Node.js Diagnostics Working Group, 11 Jul 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Diagnostics WG meeting - July 11 2018

Description

https://github.com/nodejs/TSC/issues/565

A

Think that out chat that you'd only smile, if it I think I think you have the same problem as I had because a lot of the videos appeared alright yeah.

B

I know beep remoted by Michael.

C

Yep yep, that's probably what you see until you get promoted so welcome for the viewers. This welcome to the Diagnostics working group meeting for July 11 2008 18.

C

We have our standard agenda, so are there any announcements that people would like to make before we get started.

C

Okay, if there's no announcements, you might as well move on to the agenda. I guess before we jump into the agenda, is there anything that people want to add to the agenda? That's listed there.

C

No, okay, so first thing on the agenda is informal diagnostic meet up around notes, summit I, don't think, there's anything more to discuss on that front. Let me just open the issue to see if there's anything it's been added recently.

C

Yeah I think it's just you know a few people have said when they might be able to make it to get together, and if people get together, that'd be good I'm suggesting unless anybody objects. We should just take that off the agenda.

D

Okay, so let's just take a note.

D

C

So the next item is diagnostic working group, deep dive topics.

C

This is 168 I think this is just as a reminder to us all that we wanted to do some of those.

C

So I think matthäus had proposed an agenda a little while back for.

C

The deep dive into meeting expectation on tier sports- that's one actually I- think I did put together a brief overview and yeah. So this is a starting point, so I thought we'd spend some time talking about that. So, let's dive into that. This is the issue, but maybe it makes more sense if I try and share my screen.

D

So let's do that.

C

Sure screen, hopefully I'll get the right one. Okay, so hopefully everybody can see that now.

A

C

Okay, I just want to bring up participants again, just in case somebody else joins that.

C

Okay, so we're good, um so this is basically just the starting point of what I put together.

C

I tried to make a definition of a couple tears, so maybe we can walk through those tears see if we think they make sense or tweak them that kind of stuff Tier one is basically you know it's critical to these ecosystem must always be working for all current and LTS node releases.

C

We release will not be shipped if the test suite for the tools, not green, for tool to be considered for inclusion of steer must have a good test suite and that test suite and a job must exist in the note C I, so that it can be run as part of the release process, so no commit to master should break this tool. If, then, if the next releases was in one month,.

C

So the you know an example of one that might, you know, might already be in that kind of category based on the testing we have I think is the the debugger command-line interface since we ship it as part of node I.

C

Think there could be others that you know we might want to get to that level, but there's certainly nothing else that worth it's probably already there or a lot of them are not there anyway.

B

What we could do is to just put them put all of them to t4 and then gradually sort of promote them.

C

Absolutely absolutely that's what I thought I that's what I did in this diagram? Mm-Hmm I, the one the one tweak I did on that was to put like it. We might go through and agree on what we think the target would be so but yeah starting at four I.

C

Think you're right, like let's make this list well we'll get there and go through add everything we can think of to the list, and then we could sort of promote them up through the different tiers is exactly what I had in mind: okay, um so maybe well just going to the next one and at the end, then we can tweak them or something or or if as long way so so tier two I just said you know, tool is very important to the ecosystem must be working for all LTS releases.

C

An LTS release will not be shipped at the test. We've the tools, not green, for tool to be considered for inclusion. This tier must have a good test. Suite and that's a sweet job must exist. The nope see I and commuters part of the release process, so the real main difference there is that it we could break it potentially in the current release, but not in the LTS release.

C

So right, I, don't know what you guys think in terms of like. Is that a good distinction between the two.

C

Yes, it is silent. I'll take silences as like. Let's move on to the other one.

B

It's actually important here is that tier 1, through 3 all will actually require having tests in the first place. Yep.

C

B

Not sure whether that applies to too many of the ones that we have in the list mean.

C

In terms of like it being what we want to get to, or whether it yeah.

B

We definitely want to get there right, yeah, the actual like what we currently have looks differently. I agreed.

C

Like we don't that's why I started them all in four. Is that yeah I don't think we've actually got good coverage, and even if we do have some testing like we don't know well enough. If we do or not to be able to say that it could be one of the higher tiers, I guess the thing. The key question there is like do. We agree that they'll? That is a minimum requirement, for you know basically, that the tiers that we above the bottom one.

E

I'm thinking of like even real studio code, debugger and they're doing I mean they have guess. It just seems unlikely that notes here. I would run them.

C

Right so, but you're thinking that they should still be able to be one of these other tiers, yeah.

E

I think, if anything is important, debugging usual using videos for your code and rules. If that doesn't work, I mean I. Think that's the most important altitude.

C

Then I would argue we need to run them in our CI, because if we're gonna make a release, how do we know if they're there or if they're, working or not.

B

In v8 we have test Suites for the inspector and for the debugger at various different levels, and I could imagine that in you know, we start adding tests. That test, say inspector interface just directly without involving petrels at all. So.

E

They're, just like that already and I guess, the the node inspector Conley tuned by Yan is actually is a good proxy there as well, but I guess what I'm really getting at is that? Do we need to test dev tools to say debugging? Yes, yes, I.

F

Think right rightfully it's currently broken up into two different ones where the eighty back protocol is separate from chrome, dev tools, but I think is it's good expertly v8, you back to recall, which is the chrome f2. As for the call I'm, assuming that one is Tier one or targets Tier one right now and the dev tools about tier three targeted, which you know we I I kind of think that I think that that is correct right.

F

We cannot possibly actually make sure that the latest chrome release that people are likely to use with the dev towards opening their that that works and all platforms against the theft was protocol. But we can verify the DEF as protocol is working as far as we can tell from protocol tests right.

B

I, don't think it makes sense to test the dev tools UI as well, because there's not something that you can influence to release that right.

F

Yeah but I think that's why the Tier four is very, very nice because it allows us to call out. This is a tool that we recommend people be aware of, because we think it's a useful tool for people to be aware of, and we make an effort to make sure that note is compatible with it. It's necessarily something that we verify on release, so the chrome dev tools you I, would be an example of that this useful tool. We think it should work with node, but we can't really verify it on every release.

C

C

So Ali, how does that fit with what you were like I think you were commenting on like requiring tests for the first couple. Tears might cause us to exclude some things.

E

C

We we wouldn't want to so I think it.

E

Does make sense I'm just trying to square away the there's a difference between tears? So if you look at the definition of the the tears they talk about how important a tool is to the ecosystem and and I, don't think that's. So what I'm concluding is? That's not the right way to put it it's not about how important is it to the ecosystem, because the number of users of the video studio code and chrome dev tools are significantly higher and if we break that like be a broken, the ecosystem right, people cannot debug.

E

So so it's not about importance to the occurred being critical to the ecosystem. I give it's it's about. What kind of testing node is going to do on a given bathroom? That's the only thing that there are animals out here.

B

E

B

Setting up some sort of end-to-end tests with chrome, dev tools with nodejs, it's probably going to be hard because the way dev tools, mostly Tessa's UI, is using their tests and I imagine running their tests and, at the same time, a node.js instance and then connect. You know sounds a little bit tricky right.

C

So I certainly take the point that I think Ali's point is that I've mixed into these definition, two things one is like I've tried to rationalize why it would be a tier one as well as saying what you, what we'll do at tier one right so and the fact that it's critical to the ecosystem doesn't necessarily mean that we're gonna get it into Tier one. So I guess do these standalone. Do you know we think we just took that part out. You know the.

C

If we just took out the first bit, which is the tools critical TSO to important or tools? Does it still stand as or do we need to have something that replaces those so.

B

I think, having sort of standard for how we test things is pretty important like, but they may not dark him up to the importance of the.

C

New standard way, I mean is how we would decide whether they can be Tier one or should be Tier. One.

B

Yeah sure the thing is some of the tools is important to have like working as a foal and to test and to anything like chrome, dev tools, but on the other end. On the other hand, it's kind of hard to test at the same level as, for example, you know profiling right.

C

But I guess it's that, like there's the the the tier one could be like just because we don't get chrome dev tools into tier one doesn't mean that it's not critical to the ecosystem. I! Guess it's I! Guess it's! Whether Ellie in your mind, is it's whether we should have like we're basically saying these things are not important if we make the target tier lower than we'd like it to be, if it wasn't for the test test, consequences right, yeah,.

E

I think the deer in mind from what I'm doing is not related to importance at all. It's it's related to what kind of testing the node project is going to do before relieving a version, and- and that's that's about it like it's a Tier. One thing is something we test and we block and Rudy's on yeah and.

B

For that I don't think there should be a distinction between tier 2 inches. We. If we do have tests, then they should pass right.

C

The only difference was that they could not be green for a current versus a release.

D

Right, it's sort.

C

Of like you know when for an LTS release, they have to be green, they all have to be green. It would be that, like you could say that, like because current gives you potentially six months, you might sail and a version of v8 that breaks the tools in the current, but then you've got six six months to fix that before you go else. Yes, you can't actually go LTS until you fix that in the current and.

D

That was just I'm.

C

Not saying that's the right thing, but it was kind of like one way you could say is that, like okay, current being working in current is a little bit less important, because we don't necessarily recommend people use current in production and therefore the tooling is maybe not as critical right, but that.

E

C

Just my rationale and I can easily see that you know we could combine those tears or whatever.

C

So is everybody like if I edited this so to just basically say to must always be working.

C

Must be working.

C

Yeah does that actually make sense to people mm-hmm.

C

There's no objections. I'll update that.

C

Are there any other things that people think we should include in the tiers like this now? Basically, just is. The tiers are strictly like what kind of testing are they gonna get in where they have to pass.

F

Is there any way to so? For example, let's say someone has a pet performance profiling tool. They submit an automated test suite against node.

F

We tear it as tier 1. Is there anything that makes it anywhere like well I'm, assuming that we could then make a decision for that tool? To just say: hey, yes, it has a test Suites, but we don't accept it and then that would make it tier 4 or right.

C

Maybe maybe because we've taken that out of here like maybe we need like something like the choice of which tier a.

C

Tool will be assigned to.

C

C

So this is where it could get contentious right like if our group just says, hey sorry, you can't ship anymore. Actually, maybe it needs to be a combined decision between the diagnostic workgroups and release working group. That will be it a decision, a.

E

Like a crack ad, or how do you make decisions right, so we can I contributed to that, doesn't make it more important to the ecosystem, then maybe somebody who's, not part of the vertical.

C

E

Was objectivity yeah right.

C

That's that's what I was trying to get out and these stuff we just removed in terms of like it's critical to the ecosystem or.

E

I think that's more related to what we add where.

C

So what would those criteria be.

E

Number of users using.

E

Yeah and not be the only criteria because a tool that enterprises use I know I, have mass right. Jana I, interrupted you sorry about that. Yeah.

F

I was I was trying to get to something like you were saying. I was trying to get something like key enablers or the the availability of alternatives like if it's just the only way to get production tracing in Windows, then it's kind of doesn't really matter if it's used by every single person wait.

C

Yeah and so number of users available alternatives, impact to the overall ecosystem. I think we already talked about, like we may say, yeah we'd, like we'd love it to be in Tier one. But sorry, it's gonna be Tier four, because we just can't run the tests right. Yeah.

E

The other thing is not just that whether the tests are available, whether the tester flaky so I'm I'm. Sorry, if you do it important, but your tests just flake out 100% of the time, then we can't have it.

D

To consider Bill, you run the test, wieners, fish and and.

D

Okay, I will just say if the test suite passes consistently right.

E

That can be indicated with our CI okay.

D

Availability of reliable tests we can integrate.

A

Citycam also have an add, a requirement that there needs to be a person, a contact person that that can you think, like contact and say hey this. Something is not working.

D

Yeah actually I like that.

C

So it's it is it that should be neat to be more than a contact person like basically there's a maintainer willing to support it like to help resolve issues. Yeah.

E

I think we should it should be a superset of the certain criteria right. So that's what I ever use for that buddy, more testing in or CI yeah I know. Let me get it right.

C

So I'm, just wondering like instead of just like I, would almost say, availability of maintainer or maintainer of.

D

D

C

So that sound, reasonable yeah.

G

C

So, like that's sort of contact person plus right, it's like you better, we need we need somebody who's invested enough to say this is important and we need to help get it fixed if it breaks.

C

Are there other ones if you were looking at the sitchin list? Are there other ones there that we should pull in.

E

Would be reference this animated list instead of sort of copy it.

C

Let's just take I want to: let's pull it up and see whether they it looks like it's that makes sense.

D

Okay, that's on the main readme. Is it.

D

Okay, here the requirements right.

C

Right, so are these like his module source being on github a hard requirement.

C

Published versions must include a tag on github, so the first.

A

To see a new window on his bean I know.

C

But I can hear I was just looking at it. Let me share it new share.

D

Actually know, instead of doing that, I will just try that over here I'll try to drag it over here come.

C

On try again, okay.

D

C

Because I think, like the first three main well, this the third one doesn't make sense, necessarily the first two I'm guessing, not necessarily either.

A

Yeah I, don't know why why we would make those a requirement.

C

So I'm not sure just referring to that list that I'm looking for any others, we might want to pull over yeah I.

G

F

Like the soft requirements.

F

Because it's pretty much the same thing that we already wrote up so.

C

Let's see here as this, the must be actively used by the community okay. So it's kind of the.

C

Totally dependent on unique portions I, don't think this one applies right, I mean.

F

It's basically the ability of alternatives already cutlets.

A

Heavily depended on also in the case of Sitka means, like literally mmm different rice, yeah,.

D

A

Case it doesn't mean that it means to pendant on in a different way. Obviously,.

C

And some of them would mean would make like we could make these be something like level.

C

Like if we yeah so as opposed to criteria, we could we cast them as requirements if that's better.

F

It will be hard to come up with hard requirements for, for example, number of users, especially since we don't have a good way of measuring it. Right.

D

F

Is also actually an argument against using it as a because it wouldn't mean that we would have to come up with a way to actually not only gut feeling, but actually talk about. How do we tell how many people are using beat race.

C

Alright, so I guess it depends if we're worried that somebody will argue that that alone justifies that it needs to be there like, because often it'll be like okay. We know, we just know like lots of people use this or it's so you'd say well. This one qualifies just because every you know lots of people use the chrome dev tools, for example it's the sort of, and then there's one on the other end, we're like you can probably look and see well, yeah, okay, we've never heard of it.

C

So it's probably you know small usage, it's the ones in the middle, where it might be harder to yeah.

F

So maybe actually rephrasing it as the first one this, the the mode you must be actively used by the community off the pool, must be act just number instead of number of users, just there have to be users.

F

Let me know else: okay,.

C

So there have to be active users, so.

C

C

Although I mean that that one I would almost say the the level of use in the community, I.

A

Kind of like the the phrasing of the flow of actively used by the community- that's kind of it doesn't say too much, but it still gives like a feeling of. Oh, it's the gist of what we're trying to convey right right. How about this? In my opinion, at least so.

C

As a requirements, whether the to look that would be considered like whether the tool is actively used by the community.

F

Yeah and also the module fits into a key guitar category brings us to an interesting one off should be define a scope of the kinds of tools where we say this is something that we are able to add to the release requirements.

C

In terms of, like you mean the heap profiler debugger tracing yeah,.

F

So, for example, if there's a tool that is a super-great checker for licenses, which is clearly not in our scope, yeah.

C

F

This is something that we would not be able to add to the release as a blocker, most likely.

A

We just take those types in the mill column there and say those other. Is that a complete list of areas or I don't.

C

Think it's a complete list. But but what are you saying so like we would take that list and say tracing can't be one and profiling can be or I.

F

Would just take those could categories as a starting point, and maybe people need to amend them the future just to set tell people hey. This is the kind of tools that we would be able to help with of be able to talk to.

C

Okay, so this is basically a separate thing. The current occurrence.

A

It's like in 9:45 right.

F

Yeah yeah, basically you're defining a scope of what we're talking about we're talking about that.

C

We will that that fall under this I don't know what we would call. This call fall into these these. These tiers.

C

Now, I, don't think it's gonna, let me easily, but we can do that and we probably need a bit of a description and so forth. Right.

C

Okay, I can pull that over.

C

So any other one of these that we saw we covered off this guy, the heavily dependent I think we kind of unique portion is the availability of alternatives.

C

Okay and right so maybe a tool that's into and now I understand better.

C

We don't necessarily need, although we as listed because I, don't think it hurts to actually list what those categories are all right. A little bit of info.

C

C

Do we want to say something about modules? The tool is under the node foundation, github org.

A

I didn't then other people can submit anything we've.

C

Just basically said that that you know yeah I think we can leave that out in this case, I think in the end, in the case of the modules it was like, if it's one that's under our org, we should likely be testing it. In this case you know, maybe that's the case for our own tools, but I, you know, but I think we would I, don't know we could put it I think the reason it was there is.

C

It was sort of like if you want to say, give tools that are being maintained by the organization and easier way in to just be all say. Well, we're part of the org. That's what it gives us. Oh.

A

Yeah, that kind of makes sense, I mean hey, but it, but it still has to fit into like one of these categories right. Oh.

C

Absolutely right, yeah.

A

But it's not a big, a requirement like in the other list from from Sitka me. This was like an or like it just had to be one of the right.

C

This is sort of it, but though this says some of the criteria to consider might be.

F

I think this is also a good signal for the stability of tools that we are exposing under nodejs, so the other way around. Basically it's our commitment that if we expose something as hosted under node.js on 'get up, then we want to make sure that it's actually tested and the tools at least are tier X, yeah.

C

I think that that makes sense, and then this last one the modules identifies an important module. Basically, that's the other out of like what we just think. It's important.

C

So I don't know if we want to leave an equivalent to that one in or not I.

F

Think by the fact that we have already kind of the owners of deciding whether it gets in a lot yeah.

C

F

Think if we can't explain using the others why it's important it might share how.

G

F

G

C

Okay, so that's good.

C

Anything else like that that you see is major thing. Major thing we should add like this will obviously be you know: I'll convert it to a PR where we can add more discussion, but is there anything else like a major thing that we should have in that first cut.

C

I'll fill in a try and put a bit of a description of these the categories and then I, don't know if anybody's had a chance to look at the you know the the ones that were there and like the target, tears that I put them into just as see if people agree with that or not because it was just a first cut off the top of my head.

C

I guess that for this discussion and because we're we only think we only have 20 minutes for the rest of the agenda. It would be mostly like you know, do we think this is close enough to something we should submit a PRN for or they're like? Do we want to schedule some more time to continue discussing it before we do that? I.

F

Think pulling in an a-list early and then iterating on it after it's on github and we can have requests for discussions and people can easy a collaborate, I think that makes sense. Okay,.

C

Now, okay I will take take that as the next step, so I'll just update that yeah so I'll take the action. I'm gonna go back to the minutes.

D

Your support we discussed such reviewed. The draft michael, has action to add.

C

To create PR and I guess the one thing I didn't we didn't talk about, but just quickly. Maybe we should, as the do people think, that the group that we've got the right group, like it's a collaborative decision between the diagnostic working group and the release working group as to what tier something can be in I.

C

Think, that's that's good bad.

C

I guess nobody violently objects, at least so we'll leave it we'll leave it like that for for now, unless somebody has a suggestion for something else,.

C

Okay, so next issue on the list is async hooks, stable, API tracking I, don't know Ali. If you have an update on that or four anything's going on nothing.

E

Has been going on anything books in the past two weeks, so I don't really have an update. Okay,.

C

Do we need to leave that on the agenda or is it I.

B

Think there's some value for Mateo to keep us updated on.

B

What's going on with the PR we're adding a resource to some cooks, okay,.

C

Okay, so I'll leave it on the agenda to get updates from Mateo. So I said: okay, a sync contacts, formalization of.

A

A quick note, I've been playing around with the I, wanted to say, there's a beginning but forgot about it. I've been playing her with embedder API and would actually, if anybody's, interested, like some feedback on a module, lag I'd, try to release with the embed API and we're gonna just paste it here. It's it's for working with new world of threads.

A

So if, if anybody wants to do go with a code will be there, we awesome.

G

Yeah, take that from can I find the chat it's just because we have recently.

A

Or several times asked if people are using it or in in real life and stuff like that, so I thought I might as well give it a go.

D

Okay and paste them that in here right.

D

F

You have any early thoughts on using it yourself.

A

It was a bit confusing exactly what to do. I had to look at another module that was using it to figure it out.

A

Yeah but I mean once I kind of got got, got the gist of it. I think it was easy, but my problem was still I, wasn't really sure. If I did it correctly yeah, it was also why I wanted feedback on it. Yeah.

E

So quick glance through looks pretty nice nimble but and I think this is something we should probably be highlighting more because I think anybody doing cooling themselves. The reason why I think resource exists is so that they can actually communicate with this in cooks and not break context, so so, having example, modules to point to maybe in a direct you know, Docs would be a good thing, and this could be a good model we can point to, and that makes it clear is what was the other module? You looked at that.

A

Was Jas which recently started using it, which is a MongoDB driver? Okay,.

E

C

And that's that's using the new experimental worker support, no.

A

No that's using the embed API, okay,.

C

So the embedder api, so right, okay! So that's that this is not related.

A

To sorry, I probably misunderstood the question original question: what's the question about how it was to work with a new worker API or how it was to work with embed API? Well,.

C

My question was I'm: just clarifying like this is your you're managing a new set of workers that were created with the worker API lately yeah and then it's under async hooks, because the embedder api from async hooks helped. You do that.

A

It's that helps me make sure that when you have a pool of workers to correctly pass make sure that, in the right context, okay.

C

That the contacts and everything.

A

Okay, yeah, so I thought that having a cool water pool would be tagging eyes. Use case for for working with the embed, API I.

E

Think does anybody know why it's called them better api, because I don't see any invaders in this discussion. I, don't know why it's called embed I should have a better name. I. Think.

A

It wasn't once called the providers API. Is it just my brain? That's playing a trick on me. I.

E

Think yeah I, don't know, I, don't know the history I think it's just a third party want you connection propagation or something just not not foreign betters, not necessarily brothers.

D

Agree that doesn't quite yeah.

A

Sure, whatever means, in this context, all.

D

C

So what was your suggestion again? It was like maybe something like I.

E

Don't think about it: I I anemia, the hard for me. So I am I. Don't have a crisp enough. Nice alternative, ok,.

C

I was just putting in, like whatever you'd said, just as like to show that something else might come up, make it come across better. But it's.

E

An API we give to module authors to do context. Propagation is.

A

It only constructs publication, I mean that's, definitely the big use case, of course, but I.

E

Guess we should say an async resource provision, because that's precisely what it does right, it says: here's a resource run this thing in this resource.

C

Okay, so maybe think about it, a bit more but but I just put that as a minutes for people to think about, and maybe we should to make a suggestion on that finance and Ali since you're sort of leading your the champion now for async hooks I assume you can sort of raise that or and whatever, whatever format we need to in terms of deciding. If we should rename the API yeah sounds.

E

Good, it doesn't affect just documentation, changes, yep, I,.

A

Could probably also now that I still have this little bit fish in my mind, Lugo with the documentation and see if it can probably be improved in a few areas to make it easier for other people to use it.

C

A

Of the hot things for me to figure out exactly how I should implement this.

E

I think it would be way quite good if you can, if you're into some example modules in the documentation. So then that that's a really good way to find out how to do it.

G

C

Okay, so any more discussion on that one before I move on to the next issue.

C

Okay, the next one is a sink context: formalization and diagnostic support. I, don't know if there's more for us to discuss today and especially given that we don't have Mike, and we only have nine minutes left but I know, there's still been sort of discussion going on on anything. People want to talk about right now on that front.

D

No okay discussion.

D

C

Go ahead, I know.

A

Sorry I probably skip it, skip that and item, but we didn't do that. Okay,.

C

And then the last item is the trace event. Tracking issue number 84.

E

So it's a game champion for that one. There are a few palookas open. So what continues there some some shutdown issue. Let me see occasionally like if a trace amount had enabled and process the shutdown under gracefully, even bad things happened and I painted the photograph from Eugene, but I think he is currently doing so so so progress has been slow, but there's some okay.

C

And the main, the main thing is I understand to unlock broader community work. Maybe is the intrinsic for logging from JavaScript right yeah.

E

So so James was working on a pull request. I think it's. There were some test failures that were, he was looking into I'm, not sure if he's been able to get to those at all okay. So that's a poor acquire this C alone on the v8 rebo yeah.

C

Okay, yeah I just want to confirm that then the that's, the end of things were actually tagged now. The other one I think I was talking to mike coffman that we wanted to tag that we didn't I'll just paste in there.

D

Diagnostics was this.

C

And then I actually might have one other thing.

C

This guy, which is just you know he a suggestion that hey maybe we should come up with a best practices guide to try and cover, like you know, for the different sort of problems that people run into. What are the different ways that we should you know what tools do we recommend using and sort of a how-to guide?

C

I, don't think we have really time to dive into a lot of discussion, but I think what he wanted was some feedback from this group to say: hey, that's a good idea or there's some fundamental issues that make it not practical. That kind of thing so.

C

If maybe at this point, it's like a call to please take a look at it and comment.

C

And the last thing I wanted to bring up just before we finish this and I just reminded me because see the readers membership readme might need an update, I'm thinking that maybe we need a membership review. We did that in the the the security working group, just by basically sending out an email that says hey, you know we're doing a membership review. If you still plan to be active, let us know if we don't hear from you, you know you wait, you'll be moved to me.

C

Emeritus if you say you're still, gonna be active, you still will be and it was like we're. Gonna send out the an email wait, two weeks and a reminder, err and if we haven't heard anybody from anybody after four weeks, movend emeritus and then I think we even said if you come back within two weeks of that, you know say: wait a sec, I'm still active, we'll just move you back so that sound reasonable to people something we should do.

C

Ice know that you know people people are talking about like hey, we've got thirty people in the working group and there's probably at least a few. We haven't seen for awhile and I bet you if we ask them, if they're gonna have time they'll, probably say no right, I.

A

Think it's a good practice to, as you suggested, to give people a chance to just to raise their hand if they still want to be members.

C

Okay, so if there's no objections, you know I'll raise an issue just saying: hey work. This is what we're good. This is what we talked about doing and as long as then nobody objects in the issue we'll go ahead and and do that make sense, okay sounds good, that's it and we're probably have four minutes anything more other things. I guess we can ask if there's questions on the YouTube stream, see if anybody's watching they can ask questions, I, don't think ice from what I can tell right.

C

Now it tells me one person was watching and there aren't any comments.

C

So if you do, if you are watching and have a comment, you got three seconds and then we'll close out the call.

A

And there's buffering Hey.

C

Okay, we're probably safe now, so thanks for everybody for showing up and yeah I should have mentioned that up front Mike said he couldn't make it today, but I'll obviously be back next time. So talk to you.

D

All day, okay,.

C