.

E

I mean I, think a lot of people are interested in metrics and I think just learning about our project and that there's a whole group of people that are actively thinking about this talking about it, trying to create guides for others that are pursuing. It I think that I don't know I just like it felt exciting to see how much interest there was and just like. Even if you don't end up joining the project, knowing that the project exists and that we could be a reference or resource to you, I think is just.

E

It was really exciting to see that um and just kind of also nice, because our project is so applicable to the entire Foss Community, which I think is really fun, especially because not all projects can can make that claim and I think this kind of solidified that for me, seeing chaos in this space versus an eloff event where we were very much providing something that was valuable to a large number of people, so great job. Everyone I.

A

Would tend to agree and um I I feel like we'll, probably do this conference again. This was the first year for Fosse, so it was a smallish attendee pool of about a little less than 300.

A

um and we did have 27 folks sign up for the Lego Globe. So we had about 10 of the conference. I guess and you could say that um that that actually entered the raffle and filled out our form told us about their metrics. So that was very, very cool. Congratulations to Sasha Reed, wherever you are um I'm winning that big Lego globe and getting that home. I'm glad I.

B

A

Have to do that because that was a yeah that was fun um but yeah. It was really I think it was really a success. We didn't really have any any goals other than just doing some Outreach and chatting with folks. So um I think we we met our non-existing goals, whatever those who are we did. um We did have a good time and and um got a lot accomplished, I think at that conference, um a few tweaks for next time we are gonna, maybe get some smaller stickers printed.

A

um People liked the size of our stickers, which were I think four four inches by one, but you know laptop space is precious and so uh it's Prime real estate. So um we had been asked by a few people if we could have smaller ones. So we will also get some smaller ones printed as well.

D

Well, we should have some bigger ones for people who are truly committed as well. That's.

A

Right somebody was telling me I, don't remember who it was somebody from some um community and I. Don't want to I. Don't remember who it was so I, don't want to say something wrong, but.

C

A

They ordered their stickers, they came out bumpers sticker size, they didn't realize so they had like giant ones so I feel like. We should also you're right Sean. We should have some bumper sticker chaos, just like splattered everywhere that'd be great that'd, be great yeah.

A

um We also did not really have anything. Besides our form for people to fill out. We didn't really have any convenient QR codes for them to just click and find specific information on our website. So I think we're gonna print off. If we don't print off stacks of Flyers, we might just print off a few different things and just have them there at the table, so people can can find um because other the only thing we really had was the sign, which was awesome and perfect.

A

Perfect size, perfect everything, um but it just had the website listed. So um we were pointing people to that, but it would be better to have some QR codes with like links straight to the metrics or straight to Auger or grimoire lab or wherever we want to send people so we'll figure that out um and then we it would be nice to have a tablecloth with with the chaos logo. We had the gnome um folks were right here, and so theirs was really easy to see them.

A

We just had gray, so you would see this sign, but maybe it didn't pop out at you enough. So um maybe we'll get a big tablecloth too just for next time and then Justin said that when they do Fedora shows they have an actually like a.

B

A

There, with a slideshow showing like some of their folks some of their events and activities, so that would be really cool if we could eventually do that, um not maybe for all things open, which is our next conference that we're gonna have this table at, but maybe in the future we could have. That could be a project that the design team could work on or whoever wants to work on.

A

um Just have it like looping, while we're, while we're sitting at the table and chatting with people I think that would be super cool, but a little bit of logistics there, and maybe some costs I, don't know where we would get a monitor. I, don't know how that works. So um yeah shawna's thing and chat. The tagline for his family is chaos on North Lake yeah. You got to get them to get that extra s on there, Sean for sure.

D

Yeah I'm really software person in The Brood, so they're.

A

Not gonna get it successful.

D

A

You could just hand write it in like just wherever it's.

D

Yeah I may just do that. I may just take my t-shirt and add an ass. That's.

A

It that's it um so, does anybody have questions about Fosse or kind of experience there or have suggestions on what how we might make that better in the future.

D

I'm sorry I wasn't able to go because it sounds like it was fantastic.

A

It really was, it was extremely well run, I think for especially for a first year conference. Oh my gosh, like extremely well run.

A

So yeah I would really like to participate again, if possible, um maybe some formal, more formal partnership with them next year between chaos and and those folks so we'll see, but.

B

E

A

A

um So the next thing on our list is just want to remind everybody. We are having a badger orientation tomorrow at 12 noon: U.S Central, Chicago time um there is a calendar, it should be on the calendar. If you would like an invite, I'm happy to send that to you or you can just show up, you don't have to register or.

C

A

That, but if you are new to chaos, it's a great way, you can contribute um essentially you're, just double checking things that an event organizer is telling us and we use those Dei metrics to um for that application. So they'll have to tell us how they're attending to different metrics that we've developed in the Dei working group and and again all you I should say all you need to do because I mean you mean you do have to you know, know how GitHub works and um be able to to get to that application.

A

But essentially the essence of it is you just are verifying the information, so you do not have to be a def expert. You don't have to be a chaos expert. You can just um it's a great way to contribute for new folks, so show up at this orientation. If you like um it's an hour long, you can then decide whether or not you want to actually be an event Badger. It's no obligation, no commitment whatsoever at all um yeah.

A

It will take about maybe 15 to 30 minutes uh of your time a month if, if maybe maybe a little more, depending on how many applications we receive, but we try to really spread them out for folks, so that not everybody's, you know, inundated and has a very large time commitment s on that.

A

Okay, fair enough, um the next thing I wanted to bring up was I know. A few weeks ago we had talked about our governance stock I. Think it's down here, yep and I. Don't I, don't think I don't know Kevin. Did we get this on the website? I know I haven't done anything with it, so I'm, basically I'm just asking I'm if I'm holding up the show or if I need to do something.

F

It's uh it's not on the website, yet uh the knob just pushed a uh a pretty big uh pull requests that kind of that restructured. The uh the knowledge base right so I have an action item to uh to move that restructure to the website, so that because they're not matching currently, uh that action item is to happen before the next uh before that next knowledge base meeting, so so I haven't quite gotten to it yet, but it should be done within the week.

F

Okay, at that point, uh I'm not sure is the government stock currently in the community repo.

A

Not sure and I don't.

D

It was developed on Google Docs. So to my knowledge, it is not in a repo, but if you point me at where it should go on a repo I'm happy to put it into one I.

A

Think well, we can just check here really quick.

F

It should be the.

A

F

Folder inside the community, repo.

A

This is that right, yeah, that's right! Okay,.

D

You give me an action item to put that in a repo.

A

Oh, it's already here: Sean we're good okay,.

D

F

Yeah, so the the action so the so actually the uh the person holding that back is me, so I have the action item to uh uh to link the uh the website to that new structure. So uh that should happen within the week. Okay,.

A

F

Will be on the website.

A

If you need me to do something with that, if you get busy or whatever just shout at me and I will be happy to create a page and Link, it.

F

So and then the uh we do, the the general knowledge item that we have for the uh the the group. That's that's meeting on, so we meet on Wednesday every other Wednesday at 8 30., let's see, uh what's the community handbook working group, currently uh the action item that we have right now now that we have a a base structure for the knowledge base is to is to start sorting through the documents that we have to see.

F

What's missing, what documents maybe need to be created and and which documents that we're using need to be edited to better reflect uh chaos. So if anyone on the call is interested in doing some Community handbook, work you're welcome to uh join us on next Wednesday, not this coming Wednesday uh for that meeting and uh maybe take a take a moment to to peek at the community handbook and and come with feedback.

A

um Can do you think we can post that because there is a um a technical writer kind of subgroup under the chaos Africa umbrella? Do you think we could post that in there as well and maybe just um see if any of those books will be available? Sure awesome I think it's uh yes,.

C

Tech writers.

A

Thank you, Ruth sure, be honest.

A

um I think it's been a little quiet in there, so I'm sure they would probably love to have something to work on. So that would be great how.

F

About if I I'll send them an invitation after I uh after.

A

I make that change on the website. Perfect awesome thanks, Kevin.

A

Anything else with uh governance, stock or knowledge based stuff from anybody.

A

Okie dokie, um the last thing I wanted to bring up was something that I heard about at Fosse called allcontributors.org and I I feel like maybe we have talked about this um here but I know I know: we've been talking about Sean you're, breaking up.

D

Distributors, but not all.

A

Yeah, so um we have talked in the past about how we recognize and surface and kind of keep track of these non-code contributions, um and this was I was something that I heard about at Posse and it's a bot that goes in GitHub and it essentially lets you kind of just give credit where credit is due easily. So um you would just say this: please add this person for design and then it opens a pull request and then it just kind of gets counted.

A

So I do I, agree, Ruth, I think it is really cool and I just wanted to I mean it's not perfect, because people will still need to have a GitHub account.

A

So, if someone's contributing on something that doesn't even touch GitHub, they will still have to have an account, um and it will also like if we're running overall numbers I, don't know if there's a way to pull those out or if it's just going to get counted as a regular old, same old, same old code contributions, I'm, not sure about that piece, but um I I was going to look into it a little more but I wanted to kind of just show the community this and to see.

A

If it's something that they would, you think we should add to our repositories or try out or what do you all think about it?

A

Here's the link, I'll drop it in chat.

F

I've tried I've, tried, maintaining contributor documents in the past and it is, uh it is a beast, so I'd be in favor of trying this.

F

uh However, uh it looks like there's some you can set it up to add the name and for a specific contribution, so we would probably have to talk about what those specific contributions are uh that we would uh that we would want to recognize, and then maybe we could take a peek at our uh contribution, attribution metric and see how well that aligns uh with this, uh which might include even adding an organization so in the name and organization and contribute contribution, type I.

F

Think that that's the way the contribution attribution metric is set up currently yeah.

B

Just my thoughts but.

F

I think it's worth exploring.

A

Like maybe we could just try it in one of the repos, maybe in the community repo or something just see how it works.

D

We have tried some things like tagging things for first time, contributors and so I think anything we can do that might encouragement. We should experiment with.

A

I mean we got nothing to lose right exactly.

D

We do want to encourage new contributors, so yeah.

E

Yeah I feel like there's questions that I can't really make sense of in the documentation alone that I guess trying it out would answer like to Kevin's Point how we're creating a lit like a standard list of contribution, types that we would be more specific about as well as I. Don't know what the data looks like like. Where is that going? How we're going to pull that back into like I'm, not really sure how it works?

E

Yet, um then I could try to peruse in the documentation, but it might not be clear until we actually try it. What it looks like to see if it's worth it for us, so I think if you're up for a trial, I don't see any any downside other than the maybe slight personal attacks of adding that in now,.

A

Yeah Ruth go ahead.

C

Yeah actually I go with Kevin on Sofia because, like it's really good that we should try. But there's also like some questions. We can't answer like, for example, something I just thought about is um for people that make contributions on GitHub that are not related to code, their data gets stored or their contribution gets stored and then or the other contributions actually, for example, design. They will also need an issue to back it up.

C

We can't just like call the Bots to add a contribution without like a linking issue like the picture shows there, so it might need us to start opening like issues um you know for different types, even though they are not being implemented on GitHub. So there's there'll still be some tricks that we need to do on our own side or on our own processes to, but we should definitely try um something we have. We had done in chaos.

C

Africa some time back was to have like a documents um that kind of lists out like uh and that that's very hard to maintain, because you have to go update that document. Every time I think Kevin mentioned that as well right, so that document is even outdated of people that have made like none good contributions, so yeah well, it seems like this Bots is going to kind of automate. That kind of document just add them manually, um so yeah.

C

We should try it out and then we'll also see ways to improve and also maybe give suggestions to the the project itself.

F

I agree with everything, I think Elizabeth, I, think you're right, I think the community repo is probably the best place to try this out. So we've already kind of we've already rolled out uh a working group, that's specifically uh kind of building content in that space. So if we can find a volunteer to implement this, maybe we just roll this into the into that community handbook Community, not knowledge based group to discuss and talk about.

F

And I am I am interested. What the presentation of these contributors would look like after everything is collected. Is it just a? Is it a markdown file? Is it? uh Is there something else behind it? um Oh and Sophia has her hand up I'm. Sorry,.

E

No, no! It's just kind of building on that, because I think Ruth's comment is very Salient to me around this. Just like changed the way that we might work. If we want to use this, and so just like I guess as we're paying attention to the trial sort of also noting how we would have to adapt processes if we do want to use a tool like this I think, uh oh a conversation at Fosse very relevant to this um I was chatting with wait. Can you still hear me my entire screen just stalled.

D

B

Okay, I'm frozen everything is black I can't see anything so I'm gonna make my comment, and hopefully you hear it. um I was chatting with Josh burkus.

E

Who works in the community, the kubernetes community and, if you're in my presentation, I just kind of drilled in on that space? So we chatted a lot about how the kubernetes community has also chosen to put more things on GitHub because of this kind of thing um where, like their events, team and events management does all of their issues via GitHub as a way to just um increase ability to track who's.

E

Doing how much work in this space is too versus just code code work um in a way that, like I, think our our community has always used a number of different tools and platforms that kind of fit what we do and also not forcing everyone to use GitHub who's, not familiar with it, but I do also see the benefits and having more things in one system that has more ability to track measure and collect metrics about, and so I just I think that I'd be curious as we trial something like this.

E

If it would essentially encourage more activities to come to GitHub as issues even if they're not code issues, they could be event issues or website issues or what I mean. That's already happening there, but uh that's a bad example, um but like other things, that aren't necessarily happening on GitHub.

E

I'm going to turn my video off because I think that seems to be struggling.

A

Yeah, that's super super interesting um I. Think it's a I think as a separate issue. I think that's a really cool thing to do is to plan events more visibly on something like GitHub, so I think it makes it easier to delegate tasks as well and to see who's doing what, because I know when we do chaos con. It's very fluid, and um you know it's mostly just in discussions of okay who's.

A

Doing what and we have those we have like a you know, documentation, but we should and as Ruth as we should open an issue for this discussion right here. I think that's perfect. So um I will take that action item um to do that.

A

Because you know a lot of times, people are like: oh I want to help with this event, but it's you know, at least from from the chaos kind of I'm, not sure about the chaos kind of Africa side, but gascon, the the um the other cast con. It's a little bit more fluid and we have a smaller team. So it's just more informal but I kind of I kind of like this idea of um having that those issues to so everybody sees what's going on um open an issue to discuss this issue. I guess yeah.

C

I think I think something that kind of uh maybe limits or on the issue. Opening the issues part at least for my end, is sometimes when I think about the wording to go and putting issue I'm like oh, no I'm, just going to do this like that right. So, if probably when we decide to go through the issue route, we can have like issue templates like say, for example, it's a program. It's uh an event.

C

It's a question, you know have those templates, so it helps people easily open up issues because I know that's something that I struggle on. On my own hand, like anytime I, want to open that issue and I. Think of okay, I have to like structure the world and I'm like okay, no I'm just going to do this.

F

So I would uh I would add that uh those issue templates are documents that would land in the knowledge base, and we do have some templates. But if we are missing templates that we need once again, I would encourage you to come to the meeting on Wednesday and and let us know um what templates or what documents were missing and they need to be created.

D

Then not tomorrow, but a week from tomorrow, right a week from tomorrow, yeah.

A

Yeah and I think um I think Chris, uh referring to issue templates right. Those go. Would those do this go into GitHub org, or do they go in their repo level? Aren't they.

F

Those are repo level yeah.

C

F

The Master Copy would be kept in the uh in the knowledge base. I'm. Sorry, the main copy apologize for.

A

The wording yeah copy and uh okay, okay.

A

So we kind of have two separate things going on here. um We have this all contributors and then we have uh kind of this.

A

This issue of um more.

A

I'm curious to know what people think about using using GitHub, more, especially since we do have so many folks that are so new to open source and maybe not familiar with GitHub like is that going to be a barrier or is that something that is a good on-ramp for them to learn? Github like how do you all feel about that go ahead? Ruth.

C

Well, yeah I think. um Yes, we should do that because, like like you said like so people getting you to open source, it will help them learn just that, while we're trying to why we use GitHub, we should put in pointers like the key for people that do not know how to use GitHub and I think this falls back on that um all the courses that we want to do.

C

The chaos 101s and all um kind of disappoint us on how to navigate GitHub I recently did an article, and there was a comment um from somebody that read the article that they it the the article the article had like Graphics of how to navigate to GitHub, to find um you know open source project, and then the comments was that that article helped them navigate GitHub better because they felt like it wasn't beginner friendly to them so kind of putting pointers out there and then I do agree with Anita's comments on uh Gita and GitHub workshops.

C

I used to run that early on when I say open source, and it was really helpful for a lot of people getting to learn open source.

F

F

So I think I'm a little uh agnostic on doing more things on GitHub uh it. It could go either way for I'd, be okay, either way. uh So the the comment that I would make is that uh so we we used to be a lot more GitHub heavy than we are now uh we kind of moved. We moved towards Google docs for a lot of the work that we do, because they're collaborative in real time, whereas GitHub is kind of uh the the collaboration, can be a little delayed right.

F

So it's collaboration over a longer period of time where it's GitHub docs are kind of collaboration in real time. uh So I think that was kind of our thinking in uh in transitioning to that. So we've actually we've moved from being really GitHub heavy uh to kind of doing work in real time in GitHub docs and then moving things into GitHub afterwards.

F

uh So that's our current process. If, uh if we were to uh start doing more things in GitHub, that probably involves uh reducing the amount of work that we're doing in Google Docs, which once again I'm I, could go either way on it. uh I do agree that the Google Docs do have some uh opaqueness to them, they're, not as open as GitHub is, uh but so that's just uh my comment there. This is if this is probably something we would. uh We would need to discuss over over a few meetings.

A

Ruth, do you still have your hand up, or do you have something new.

B

A

You're talking so I, don't know, maybe you just have to wait for a minute. That's totally fine, um yeah, I, I, agree. Kevin I think um we could certainly talk about how we do our metrics. Maybe it's time to change that up. I, don't think they have to be um I. Think we could do things also separately, so adding an issue adding an issue to do an event does not mean we have to take everything off. Google, Docs, I. Think we could. You know, um keep those things separate if we wanted to they.

A

One necessarily doesn't have to be tied to the other, but absolutely I think we could maybe talk about.

A

um How we integrate uh Google or integrated.

F

And we do uh in those in those Google Docs, we do always collect contributor names when we do that, it's just uh we.

C

Don't we we don't necessarily.

F

Take that next step of of moving that uh that name, that we collect to a uh a more open space.

A

Yeah, so if we had something like that, all I'll contributor bot, we could then easily just be like add these people as metric developers or whatever we decide that contribution is called yeah, because the the each metric will have an issue right. That kind of is the prompt to add it to the website. Right still, we.

B

A

A

So we still have like a stream or an issue that we can keep that.

A

um Okay, what else on this anything.

A

Going once changes.

A

Okay, that's the end of our agenda, who has other stuff to talk about.

A

I will say just as an aside um in my talk with Justin we're talking about onboarding and some of the things that we're trying in our communities, and so people seem to really dig the the idea that we would have a a course and like a learning management system that was very clear and very sequential and um easy, and something that people can do right away.

A

You know when they come to the community and they're excited and they want to do stuff, it's something for them to do and to work through and then at the end of it, then they're a little bit more confident in their contributions to to chaos and I also um really want to just stress the point that it's, it is going to be a community effort that whole learning course I think because I really want to make sure we're listening to our newcomers when we're developing this, because you're only a newcomer once right.

A

So it's hard for me, I mean I can try to empathize with the newcomers, but I'm, not one. So it's it's really important that we can include our newcomers and it also gives them some place to contribute and and the fact that they don't know anything about. Chaos is actually a good thing and a benefit to us.

A

So um if you're listening to this call, if you're on this call- and you want to get involved in in this project as we're building this out- you can reach reach out to me or Ruth um I guess maybe we should start formalizing this somewhere, um because I know we have a doctor, maybe just kind of dropping ideas, but do.

D

You think yeah go ahead.

C

Oh no I'm just going to ask even ask a question: um we we had. We have a user group that is focused on newcomers, I think that was in the um governance, doc rights, so we can maybe either have the conversation in the newcomers group or another slap Channel.

A

Because I know um I think Kevin somebody Sean, maybe you were gonna, say yeah Matt's been looking at platforms and I know you all talked about that last week when I was not here um so I feel like like this is a piece there's a piece: that's you know, ideas of what the content will look like and so I just want to bring the pieces together in one central place.

D

Yeah we discussed platforms last week and the challenge is that none of them are uncomplicated and so Matt's doing a little bit of work to figure them out, but we're we're still in that evaluation phase of where to actually put it so that we don't create this complex thing that we have to maintain.

A

And um funny enough, I was sitting next to the imperio Folks at Fosse and they went that's what they do right. They have open source, Learning, Management Systems, so um that was not on our original list. Sakai is the one that they offer, so um I thought I would bring that up as well.

A

They were super cool and I had a really good conversation with them as well.

A

Okay and I see someone has added this with our last nine minutes so going to bring.

E

This up, it's me and it's random, but I. We haven't really had a conversation like this, so it's more like I'm, just curious to see how this group reacts and whether or not it's something that we want to take on.

E

uh Metrics related, but I've been working on a project and about halfway through my analysis, work I discovered that what I was seeing in the data was more a reflection of the GitHub workflow versus a reflection of actual data change or Community composition in a way that I think as folks who collect and review metrics constantly.

E

There's always this like sometimes a dearth of knowledge between the full system, architecture, Community process and structure that changes fundamentally, what the metrics look like and how we interpret them, um and it's just something that we all deal with in our work and I'm just curious um I. Don't we haven't really been discussing that in our metrics, but there are some metrics that would be more subject to something like this um I can give the explicit examples that helps folks make this more concrete for them, but yeah.

D

B

Just kind of helpful.

E

Yeah, okay, um I've been working on a project to try and distinguish work done in work context versus non-work context across our employee and contributor base inside the company um and looking at various ways to group sets of repositories acted on by employees that look more work-like versus personal like and the most obvious thing to do immediately was to isolate out things and personal repositories.

E

But the problem with that approach is that a lot of the work happening in personal repositories are actually forks from other repos as part of the pull request workflow.

E

So it's the more like there was so much noise in there from that alone that it was difficult like I, had to just kind of remove it all from it. But you have to actively manually remove it because of the way that the events I'm working with data from the event stream. That's also part of the problem. But um it's just more that again, because it's such a large impact. It totally made all the initial numbers that I pulled irrelevant, because most of that work was in fact not personal work, but forks from something else.

E

That eventually ended up getting merged back into the other project. um So again, like I was fighting with the system process and design uh versus just being able to pull data and understand it um and I I, don't I, don't know. I haven't really seen us address that Nuance or I. Guess like be aware of like there's always like I mentioned it in talks when it's like automation could be doing stuff and messing up your metrics, but there's also just again the system, design process and workflow.

E

That could be like, in this case directly influencing the structure of the data and the metrics that you might pull from it.

D

I want to clarify something so there's a I think, there's a principle implicit in most of the metrics that we do especially around permits and pull requests. Yeah. It's really the only things that ever get accounted are the commits to the primary branch of whatever repository that you're tracking.

D

So we won't necessarily see the forks unless we explicitly say that we want metrics on these Fork repos. Typically so.

E

You basically just removed that funkiness from your data by only looking at the main branch.

D

That's that's what most like! That's what grimorlab does that's what auger does and if we want to monitor works that we know are significant. Those just get out of your separate repos, and then we can see the action in there, but you're right and- and so we I don't auger- does not use the event stream, except supplement for supplementary data to check when things close or open, and so, if you're, relying largely on the event stream, then I can see where it would get walking.

E

So I guess it's good to know. I mean, like I, think that's and I assume that was happening in these tools, because I know that there is some manipulation and Engineering happening to ensure that you're looking at the real data and not the noise around the data um but I think depending on it. The tools that you're using that might like you might have to do that as well or like have some correction in it to look at the accurate Thing versus or just like, acknowledge what you're not looking at I.

E

Guess I'm, describing this poorly but I I. Just like I'm, curious I, guess it makes sense that in our tooling we've or the folks that are working on it have insured accuracy by engineering. The output versus the input versus those that are just looking at a raw field of data and having to deal with it on their own.

E

But maybe we don't address this.

D

E

Just kind of wanted to float it as a problem well,.

D

I think I think I would I mean personally I'd like to understand in a little bit more depth what what kinds of issues you're encountering to make sure that we're not missing something! That's really important that people care about, because when I listen to Youth talk, I'm.

E

Thinking I mean.

B

I think they're describing.

D

Maybe the maybe the Box we drew around this is wrong.

E

Not necessarily I think it's always so like the design parameters. What are you trying to solve? What are you trying to achieve and I'm going into a giant data set with another question that was not part of the structure of what this thing was designed to answer. So that's sort of again on me, I'm, trying to see if I can figure something else else out by looking more comprehensively at the entire set of events and activities, but um yeah I, don't know I, guess that's one of those things where like.

E

If you were totally new to this sort of thing, and you weren't aware about how about this about I guess the system structure? Maybe it just comes down to the fact that, like I, this comes back to our previous conversation. Maybe we need training on GitHub workflows, because understanding the workflow and processes better could maybe help to shed some light on where you could get caught up in this.

E

The system design itself, if you're, looking at data from GitHub, which is a little bit meta but related back to like what what we do as chaos and extended data analysts around open source projects.

D

Well, I would say so. If you have somebody like you, got confused or unclear about it, then there's a pretty good chance. Anybody new to the community would be confused during third battle, so it's worth a conversation and figuring it out. So it was explained back to people.

E

It wasn't I wasn't as confused.

E

Like trying to remove the system level structure on what I was actually looking at, so it's like I'm going through my own sort of filtering and normalization process to try to answer my own question but I think yeah.

D

Maybe there's nuances.

E

That I'm missing I.

D

Suspect you're, not alone.

E

I actually did have a conversation with some of the jslash commercial about this Pro, particularly and I. Think they actually have been look been looking at it because for someone else had a similar question, uh so I'm curious how they, how they're approaching it as well um again hard hard conversation, I I, know I just post it. Maybe this is a good Dawn question for the possible data scientists joining our team, uh but, like I, don't know if there's a home for that in this project, but I just wanted to raise it as a thought.

D

Because that I mean, if I'm under am I understanding right that the the challenge with the data engineering that you're encountering is related to what you're able to pull from the event stream.

E

In my particular case, but it's it's more that I think there we have the metrics and we have the software and the software is removing a lot of that system. Complexity to be able to arrive at the metrics, but I think, what's not being explained, is unless you're familiar with the software design, like our software design, uh the broader projects under chaos and Gerber, labs and auger. Then we're kind of abstracting away that complexity, which is great because that's what the tools are designed to do.

E

But if you end up trying to do this on your own versus using the software, you have to either come into that aware of it or deal with it, and so I'm wondering like again right now how the project has been handling, that is by abstracting it through software and removing the complexity through software versus trying to be more, have more conversations or guidance around the sort of the data Engineering Process to arrive at those things. So maybe that is I don't know.

E

Maybe we don't do that, because that is being solved with software versus having that more as a sort of open discussion and of data engineering as a process.

D

F

So I I do think this is a an area that we should explore further so I in the past, I've tried to uh do some research around uh uh uh pull requests. Success, merge success in relation to Forks, uh so the to the point that you've been making Sophia a lot of the work. That's being done on these projects are being done out away from the the main branch right. It's being done on these, these individual repos.

F

uh So it's and those those these individual, repos or Forks are they're, basically a black box right so once they Fork them we're not including any of the metrics around the work, that's being done in those repos, we don't know there could be 10 people working in there. There could be one person working in there. There could be two people working in there. uh The only thing we care about when it comes back to that main branch is the is basically the one commit and or the the one pull request and the merge success.

F

uh So there's there's a lot that we don't understand about the work being done in a project uh in most projects, because it's.

D

Occurring outside of the project.

F

Right, I I, if I understand that correctly and that's and that's something I noticed as well, so the the the complexity of figuring out what was happening in those Forks was not something I was able to overcome in this project, so I had to kind of transition it into a a different thing.

E

Oh, that makes me feel better because I also feel like I I'm, looking at it from another Nuance, but where we do have visibility in that, because I'm looking at it from the lens of our employees and the lens of our contributors versus the lens of the repositories or projects.

E

um In that case, that work is visible to me, but I'm trying to bucket it. So it's more like trying to count it, because that is work happening, even though it's not visible in the activity around the main branch, um and so we're able to include that work, but I'm also trying to include it relatively such that it doesn't look like net new stuff. It's actually existing stuff. It's just also counted as part of that work.

E

So it's a bit of a balance like I'm trying to understand how to how to do again like how to bucket work type, basically, but also I, I, see it from a different lens, so I can see that extended work inside of our contributing population versus like I'm, not seeing outside of that, because we're not tracking anyone else, but our own employees, so I I'm, coming at it from a different angle, but I think we're kind of talking about the same problem and I think it is.

E

It is challenging and maybe something that I would hope. We could provide more guidance on to others. That might be doing this. So again. This sounds like actually a Don problem. Now I feel like this always happens when she's, not in meetings I just create ideas for her and work for her that she has no idea about, and it's okay, if she doesn't pick it up, but like it's.

B

A

B

um And I'm so sorry.

A

To cut this conversation short, but we are four minutes over, so I just wanted to give time for everybody to get to their next meeting. So thank you. Everybody thanks so much for great conversations today and when Don comes back, we will inundate her with all of the new work that she now has to worry about.

A

Thanks again, everybody we'll see you next time.