AsyncAPI Community, 2 Feb 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: AsyncAPI SIG meeting 41 (February 2, 2021)

Description

This is the recording for the AsyncAPI Special Interest Group (SIG) meeting #41.

Attendees:
We had 17 guests, sorry but didn't have time to write down all names.

Agenda & Notes: https://github.com/asyncapi/asyncapi/issues/480

Chat:
00:23:29 Zoltan: good point!, progress over perfection!
00:24:23 Laurent Broudoux: Issue for examples: https://github.com/asyncapi/asyncapi/issues/329
00:27:36 Laurent Broudoux: Issues on Kafka bindings : https://github.com/asyncapi/bindings/issues/40 and https://github.com/asyncapi/bindings/issues/41
00:38:40 Lukasz Gornicki: https://github.com/asyncapi/asyncapi/pull/468
00:42:57 Lukasz Gornicki: https://github.com/asyncapi/asyncapi/issues/480
00:47:43 Lukasz Gornicki: https://github.com/asyncapi/.github/issues
00:53:49 Martin Fietkiewicz: p’szyz
01:01:35 Kin Lane (Postman): Bye everyone!! ;-)
01:03:42 Ian Cooper: I have to run, will be along again. Thanks, helped to get a feel for the community
01:03:43 Lukasz Gornicki: https://github.com/asyncapi/shape-up-process/issues/38
01:04:33 Marc DiPasquale: Thanks everyone!
01:04:37 Zoltan: Thx!
01:05:35 Laurent Broudoux: Bye everyone!

A

Okay, recording started hey um all of you. uh Welcome.

B

A

Welcome to our um another public meeting, which is also called special interest group asking api meeting. It has different meetings uh depending on the day.

A

I guess we use different name, and I mentioned that- I'm mentioning it because bravano recently asked me about it, so it was hard for me to say what's the name of the meeting and when we use, which name but anyway, it's a public meeting hello to all of you, and if I could ask um people that I've see on the meeting for the first time plus people that anyway would like to introduce themselves um to others. Can we like?

A

Can I I hope you don't mind if I call you out um to the uh wall board, I think that's at least how it was in school. In poland, when teacher was calling you to the wall board, um so we for sure it's for me. It's a first time. I see hugo on the meeting hugo. Would you like to? Are you just uh here to listen the meeting, or would you like to add something before we jump into um the agenda items.

C

I'm mainly just here to listen. This isn't the first time on a meeting. I've got a couple of questions in the issues, but I don't think that's necessary this week, but I'm mainly trying to catch up and get a feel for. What's going on.

A

C

A

Hello again, uh everybody so next person who's who's. Next, who wants to be next one nick, maybe.

D

Yeah sure um so, I'm from ibm and I work with dale lane um and um I'm joining this chord at the moment, mostly to be listening in um the interest I have specifically is describing kafka within async api. um But that's that's not a um concrete interest uh in so much as I will be paying attention to lots of the other ways that async api is being used and developed as well.

D

A

Welcome um martin, yes,.

E

uh Hello, um I'm from abele, I'm the technical content manager there and uh I'm just sitting in to listen fly on the wall. Basically, um I have a question I'll post in slack later, but um but I just just wanted to get a get a feel.

A

Awesome welcome. um Can I ask, I mean I'm gonna, ask a personal question, but you can you can just skip it? You have polish uh roots, maybe because of the.

E

That's, yes, that's a very good question. The answer is yes,.

A

Okay, cool interesting, so we have three people with polish roots on.

E

No, they just get everywhere and we'll keep multiplying.

F

A

That's the plan um and jess. Would you like to also, um I think it's your first time to bring the.

F

Media good morning everybody, my name is jesse manning, I'm in the c2 office at solace uh mark the best quality's uh colleague uh mark has been called to other stuff, so I'm going to take over his participation mark may pop in uh from time to time, but I'm interested in starting getting up to speed a little bit.

B

Oh hi jesse, I'm michael from solids too. I haven't, met you yet. Oh.

F

Hey mike nice to meet.

B

G

Nice to meet you.

F

G

B

On event, designer.

F

A

Okay, so three people from solace at least I can see two people, I think, from red hat too many people from boston and and these two people from ibm. So it's it's good and others. I don't remember the names of the company. Sorry. um So before we jump into the agenda items. Would you like to anyone would like to add something ask something before before we take over the meeting with the official agent items.

A

You sure it's nice last moment, okay, so the first item, it's uh from lauren uh from mike rock's project um I mean, would you like to I mean I can talk? Yes,.

H

I'm going to introduce myself hi everyone, so I am laurent rudoo and, as I said, lucas, I'm uh I'm working at riot on on daytime and at nighttime, I'm the I'm, the leader of the my crocs project. So the the aim of microx is to provide uh two links for uh easily mocking and contract testing apis.

H

So it supports a sync api from last summer and for that it takes care specifically of the examples that you may have added to your specification because it used them to turn them into actionable actionable mocks okay, and for that I've made a feature request some some months ago, and it was about uh having a a formal specification for the the message examples, because at that time still now I would say uh examples are just a map of string and any type.

H

But basically, if you I think, if we want to build some tools upon that to be able to to to extract examples and to turn them into actionable items, we have to specify them a bit more further.

H

So I put this issue on github so some months ago and just wanted to to know how we can proceed on how we can okay, how we can go yes, trying to pushing this in the in the next specification release.

A

Okay, um so the like, um would you like me, fran to answer, and then you just add if I'm missing something or.

G

As you wish, uh if you have an answer for that.

A

So I can give you on this answer from my side and then frank can say something different.

G

A

I mean we didn't prepare the answer. So, let's, let's see who answers what so from. From my perspective, the reason why it's blocked it's, um because we we are now in a position that we should not make any bigger changes to the to the spec, because, as you know, um many of us joined um joined postman.

A

So that's a that's a main blocker and so me and fran. We are now fully focused on on yeah, going to a neutral ground to a um to a foundation, but also at the same time, to to research and and model a basic, uh open governance model that can help us out to kick off and it's not a trivial thingy. I mean from my perspective, especially it's a it's a it's a lot of reading, researching, uh checking out all those different communities how they work.

A

It's um it's a it's causing a lot of headache, and it's like for me personally, it's um after listening to all different stories from different communities. um For us, it's super important to make sure that the open governance model that we come up with is is making sure that we we we all work together on the on the spag and on the tooling and there's no risk that um there can be some dominance of some specific company or some group of interests.

A

So that's this, the governance model and the the foundation. That's that's the main blocker and right after this there's another blocker I would say for for uh adding the support is that we don't have a a a good like written down, release process for the spec. So um for now we were just like I'm, not we actually before to zero. It's just like fran was um accepting changes in the spec after discussions with community etc. But there was no, let's say, official process written down.

A

That would clearly say: okay, that's the stage one stage, two stage three and then go um so that's something that we should have to write down. From my perspective, it's also a blocker, um not sure how how france is it and and others, but um I think we we need clear, like definition and and decision how we want to release in the in the future I'm so the only difference is here that the examples um proposal that you you did, the the proposal for examples object.

A

I think it's not that complicated topic, it's pretty pretty easy, pretty easy change and, from my perspective, it would be awesome to model model and or work on the concept for release process. Basing on this feature as an example, so so we can see how to proceed with this and then release and how, however, we did the process just write it down and that's the official release process, but to do it. um I think it has to be accepted by the um by whatever governance um group will set up after joining the foundation.

A

So that's, um unfortunately, the the blockers that I see, but the the the light in the tunnel is that we're working on it at the moment. Our main main focus.

H

And, and then do you uh do a sorry friend? Do you have an agenda an estimated time of, however, for for setting up this this release process and this this governance model, let's say.

A

um So we don't have any. I mean me personally, I don't have any deadline, but.

A

But for me it's like my focus and it's like uh every day. I I work on it. So it's.

H

A

It's so I would say it's pretty soon, but yeah. It's like priority. Okay, so.

G

Yeah so yeah just to clarify a little bit uh the reason this is a blocker right now and that we cannot stop this from being a blocker is precisely to avoid this company dominance or group of interest dominance uh into the the project uh so modeling the project be before we join a foundation, a neutral ground and and that's why we're like kind of stopping all the important decisions, especially these breaking changes, because that will be a breaking change if I'm not mistaken, right, yeah, it's inspect, so that will be for racing qk3, um but not just there.

G

The thing is that um yeah just to be fair with everyone right and uh I want to be as transparent as possible here uh we are right now discussing joining uh linux foundation right, as some of you know, so we're in the process actually we're working on on joining linux foundation. uh So we decided that it's a it's a great place to be, and it looks like we're going to join an explanation, and I say it look.

G

It looks like because you know it doesn't only depend on us right and many things can happen so yeah.

G

But if everything continues like this, we'll join linux foundation, somehow whatever that means right, uh we need to find yet uh the the format right and how we're gonna do the open governance uh um thing right like what uh what lucas was saying is that we want to come up with a way to to establish governance for the project that is fair for everyone and that um it's not only fair for the companies putting some some dollars there. You know like putting some backs there, it's like just because I'm paying.

G

I can control the direction of the project. uh Now. We also want to give the opportunity to people who is dedicated to the project and dedicating time to to take these decisions so yeah in short money, so not by power here right. That's uh that's what we're trying to uh accomplish and that's why it's so hard because yeah, it's it's freaking hard, because somehow we're doing something new, but it's not really new. There are no gs and other foundations are doing something similar. So so yeah, it's gonna, take some time.

G

Then that's the release process. So, for the like, I I don't think the release process is a blocker. uh It's more of a uh like lucas was saying that we can learn while doing we establish a or we decide into a a release process. We try with uh with a few not just with the examples, feature, uh examples change, but with other changes, and then we try it. We will see how it goes right and we iterate over that, but I think iteration there will keep happening over time, it's just at the beginning.

G

So so that's um that's the the other thing um then, on on the estimated time, uh I have a deadline myself that I don't want this to take more than uh this month. So at the end of this month it needs to be done, like whatever is joining a foundation and having the release process, at least an initial plan on how we'll do it uh at the end of this month and- um and I think it's it's viable- it's not a an eager um estimated time.

G

I will say that it's, uh it's really possible because we're very close to all to finishing all the other, all the other steps and um and yeah other than that. I don't think I think I don't know if we can start paralyzing things already like, for instance, development of the feature.

G

Sorry another feature of the change: the examples: rework reworking how examples work if we can already start moving forward with this and preparing like everything, that's needed for whatever it's ready to to happen.

G

uh I think that's the only thing that we could do right now to to speed things up and- uh and I think it would be a great idea actually but um but yeah um happy to hear from uh from others like uh how do you, if it's uh becoming a pain for you, because I I I don't know, but I feel myself like I'm repeating the same, I was going to say something bad at the same thing,.

G

Meeting over meeting and uh and it's like uh yeah, but I really want this to be unblocked, but uh you know when things grow like it's happening right now, it's really hard to move forward fast, especially when you have to uh change governance, and you want to make sure that you do it well right, because if you mess things up on governance, then you probably mess things up forever, so so yeah. We want to take this step very seriously and that's why it's probably blocking a lot so yeah.

I

So I understand what you're saying about not wanting to make changes until the government stuff is sorted. Is there anything else on these sorts of uh issues around chain? Future changes? Is there anything else we can do to progress it in the meantime to to get to a point where, once the government is sorted, we have a fully specced out fully considered fully debated thing. That's ready for a decision.

G

Well, like I said, for instance, uh with the examples issue, I think individual issues we can get individual issues and move them forward as much as we can and start debating their own issues and coming to a consensus there, and I think that would be.

G

That would be fine, like one of the things that we'll have to do for for governance is.

A

Hey others. Can you hear your friend no, no.

J

C

We lost him yeah again.

G

A

G

We can hear you.

A

Now, yeah, that's good! Okay, that's good! I.

G

Don't know if I okay, I don't know something happened. um So the thing with.

B

A

Whenever you want to say something meaningful, you get cut out.

C

The wrong thing: are we having an intervention to stop you from getting this on the record.

A

G

Okay, so I think you can hear me now.

A

G

Yeah, but I cannot hear you okay now. Yes, I can just take this out.

G

Sorry, so I was saying that um that one of the things that we'll have to do is to do is to designate a technical committee, an initial technical committee or a group of people right. That will decide on what gets included and what not.

G

Damn sorry it's my speakers are changing. All the time like this is becoming crazy uh and sorry I'll. Just try to to finish this sentence and then I'll I'll see what's happening, uh and the thing is that whatever we decide in the issue, we can be sure that this is gonna get into the spec right. uh It's not that we're gonna work on something on specific issues, and then we just we designate a technical committee and suddenly this technical committee says that no this is not gonna get in uh in the end right.

G

So no, this is not going to happen. um I mean the the first reason that that the reason I'm saying this is not going to happen is that the people who are working on on defining or helping or guiding you uh to finishing these issues to solve these issues is most probably the people who are gonna be part of the t of the technical committee right. So so, in the end uh we might have someone else, some extra person or someone else in the beginning, but uh yeah.

G

If everything is almost um designed already and decided, that would be fine, so yeah. So so just uh I, I will say that the best way here is to is to work on on specific issues and try to get them as far as possible.

G

Even if done maybe we don't merge them, but but yeah like even with the pull request or something, and that should be fine.

G

And now I'm gonna see if I can actually hear.

G

I

Yeah I mean so I mean I've sort of got more the the couple of issues that I've raised around the kafka bindings in mind, so because I know that I sort of raised the issue very much as highlighting an area that I thought there was an opportunity, as opposed to proposing a thing that was ready for a decision.

I

um So I was just wondering like how helpful would it be to to try and nudge that further towards being a proposal, instead of just a here's, an idea that needs looking at.

G

Well, I think I think you did the the right the right step right. Let's say it's like just uh highlighting that there is an issue on the on the spec and uh like with the kafka binding and- and I think the next step is probably proposing a solution right. It can be you, it can be me, it can be someone else like whoever wants to work on it, but if you want to take it further, that will be perfect and if you want to propose a solution, that would be awesome even in the same issue.

G

Something one thing that I will recommend is describing your solution there and then, once you know we're discussed there, you can create the pull request or if it's something like something really minimal, maybe you can go straight with the pull request and we discuss there and um yeah whatever works better for you. The only thing we don't want to happen here is that you lose time working on something that is not going to be done. This way.

G

Yeah, sorry, so yeah, but it's something. If it's something minimal, uh then we can discuss on the pull request. If you're happy with that um so yeah hope that makes sense.

A

But you're talking about bindings right.

G

Kafka bindings, yeah, yeah kafka authorization. I think right. Yeah.

H

Yeah this was the the other parts of my uh agenda suggestion because yeah, we are also a rfe on kafka by innings too.

H

So, okay, if I understand correctly so we'll continue discussing on the issue and then pushing pull requests once we have some kind of agreement around the solution and then we'll see after the the the government's committee set up how to validate this and how to merge the changes within the next release. Okay, makes sense.

J

I think I think, if I can, if I can add something, I think that this kind of things are actually good previews for this kind of governance model. So I think that's what friends trying to say that we shouldn't we shouldn't stop, because there is.

I

J

Yes, so I think the community is working and there are some great suggestions and findings, and I think we shouldn't stop that so and that's that's. We have to exercise and of course there is a point because of the change that maybe we cannot push hundred percent. But let's try I mean, maybe that will be the first version of that. That's beautiful and we say that okay by agreeing in the community as we as you guys, work and we working there. So that's what we do.

J

I think another thing I think friend again: it's like communication. Yes, so if the timeline is more visible, sorry I'm also repeating myself, but maybe that's how what we do. Then people are also kind of more confident at what.

G

Absolutely, and actually um I I was going to say like something that I'm going to work now, but I can't work now, um even if I wish.

G

The next thing that I have in in my to the list is in in, and it's sorted by priority is a crit to create and publish this roadmap right and um and also I I've been investigating a little bit more about creating roadmaps, because I usually don't like the kind of roadmaps that you usually find everywhere, like this straight line that spans over time, and that puts some things one after the other right and um that it's so difficult to commit to because yeah.

I

G

Knows what you're gonna um do in six months from now, let's say or priorities must probably will shift, uh because you learn more and then you decide to do uh to re-prioritize things right. So so I'm working on a new kind of roadmap, which is not invented by myself, but more in a in a roadmap that you know in a format that it's visible.

G

What's coming what's coming next and then what probably will come next uh after that right in terms of priority and all the stuff right, so so yeah. I hope to have this ready, like a myth of uh this month, to have this kind of road map. uh It's gonna be hard because of uh what luca said that we're working on the governance as well and uh and yeah. This is gonna, be hard, but yeah. um More visibility is something that we have in mind and and that we want to.

G

We want to do more transparency if you want yeah.

J

One other thing I I came into my mind just looking at these issues that one of the on one part is like features and and and some new things that out changes, but as you mentioned that sometimes our fixes- yes, so maybe spec fixes and something- and there is always there is always a better path to do that. Yes, because if it's a fix, then it could be harmful or painful for people.

J

So that's also could be an interesting thing to say that what if there is a possibility to distinguish between these things, yes again in a in a release process, sometimes you say: okay, it's a bug fix, so there is a way of fixing a buff on a spec or it's a it's a feature. So when that also could be something that kind of qualifying the question. So not everything is just flat and have to go through the same governance process, but it's a fix. So we act. We agree or you guys agree on.

J

Okay, this is really about. So we have we recognize. This is a as a bug, not a feature yeah. So then it could go through on this temporary time and of course, but it was of course it could be a thinking also for the future that it should be different things. Yes, as a normal release procedure that there are different things as major minor things, minor changes and so on and so on. So maybe that's also something you can you can say that?

J

Okay, that's how it goes or that's how anyways it's just about because of the uncertainty, a little bit of uncertainty right now, people are kind of yeah, that's it so.

G

Yeah, no, no, no, that that's that's a really great suggestion and I think uh what lucas has been doing on all the reapers of the of the organization which actually you're gonna show later right. I think um so lucas has been uh automating the way we release on on all the repos, and uh what was the name of this? uh I can't remember lucas what was the name of this syntax that we have conventional comments like conventional comments exactly so we were using conventional comments and semantic release, or something like this. This is back.

G

This is a package right and, um and so then you tag your pull request as if this is a bug fix or if it's a minor or a major right um I mean the the problem with this is: is that it's it's a spec right, so we we want to group changes hms as much as possible so uh and and release, or I mean this is something that we have to think in the end, because if you think about it as well, it's like why do we have to group changes at all right?

G

Why don't we do continuous uh release? And then it's just a matter of uh I mean just treat the the spec as a software right as any other software. uh It doesn't mean that, because we have a new version of the spec, you have to instantly adopt it right, but um like every every x amount of time, then uh we decide that uh all the tooling should be updated to the new version of the spec and blah blah blah.

G

You know I mean this is something that we have to discuss like how we're gonna do the release process, like you like we were saying and um but yeah for sure. If we end up deciding that we group minor changes and major changes into a single big release of the spec that shouldn't block from releasing patches, you know small patches and fixes, so fixes will be like immediate as soon as as soon as it's merged there's a new version of the spec.

G

I don't know three point zero point, eight whatever, because we do a lot of mistakes, so so so yeah. uh So that would be that that's the be a blocker but yeah again. This is a release process.

G

This is something that we have to decide and to work on and and- and I feel that we we could decide this ourselves now, but right before we create a technical committee right, uh but is it fair for the technical committee then working on that like uh why don't allow these people joining officially the project to think technical committee to decide this themselves right as part of this, so as part of this being the technical committee right so yeah? So so I hope this.

G

This makes sense.

J

Maybe you have you have to you, have to start calling these recommendations for and not decisions, and then you are in the same.

C

Recommendation.

J

Right, it's a recommendation.

J

It doesn't matter.

C

Sometimes the terminology.

J

Is important you know.

I

If you want to, if you want to.

J

Treat that upcoming technical community right so terminology is everything: yes, because that's.

I

J

How it is so that's how we give our empower that coming thing, but I think I think, from from from a decision-making point of view, it's just like progress. That's over perfection, so it's not like you have to find out sit down now and in two weeks have sort out everything. Yes, it's like having the intentions on place and and and helping us helping people who sitting there to do that. Like this, like this in this meeting, I think that's just perfect. Yes, because now we have a. We have an insight.

J

People have an insight how this video works or works already, maybe maybe not documented, not communicated not very nicely put, but I mean it is here it's out. So it's okay again, it's part of it. I think part of the journey.

G

Yeah, no, no, I I totally agree, but we uh sometimes it looks like we're talking about this technical committee, future technical committee and ourselves as different totally different people, but think that uh most probably technical committee will be ourselves uh plus some other people.

G

So in the end, so in the end, it's the same people uh and we'll make proposals. Of course uh we have ideas, we'll make this not suggestions we'll well. In the end, there are suggestions, of course, because they have to be voted but we'll make proposals, and the committee will have to vote right and uh we may end up having multiple proposals for their release. uh Release cycles right so yeah.

J

A

And, and just that um I think, like fran said we also can do work in parallel, so it doesn't mean that, like, if, if me and front, are now focused on open governance, then scratching gabriel. This process has to wait for us. um It's like um it's like with the uh bindings. What what fran said that you can already start pushing those topics, create pull requests and get the stuff ready.

A

um The same with the release process, I mean um during the uh the meeting I was looking for, one pull request, which um I think will be pretty interesting for you.

A

So in december last year, walid from our community already wanted to push forward the the topic of releases and release, candidates etc, and he opened apr and we had a longer discussion in the pr and uh already we kind of had a brain fart of all the ideas answer, questions and answers um that we that we have to face while working on the release.

A

So if there's already in this audience on this call someone that would like to start pushing forward this topic uh start answering those questions. Giving suggestions then feel free to take this pull request as a reference open, a and a separate issue in nothing, api, repo and let's, let's, uh let's let us already start discussion and proposals how certain uh certain steps can be solved.

A

Why not I mean at the end um for us it's it's very hard to jump into it and start driving the topic, but if somebody drives it and and we're just commenting and and giving questions and answers, then then we should be able to also in parallel drive the release process topic as well.

G

Yeah and just to add to that is the more the more you uh participate on that, the less it will be about us, and it will be more about the community right because, right now it looks like the group uh uh working at postman um like it's leading the the future of the project and and deciding everything right and- and we want this group to be more diverse right uh more in terms of people, not only people working at postman. We love our people working at postman.

G

I'm saying this because king is here. Otherwise it's not.

G

So um but yeah this needs to be a more healthy group and diverse right, so so yeah.

G

So I'll really recommend that you make some proposals and stuff, even if it's just triggering the conversation right, it doesn't have to be a formal proposal with lots of text or anything just raising the topic and uh and triggering maybe just uh describing a little bit the idea that you have in mind. If you have someone and and from there we we continue right.

G

Sometimes that's uh that works better than just waiting to have the perfect idea in your mind and the perfect text to because then it will never happen right. You will never do it so yeah.

A

Okay, anything else in regards to this first object in the item: can we jump to another one yeah, okay, cool?

A

uh So next one is boring comparing to the previous discussion, um so I just want to give you a um some summary of uh of changes that were done over a couple of over the last couple of weeks on in in a ci related async api organization.

A

So I'm not gonna talk about everything. I'm gonna talk about changes, important changes for contributors, uh because committers maintainers, most of them already like all of them, already know uh what changed except of michael, because michael mentioned at the beginning of the call that um he will contribute back a few features again so michael, if you will have like more questions after what I say then feel free to to ping me also um on slack or just during development.

A

If, if you need some help so um regarding the changes, I've already put the the agenda, uh the the notes for the for the meeting in the issue. So I just pasted the link to the issue in the chat, um so one the most important change um or not- maybe a change, but something that we now made official is what also already front mentioned so last year, march or april april. I think we started using conventional comets um for generator.

A

Then we started using it for parser and then other projects mainly to to be able to automate releases uh that worked out pretty well so since last or last week or two weeks ago now we have it like official. Every pr is um is checked against the conventional combat.

A

So if the pr title does not meet the conventional comments, where we're blocking the release- and I don't want to now- go into the topic of the release process again, but just to mention that conventional comments is flexible enough, that we can even use it for the spec, because it has also the concept of scopes, etc. So it's um it's it's I think it's pretty doable, maybe not with semantic versioning package but we're developers. We can write our own package, another interesting topic for contributors.

A

I think, because in the past we had few contributors using windows and they were facing regularly issues with using projects, building projects locally on machines, different that mark then max. um So now we are testing um on aci all the scripts, using all the scripts um on all possible platforms. So it's mac linux and windows.

A

We solved a number of like issues that we faced on windows with our scripts and now they're passing, which means like doesn't matter what platform you're having if you're, um if you're gonna need to run some npm scripts on your local machine and you're, a windows developer and not windows, developer developer, using windows, you should be happy and if you're happy, then we're happy, then also something that I asked community a few months back.

A

um We started we wanted to like inform you about new releases on different channels, and we picked that twitter is also a good channel uh to automatically let you know that we have a new release of some package of some of some in some repository. So now it's also done for all the repos.

A

So as an and as agreed with community, every major and minor release will go uh automatically to to twitter, so you know it happened. um Plus all the releases not only minor and major but also the and the patches will be, uh will go to the slack channel github releases. So if you want to listen, uh you can join the channel and and monitor um this information.

A

If you think we should post it to some other channels, just let let us know we can. We can always configure that. Also. We have a separate channel for new pr's and issues. So if you want to be more active in the project, you want to see what's happening, where prs are being created, uh what issues are are added to any repository under async api?

A

Then you can also join this dedicated channel, where you can see all the notifications about only about new new issues um and also the I mean issues that are not created by bots, but I mean like real humans um and there's a dedicated channel, which name is pretty difficult. It's all in the link I've sent to you. I mean difficult difficult for me to spell and for potential contributors, so, if um like contributing to asking api, is not only code or spec if someone wants to contribute by um having fun with ci cd setup.

A

So after the refactoring of the ci, there are two issues left, um which are also important for the project that someone can jump into these topics and then such solutions and even do the configuration. I offer, of course, my help to to onboard someone. So there are two issues. One is, of course, I'm gonna. Send you the link. First.

A

One is that we would like to start automatically uh do the test coverage testing um on in our projects um plus have this testing on pr's level so um be professional enough that on apr, when there's a new feature added, we don't go, don't go down with the quality of test coverage, uh so I think it's pretty interesting topic to work on uh with google acc with github actions, of course, and that was my wife in the background and the the other one.

A

I think it's much more cooler, especially that discussion that has to be done around the topic, so we want to automate the way you can merge changes, because we have many projects, many many code repositories um and to to be able to emerge the the pull request we uh you have to be a member of the organization, have right access to the into the repo etc, um and at the moment we have this problem that we can only uh we are accepting.

A

Whoever is in the organization we give people bright access to the organization with which means people have right access to to all repos, which should not uh not happen by default. That's why front was recently doing some cleanup, uh so we can be like safe here until we're done with this um this workflow, so we want to basically using github actions, have some kind of automated way of merging prs.

A

So if the ci does all checks like pr title, linked testing goes through um test coverage is okay, etc, etc, and then one of the maintainers like in kubernetes does um okay to merge, comment or add a label okay to merge. Then the merge should happen um by a bot um in an automated way.

A

So it's much much easier than like, remembering to add people to teams and then manage those rights and um believe me, it can take a lot of time like if, if you want to manually, manage the organization, um then it gets complicated and we just want to automate it with with ci and github actions. And if someone wants to take this over, then again I'm happy to onboard you and that's it if it comes to the changes in the ci.

A

If you have any questions, then please ask.

G

If, well, I don't have a question, I don't. I just want to highlight something that you did and you didn't mention is that uh lucas has been doing an amazing job, cleaning up all the repositories in the organization. So uh thanks for that man, because this isn't- this has been like a very detailed and exhausting job.

G

So thanks for doing this, you know little by little on each reaper and and also uh something amazing that he did and it's not and that can be useful for other organizations is the way we handle um the global workflows. I think that's the way you called it right like uh or yeah.

A

Yeah global workforce.

G

So maybe you can explain what is it about uh better than myself? I don't know.

A

Thank you, so I'm not sure how much you're familiar with with github actions, but I think it's also concerning the circle ci or travis ci, all or all the ci systems that are based on the file configuration that you have in your repo. So we we faced many times during last month's working with github actions, issues like, for example, with pull request action in the release. Workflow, we have a pull request and we use a pull request.

A

Action that, at the end of the release, creates a pull request with the with a change in package.json um to the repo. So then automatically we bump the the version impact json um in the repo and the problem was that this action got broken and many other actions also had the same issue over the last couple of months, and this was causing issues on the ci.

A

But to fix uh these issues, we had to bump the action that we use, which means we have to update the version of the action that we use in those configurations, files in 20 repositories or 30 repositories.

A

So you can imagine like doing 30 pr see that I mean first time it might be fun next time, you're starting getting depression, and that was in my case and my my psychiatrist told me that it's better to automate. So I I simply created an a github action that it works in the way that we we have a workflow that we put in dot github repository and this. This action grabs the the change checks. What repositories we have in the organization.

A

And push pull push pushes puss push. Push is pushing um this change to um to these repos uh through a usual workflow, so it grabs the repo makes a change creates a pull request and then we have auto merge workflow. That is merging the changes. So we change in one place and then it's it's done in all the repos.

A

So if you're interested, I can later on, share the link to the action with you and tell you how it works, but it's it. It helped us.

A

It helped us a lot um now. It's it really helped helped any other topics, except of my poor english.

A

We have eight minutes till the end, so.

G

Well, I would love to to raise a question here, for everyone is that how how can we make it easier for you to to start contributing right, because we we see, uh we see many people joining this the skulls, especially today right. I think we had a record today, uh but um but it's it's always hard and and I and and I get many private messages hey, I want to contribute.

G

How can I get started right and uh we always do it like in a kind of manual process, because the project is evolving so fast that whatever we do is not is not valid in probably six months from now so um so I was wondering: how can we improve this process or maybe what's stopping you from contributing, or I don't know if you have any ideas of what can we do? What can we be doing better right now uh that we're not doing.

J

I can speak only myself for myself, I said to you also, but I'm saying here, I'm really considering I'm trying to go into into the space and and and start contributing with my contribution. uh It's it's restricted right now to participate here and trying to give my input right now.

J

I don't think I don't think I missed something from you guys. I think you you're great. You have created a tremendous job here. I can also just say to myself. I think it's difficult to find a time for the contribution from my side. So it's nothing. I wouldn't say that it's something nothing hindering that I can start like today. You just listed I like five or six areas.

J

There are, and there are some very interesting one, I'm very interested, it's just a question of how how it starts, but I think you have the slack you you promoted that I think there is a very good discussions there. That's that's that's just can you can just go in and I can go in. I think it's just a matter of I'm dropping some of the boring stuff from my work and doing this one. Instead of that, that is going forward. So that's my point, that's that's my.

J

Do it so ditch me more and ask me more like this one and then I I feel the urge to start doing it. I think, but I speak myself so so that's my and I really.

I

J

Like to participate on on reference for the spec or the architecture side or even on on other other aspects, so I'm ready to do that. I just need to find the right time and.

K

That's what what's what's your biggest blocker with your time? Is it is it? uh Is it in your head or is it your boss's head, or is it a little bit of both? What what are some of the things that we could do to to make things easier for you to to carve out time.

J

I think it's just personal priority. I I think it's it's like that. It's like you, have to be in this corona times everything happening in parallel. You can see that I'm looking outside. That means that I have family here, they're coming back and forth and doing that and then another monitor monitors just still work. So I think that's that's the best. That's it's just personal. I think from you from you guys. I think the value proposition what's coming out as a communication from the from a from an employer point of view is just enough.

J

I mean it shows that there is something here, it's important. So if that's question of priority, that's I think it's not. There is no issue that I can defend from my employer side if it's if it's counts, but I think it's more like personal issue for my side, but I will do that so trust me. I committed this, so I want to do this so no worries.

K

Well, I'm going to keep asking over time, because I I mean I have a list of things that I got to do every day. That's like this long and I I get the time to do this much of my list. So it's always about prioritization for me and then tricking myself into believing well. I've got to do that because that's really important because either my boss is is on me for it or usually it's not that usually it's me.

K

I have to say this is super important and convince myself, so I'm going to always be looking ways at convincing y'all um and try to understand where the friction is in the process. So um I'm doing this for the oai right now as well. Try we're talking about adding a piece into the the um the the contribution flow and there was. There was questions like. Oh, it's already so much friction it's so hard to get people involved.

K

Adding another piece is just going to make it worse, and so I'm trying to understand what that process is so I'll I'll keep working with fran and and and lucas to understand the process, but uh I'll keep trying to find ways to incentivize y'all and get people contributing.

B

Yeah and I'd say from my perspective, I think fran mentioned earlier creation of a new road map. I think that would help because it would help people realize what the top priorities are coming up for the initiative as a whole right. So I think people want to contribute to something they know is meaningful to the community. um So I think that will help and then also maybe on like the the async api webpage or something saying like how people can contribute right other than I think right now.

B

It just says you can join this join slack. But if you said like hey head over here and like here are like some of our top priorities that we really want to do, they didn't fit into our plan for the next month or two, but if somebody can contribute an ad to those like, we would like like it.

B

If you did those- and maybe I know in some of the repos, there are labels for like good first issue- um I don't know if it's like that for all, but maybe have those available on a page or something so people can easily find them and they don't have to go like repo to repo to find them. Those are just maybe a few suggestions that might help people find where they want to start.

K

B

Knowing what they want to do.

K

I would add to that maybe a visual to the website or you know that's like here's the process in a visual, because we all respond well to visual cues.

G

I actually like it a lot, this idea of having a central place for all the good first issues. We have this label in many many repos and we've been using it but yeah it's a pain to go through all of them and also understanding the whole organization. How many repos we have, what everything every piece is doing. How are they interconnected uh lukas created a map uh some time ago, where you could see how everything is connected?

G

um I think it was it was you or someone uh who contributed this map, but yeah anyway, there's these maps will be visible as well somewhere, because if I had to find it myself now, I will be. I will struggle. So that's not good.

G

If, if myself, who I'm working a hundred percent on this struggle to find things, maybe it will be uh easier to find and we we have to rearrange uh yeah in the end, it's about communicating better right, so yeah and and and the chicken neck problem that I'm seeing here with the community wanting uh wanting to find uh what's uh what contribution can be meaningful?

G

Is that for us what's meaningful for the community? It's probably needs to be suggested by the community right, so if the community doesn't tell us what's meaningful for you right individually like if, if you don't tell us like what's meaningful for you, uh then it's hard for us to understand what's meaningful but uh yeah. I think there are like two.

G

I can see that there are like at least two kind of uh contributor um two kinds of contributor here.

G

One is the the contributor who has experience and then wants to propose changes or or race issues, or things like this and then there's the other kind of contributor with who is more, like hey, just give me some stuff to do and I'll do it, because I just want to get involved into open source, for instance right um so yeah, so for that second kind that uh that probe, that's a second kind of uh contributor, probably is not necessary to give their opinion first. They just want to do something right.

G

uh They just want to get started with something and feeling that they're doing something meaningful, um so yeah great great suggestions. Thanks for that.

A

And yes, just to add, I mean I I I tried to do this contribution guide in the past, so there are some google docs. um Then then then, there's this map, but this map is rather overwhelming than then I don't think it should be shown to new people to not scare them off, um but yeah um there's.

A

I I just recently shared it. So I I have a a one issue somewhere in the repo where I list all the topics that I'm I'm considering as my priority list so and at the moment I'm adding um this clear contribution guide at the end of the list. So I'm gonna put some description to it.

A

And you can vote up in the comments if you think I should move it up before some other topics.

A

And if someone wants to help them, that would be even better okay. So it's uh two minutes past six. um Some of you probably have another meetings. um Thanks a lot for joining it was, I think it's we beat the record today, even during the postman uh partnership announcement. I don't think we had so many people, no.

J

I don't think so: yeah.

A

So it's good, uh it's really good! The year starts really really good. Thanks a lot again and hear you in two weeks in two weeks, of course, we have a different uh time for the meeting, which is much more apart. Friendly 8 am utc.

G

Yep and remember, we have postman galaxy tomorrow. Actually it started today, but there are no thoughts today, so so yeah only workshops so yeah we have pokemon galaxy and we'll I'll I'll be speaking there, but there will be a bunch of cool uh talks, so I and it's free you can register for free. So I really recommend having a look at the at the schedule. That's really it's really great nice and stuff bye bye. Thank you, bye-bye. Thank you. Bye, bye, everyone.

G