CHAOSS Common Metrics Working Group, 19 Sep 2019

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: CHAOSS.Common.September.19.2019

Description

CHAOSS.Common.September.19.2019

A

You're not recording okay,.

B

Welcome to common metrics meeting for the 19th of September.

A

So I had the homework, uh put it in the chat, but it's also in the document. It's.

B

A

B

A

The this this one yeah exactly yeah you're in their second.

A

So if you recall last time we talked about this, we were the way that this metric had started was whether, basically, whether or not somebody is contributing to projects within business hours, remember mm-hm,.

B

A

We kind of ended up on the notion that maybe business hours is just not a term. We can work with in any reasonable way, so we ended up kind of pairing this down or kind of when we get down quite a bit to really just capture commit hours and dates. Mm-Hmm.

B

A

What it came down to- and so this section right in here was kind of this- I left this in there as an indicator of what used to be there mm-hmm, but I think what came from the conversation, at least what I heard was that we need to just pare this down to perhaps just these kinds of items that we can pull from, get just really big-time stuff, yeah.

B

Exactly because I think you know someone could use this too. They could use this data to do something with business hours, because we would have the they would have the data they would have. The timestamps, but, like you said, business hours was, was problematic.

B

But the date I'm not sure about commit count because.

C

B

Stamps I usually buy commits, so you wouldn't have a commit count that match either and stamp yeah. What.

A

Is in github the count is, it lies.

B

A

So that was it, take it or leave it I think it's a size issue. I was just kind of trying to kind of go through some API stuff and see what we'd have access to and.

B

Then I then I started.

A

Even thinking further like, maybe we don't attach it just to commits, because if we're common, then we want these eight timestamps mm-hmm for pretty much anything on the activity side of things.

B

Yeah I was just I'm. Just sorry, I was just looking at the github API to see if I could figure out what that commit count was.

A

If, for a while.

B

A

It was like I'm like an ID number, but I. Don't think that's what it is.

B

Is that what the field is called commit underscore count? Maybe.

A

B

A

Was looking somewhere else.

A

Is there a commit, sighs, I, just typed, it.

A

A

I know where I pulled this from I pulled it from the auger endpoints. ah That's right pulled it from okay.

B

Yeah I don't know how hard to find flat so.

A

Maybe I just at least for the time being, maybe at least strike this through.

B

B

And that person- that's that's the bit that we actually need is who made the commit. Okay,.

B

Something along those lines.

A

A

And then, then, when I was doing this I'm like is this getting so pared down that it's like not useful, I.

A

Mean is, is talking about in common date time as an important metric mm-hmm. Should we formalize it I think it's kind of informally in a lot of the evolution metrics. That's all yeah.

B

Yeah I'm thinking that it's probably kind of a what is that we were calling these kind of atomic metrics that are you.

C

B

A lot of things build on, it seems like this is probably one of those okay.

A

And because I mean you can pull this kind of stuff for almost any action. Mm-Hmm.

A

B

Exactly so, it doesn't even have to be a commit date. It could be anything monthly, yeah.

A

Right it right so then we just like even the title just becomes like activity mm-hmm.

B

Yeah cuz, you could use it for things related to issues. You could use it for things related to pull requests. You could use it.

C

B

B

A

That is about as atomic as you can get Patrick I. Don't know that you can go below that and then to your point. Yes, then you could use this. Obviously, if you wanted, you say, like don''t: pivotal, wanted to start identifying business hours, mm-hmm.

B

A

Could do that right, but that would be your their own, your own values that you would set.

B

So this originally came out of the geography metrics yeah um I.

B

Guess I'm trying to figure out what we what we do with these, because some in some cases they may projects may have geographical data in other cases they may not.

B

Are there what other metrics did we have for geography.

A

B

Where those were I.

C

Remember when we discussed about engineers, maybe in China, changing the timezone in their comments, all that they are elsewhere and that made patches are not coming from China.

C

Was there any any progress in any workaround or is it just the way it is I.

B

Suspect, it's probably just the way, I'm, not sure I, mean I, think I think that's something you maybe have to address on a project-by-project basis. That was if that was a big issue for your particular community and then maybe you'd make some adjustments. No.

C

I was raising it as claiming that it may not be reliable. They have to be about reliable, not metric.

B

Yeah I would definitely agree with that. Mm-Hmm.

C

Though, for me, it will be very interesting because we have products like we claim to be the upstream for a project, and it's all done by our engineers in France in Sweden. Only and.

B

We see it when.

C

We use green moon or lab or II, see it for for healthy or transparent projects. That's being of.

C

Comment not you were mentioning normal business working hours and when and then you edited when in engineers provide concrete for their their employers.

C

What about France, because it states that there there's this rule in France that engineers do not even receive emails from their managers out of regular working hours? What if it comes up that their comments are coming at midnight, when the trade unions in France enforce no work at all?

C

B

Interesting, so we could conceivably run into ethical issues or we could go in a trouble with the law.

B

C

Yeah, maybe it's not all.

C

Coming your way.

A

So in the minutes, I did drop the geography link on github, so we had contributor country, employer country and come in hours, mm-hmm.

B

A

Are the other two and honestly at this point, commit hours could just be modified for these commit date, hours yeah.

B

A

Commit but activity hours right.

A

You can get rid of that regular business hour comment.

B

The other, the other two look pretty good too I mean contributor country and contribute our employer country the headquarters for the communities to have that data.

A

Can we does, what does github give you on country, stuff, github,.

B

Gives you a location field which a person can put whatever they want in that so I've seen.

C

B

Put like Planet Earth I've seen, people put, you know like just the name of a city, but no I think mine, probably just says London, is that London in the UK is that wonderful.

A

B

There's a London in Ohio, too I think there's another in London and like yeah, so that data is not particularly precise. Okay,.

A

Is it so imprecise that it's not even moving things forward, I, don't.

B

Know I mean it might be, it might be useful.

B

You know in depending depending on the community I mean some communities have you know some communities have better data than others right like if it's, if it's a small community, you might, you might know whatever ones. Location is if it's, if it's a community that has you know like other user data, like some, some communities have additional profile data. They might have additional information about location.

A

What you're talking about is kind of like social knowledge, not something that I did from trades data. No.

B

It could be traced data, so, if you think about a community, for example, that has say a discussion forum that is tied to your github account, for example like uses OAuth or something for authentication, and it has robust location data in it. You and you might be able to get trace data that way. I see.

A

B

It's a if you're analyzing a community, that's actually part of a part of a company like with inside of a company. You probably have location data that you could pull from. You know: Human Resources, trace data, okay,.

A

B

Analysis, there's lots of places. People could get this data.

A

Because I guess I'm the the root of the question was you know how we have two different templates for metrics.

B

A

We have the template, that is the date/time one and obviously we have the template. That's for the DNI.

B

um I think it's probably probably the looser template the one that we use for DNI, but.

A

Here would be some qualitative places you could go here are some quantitative ways that you could track this down. Yeah.

B

Here are some ways you might be able to gather this data.

A

Okay, then I can start kind of like this. One started building that out a little bit okay and then we can work on it in a couple weeks.

B

Yep that sounds good I.

A

Think also, if we on the commit at all the one that commits our one, the activity date and time, one I could also, in the interim reach out to folks at or more lat mm-hmm and at augur and to see how they're doing this so that we can are talking about. I can fill in things like visualization mm-hmm,.

B

Yeah and we can see possibly other ways- they use this data, so there might be some more metrics that are more. That kind of build on this data that are also.

A

B

Of related to geography, yeah.

A

Would you call those filters or visualizations they're, not exactly filters? No.

B

I'm thinking more like, like they're just they're, just other metrics that use these metrics right.

A

I'm thinking about how to how would I put that in the template? No.

B

I think it would be like it might be a separate metric. Oh oh.

A

I see what you're saying yeah.

B

I'm saying once you dig into some of the data in grimoire labs and.

C

B

Augur they might have some additional metrics I.

A

Seizures that are also.

B

Sort of location, geography, time zone based that might be interesting for us to define I get.

A

What you're getting.

B

Getting back to the, we should be better at defining the metrics that we're already using in our tools that we don't have in a release. But everybody's using I will.

A

I'll put that in my like agenda for the the software meetings next week, cool they both pass this week, but.

B

Okay sounds good: okay,.

B

Is there anything else you want to talk about related to geography, metrics, no.

A

That's it. Thank you.

B

Do you know, will Shawn me back next week or not next week, but the week after for the for the next one.

A

What date is that it's.

B

The what's today, Tuesday the 19th, so October 3rd.

A

When, when is Grace, Hopper.

B

I'm sorry I when to put him on the agenda. It's Chuck, my daughter so.

A

He should be there he's doing the OSS day, mm-hmm for Grace Hopper, but that's the day before the conference, which I think it's the last day of September the 30th, because Grace Hopper is October 1 through 4. Oh.

B

So the 3rd of October he'll be at Grace Hopper.

A

But I don't even know if he's staying the entire time, so he should be available. Ok,.

B

We can check with him the other yeah.

A

I can check with him. Ok.

B

Yeah that'd be good because, if he's not available, we could have someone from grimore labs come and I'd do a similar thing be.

A

B

They talked about some of the metrics that there they've, already defined and they're already implemented in the tools they figure.

A

B

Should put some definitions together, I thought we'd just start with augur because um you know Sean's been and so many of these meetings, but we could pick appalled, Georg or Danny, or anybody else from the Guru Labs project into I. Could.

A

Tell you about that I could ping Georg to see if, like Santa or Alberto, mm-hmm could come to do exactly what you're talking about yeah, ok, yeah.

B

Because it'd be interesting to just just kind of have that have that conversation, excite yeah I do want to get more of those metrics actually defined in the release. Ok, cuz. We have so many things that are implemented in the software that were not in the release at all and I feel.

C

B

I feel like that's like our biggest gap, we just keep working on new metrics because they're like oh well, there's lots of stuff. We haven't defined it on, so we yeah we keep working on the new stuff and we haven't really defined this stuff, we're already using.

A

B

A

And they're just an AMA, they have visualizations, they might have existing implementation. I know.

B

They have existing implementations of lots of metrics that we have not defined and put in the release, I yeah for sure, okay. It will be really interesting to better understand which is sort of like step. Two of all of this, like step, one is figure out what some of these metrics are I'd, be willing to bet that there are metrics that are defined differently in grimmer labs, then in augur, I agree.

A

And that's I actually think that's right, metrics committee. This work can help in that regard. Yeah, cuz, I, don't think we want different definitions like this is what evolution had done with hey Seuss and Sean.

A

It felt like an inordinate amount of time dealing with what like a contribution, was and how they both find contribution and they took it forever, but I think it was important. Yeah.

B

Yeah totally, no, it is important all.

A

Right well, I put that I'd have a couple action items now: okay,.

B

Perfect, okay, anything else. Anything else I want to talk about in this meeting. No.

A

I'm good, okay, I.

B

Feel like either everybody's been traveling, or we need to do something to sort of drum up support for these. These meetings, like.

A

B

Feel like it's Wendling a little bit well,.

A

I do think travel is I honestly, a part of it yeah.

B

A

Be honest with you, I think, reaching out maybe reaching out directly to Georg and getting Sean here.

B

Mmm-Hmm that'll.

A

B

A

People to a half a dozen people makes all the difference in the world. Mm-Hmm yeah.

B

A

Yeah, that's about it.

A

B

Cool well thanks everybody. Thank you.

A

Well, next time have a good weekend bye.