Open Telemetry Uncategorized, 12 Aug 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: 2022-08-12 meeting

Description

cncf-opentelemetry@cncf.io's Personal Meeting Room

A

Hey good morning, how are you doing really.

A

Hopefully your day's getting started off. Okay, there.

B

Morning, meeting seven: this is my third meeting today.

A

Got up with the birds huh yeah, I understand some of those days, um but fortunately it's the end of the week right and uh can perhaps enjoy the the weekend. There.

A

Which is pretty cool, um I'm not sure if carlos is going to make it- or I think josh sheriff is- has a day off today. So actually.

B

I'll be back next monday, I guess.

A

um Okay, yeah actually for for a live step. This is a what we call a well-being day off as well.

A

Which they have a few days a year or whatever, but uh pretty cool. So um you know we last time we got together, there were. There were a few items that were present but, as you were talking about really some of the knowledge that we don't have, it makes it difficult sometimes to go through some of the items that are there.

A

um I think there were a couple more that's been added this week. um Maybe we just take a quick look at those and decide on whether we can or can't handle them. We can't and find um we'll let them set, and maybe the again the next uh spec meeting can bring up trying to encourage um other tc members to try to come to this meeting as well.

B

And for some specific issue, maybe we can leverage the click handle that, for example, people are saying hey. I have an issue related to profiling and we don't work on profiling. It's too specific for saying this. The profiling state then go and handle that there's something there's no such thing or the thing is not official. The ball have to set the invitation like there's. No such thing, there's no capacity wait and we'll collect the signal. If there's so many somewhere else, people think this is a high priority than all repeaters.

A

Yeah yeah yeah that sounds reasonable there. um Let me go ahead and show my screen here.

A

Okay, you can see my screen right, yeah.

A

And it looks like uh this is the latest one three days ago. So let's take a look at that.

B

This definitely is important, it's even like required for gs, but I think arden made the proper tagging already and it's very purposely appropriate.

A

And you've had two grand as well.

A

Here's my understanding.

B

Someone from the spec dignity and talked about the the desire here and arming is a tt member and and based on the conversation he concluded that we we must have this before we hit the stable version like it's just like he, he didn't add the tagging, but he's already decided it's franklin.

B

And in my opinion, this must happen before they should stable.

A

Yeah okay right over here looks like here.

A

A

All right, you.

B

Want to take a.

A

B

A

One here as well.

A

There it is okay,.

A

Clarifying the tlp metrics exporter requirements.

A

You know josh and uh and uh josh and josh uh a long conversation about this a week or two ago.

B

Yeah, that makes sense we shouldn't do this.

B

It's some it's some, like all the things mentioned here, are some known, small caveats that we decided not a blocking issue for the initial matrix release, but now doctor is suggesting we should take time to clarify those things.

B

Very reasonable.

A

Very very small there yeah.

B

And we have like downstream, like two josh's and and like we, we all focus on matchmakers, so we have the foundation. Thank you there for sure.

A

A

Let's see here, let's take a look down at the next one here.

A

B

It makes sense to me as well- although I I don't have time to drive it but like given like josh, has been very involved in matrix and he seems to be willing to drive this. I feel like he's. Okay, if he's starting it then.

A

Extremely okay.

B

Yeah I like I'm, definitely happy to review his proposal and guide. You approve.

A

That sounds good. um I know he's been pushing to try to resolve some of these um these issues and get things finalized uh from a spec standpoint cell uh looks like jack bird has been involved in that discussion as well.

B

In general, like from matrix uh related issue triad, because we have, we have like five active uh members here,.

A

B

They all have approval power on the stack, so it technically makes sense. It's very easy to make progress like this is well-founded and super important for many companies.

A

A

All right so uh we'll go with triage accepted here.

B

And we probably should remove tigre from the assignment it's not working metrics at home. It's unlocked, there's nothing! You could do about this.

A

um Who would you suggest if we were.

B

Yeah, maybe just me, okay,.

B

A

First thing: you need to do.

B

Is to clear the assignment, then you you add me.

A

All right here we go cool and go with that.

A

Didn't quite get that today.

A

A

A

Let's take a look at this one here.

A

And the josh here and the histogram's been working that still as well.

A

B

I feel probably need more discussion. Do you scroll down? Do you see any active discussion here.

B

Yeah so so I have similar concern.

A

B

I'm not concerned that that this is not well baked, but it's not more discussion and it should be deferred.

A

Deferred or just um I guess you would just.

B

Say, or or wait for people to have more discussion, it seems like josh has some problems he wanted to solve, but the proposal here is likely going to introduce more problems.

A

Yeah, that's the concern of pergio there right in particular.

A

um Should we say that this is just that triage, but it needs more information, I'll just leave it open and let the the conversation mature over the next uh couple weeks or so.

B

I think either way- probably fine, maybe I I'm more lean towards we will put the tree. I just need more information.

B

Ideally, we want to see like more feedback here to understand whether there's a common need.

A

Okay, all right so we'll just leave it the way it is we'll revisit it again. uh Next time we get together and see what the traffic is like. Okay,.

A

um So what I can do here is.

A

A

This is where my car laughs here.

A

So it seems like there's more discussion still needed here. You agree. I.

B

Mentioned convention, where I have no expertise, I always had my personal opinion, but I decided not to put my personal opinion, it's something very, very subjective and I think the semantic convention 6 should figure how to handle that.

A

So your recommendation is that we don't accept this, but we allow them to.

A

Continue to uh to discuss this.

B

B

We don't have that process, but I like personally.

A

B

I would prefer to position us in this way, like hey, we're, not expert in this area, and we we think it's okay, that the semantic communication to make their own decision, whether this makes sense or not we're going to give give this issue to the semantic sick convention. We don't see this either strongly against the overall open parametric direction or strongly aligned with something from the trilogy we're, okay, whatever it's like whether it's taken or rejected by this many conventions, but they should make the decision.

A

A

B

Signal that country there's no way we can. We can say, like this.

B

Triage, depending for the semantic emergency.

A

um So yeah something like this or uh uh allow: what are the words you use.

B

Yeah so so on the tagging maybe we'll just use more information, yeah.

A

Would that make sense there.

B

Yeah and how how do we send that message.

A

Yeah um is this: is this I guess this in general is a question for this group. uh You would have something like this: how do we communicate it? Is it? Is it in the specs egg meeting? uh Is it uh you know playing with the status of this thing to say um you know we? It needs more information.

B

So the stable is already good enough. It is saying it's very semantic convention.

B

So if it's aries semantic convention, it's not something generic about the like the the trace or something.

A

B

Basically about the semantic convention, maybe they should decide because http information is a specific area like we don't have the general knowledge here.

A

Okay, is that a comment that we should add to the ticket here, or what do you think comment that the semantic conventions I I.

B

Don't feel strong about him.

A

A

A

Yeah, maybe maybe the thrust would be um um not putting the comment in here, but still communicating uh to them, perhaps at the spec sig meeting that this is something that, uh as a result of this meeting, I felt like maybe the semantic convention's sake so take a look at it and provide their their judgment.

B

Yeah, I would assume like given like christian here. He already added a semantic convention tag and he's working on the semantic convention.

B

He's already taken care of this, and if nobody ever look at this from the semantic convention, then I I would assume they don't care about.

A

Okay, so um in that case it would be not saying anything but then watching it to see if it goes stale over some period of time.

B

A

A

I'm sorry excuse me, thank you all right. um That sounds that sounds reasonable to me.

A

So, let's see look at that all right. Let's take a look at this one here. I think we have time for this one at least.

A

So that's not already there. Okay.

B

It's similar, like the synchronous histogram, will be more discussion.

A

Okay, so I'll go with.

B

Yeah sometimes.

A

A

What's that change, I didn't change that.

A

Oh, oh, that's interesting! So it says it's 12, but it's 11.. It's interesting.

B

It will refresh right so maybe like when you try to copy ah okay.

A

That's what happened.

A

There it is okay,.

A

Okay, so um let's go to the next one, then we have time.

A

B

It looks like josh already assigned this to me. I know.

A

B

It seems like he, he already changes this issue and determined that dryla should be I'm happy to pick.

A

I think riley agrees.

A

All right all right.

B

Yeah, I need to take some time to understand what I I have been volunteered to do.

A

Nice to wake up to that right.

A

And thank josh for knowing that you would accept this.

B

I'm trying to do whatever joshua signed to me.

B

Oh, look at you.

A

A

And I think, that's probably about all we have time to deal with today, um very good. So any other last comments you have.

B

That's the messaging which we looked at last yeah.

A

Last time yeah, I think still.

B

It needs more information, there's no like there's no clear direction here.

A

Okay: let's go ahead and change. This then put it in there.

B

And also we can add the text. I think this is semantic convention.

A

A

um So, okay, okay,.

A

All right um anything else for this one.

B

A

B

I have a question for rejected item. We just close the issue like, what's the expectation.

A

Yeah, I guess that so when we reject something that there's a message: that is there a message that goes back to the originator that says it was rejected.

B

A

B

Is so so I I guess there might be two different ways out one is we rejected? We keep it there. So if people disagree or they have more additional information, they think they can convince us to change their mind. They can still reply.

A

B

Maybe after grace period we can just close it. I can look for some like if there's a automation, that we can automatically close a rejected item if it's been inactive for a month.

A

Yeah, that's that's how I would lean to give the opportunity for the originator to to rebut or come back with more information that might change your mind in terms of the rejection.

B

It's a little bit uh on the edge. I was basically myself if we really think we should give our chance to justify it, then the things we should use we actually need more instead of rejection and rejected is the final decision, and if that's the final decision, then maybe we shouldn't even have that time. We should just closely just.

A

Close it out, yeah yeah, um what's the what's, the purpose there, so you know, is, is triage rejected in the way that we use we've used it basically saying hey, you need more information before we accept it, and if you don't give us more information, then it's going to be rejected and closed.

B

uh I I feel truly actually need more info is for that purpose. Triage the rejected is something that we clear. No, this is not the direction we're not going to do it, so it should be closed right away.

A

Yeah yeah, I guess I can appreciate that.

A

It's not how we've used it in in the past, though right um we've rejected it, and I.

B

I don't think there's been any.

A

Any rebuttals coming back from those uh that had originated it so yeah. In that sense, uh uh it seems like there's, it's a reasonable thing to say, say, rejected and then still have a a window right to allow them to come back and try to read that.

B

A

Yeah yeah, uh whereas uh needs more information, saying it seems to be saying that we we don't have the information to say, reject. We don't have the information to say accept, so we can't we're not going to go any further and or advocate for anything further unless they give us more information.

B

Sounds reasonable to me.

A

Okay uh sounds good cool um thanks riley um you enjoy your weekend and we'll see you next week. Okay, thank you. Take care. Thank.

B

A

A