GitHub Open Source Friday, 7 Jun 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: GraphQL Mesh — Query anything, run anywhere - Open Source Friday

Description

A powerful open-source library enabling developers to leverage the power of GraphQL without requiring changes to the technologies your servers already use.

Join @bdougieYO and @UriGoldshtein in bringing a different approach for sustainable open source while significantly improving areas in GraphQL gateways.

https://the-guild.dev/

0:00 - Start
0:16 - What is the Guild?
4:19 - The Guild members
11:00 - Decentralised GraphQL tool libraries
17:23 - GraphQL Config, the future of GraphQL tooling
22:55 - GraphQL Mesh - fetch any data source
30:48 - Envelop - share collaborate your GraphQL extension layer
43:45 - Making contributions to the Guild

As always, feel free to leave us a comment below and don't forget to subscribe: http://bit.ly/subgithub

Thanks!

Connect with us.
Facebook: http://fb.com/github
Twitter: http://twitter.com/github
LinkedIn: http://linkedin.com/company/github

About GitHub
GitHub is the best place to share code with friends, co-workers, classmates, and complete strangers. Millions of people use GitHub to build amazing things together. For more info, go to http://github.com

A

All right welcome everybody to open source friday, an opportunity for us to talk to open source, maintainers contributors alike, uh about the projects they work on and uh also a little bit of how they they got started and what they're working on next and today we've got a special guest, and that is uh your gold team. So you want to say hello, hi,.

B

um Thanks for having me, I'm really excited to be here. uh I'm uli, I am a member of a group called the guild. um We do a bunch of open source, uh mostly around graphql.

B

Maybe there's like a the way that open source is being um developed today by you know, or large organizations or um startup companies, maybe there's another way that could be complementary to all of this. um So I started this group called the guild and the original and the idea there is that it's like a group of individuals that does open source so um yeah, that's our website. It's like we didn't have a website for very long, um so we have all kinds of weird ideas.

B

uh For example, one of the ideas is that all of our libraries- that probably some of them are famous- and you heard about- are not under the guild's name but under individuals uh names so um like you can go.

B

There's like a you click this platform thing then there's like a list of all of the individual projects we're working on. If you click on any of them, let's say I know graphql code generator. For example, it's a bit up, I think, probably yeah.

B

Then this project, if you go to the github page, um it's basically under dotan um under the town's name and um there's many reasons for that. But um you know the idea is to power the individuals and to make it more sustainable, meaning like if something happens to the guild as a group or to the company um yeah.

B

The individual still has the responsibility like I've, seen just so many times where the company just companies just deserted open source um and- um and there was no individual responsibility and, together with the responsibility, comes the benefits as well. I mean graphql code generator if it was under just the guild.

B

Many there were less people who would knew about the tan, but the.

A

B

Is making the work so yeah? So that's how we started and it started from something very, very small and today I think we're like the largest open source group in graphql. um So it's really crazy yeah and how.

A

Do you sit because I know the graphql foundation that got thrown together a couple years ago.

B

Are you a participant.

A

In the foundation, or do you collaborate with them.

B

I think about a year ago, or something like that. um I saw that there was actually an opportunity to join, uh so we just you need to pay some money and then you're part of it. So I was like okay, cool.

A

B

Is that uh um I think if you go to graphql.org yeah, we just um we actually helped rebuild the website. It used to be uh uh now it's a gatsby website and it's much easier to contribute. So we did the arda from um the guild actually uh did the most of the rebuilding yeah um and yeah.

B

If you go down to the foundation section, you can see, like um you know, all the foundations work which is really exciting and for us we before that we were like yeah, you see like all kinds of big companies there, and then you see us which is pretty cool yeah. So, um yes, it was nice to like join all these big companies and actually realize that if you want, you can really affect and change things. Yeah.

A

B

um For us before that, it seems like it's very hard to change and there's like special people there out there that are doing special things and um now, like, I think, it's actually relatively accessible and easy. Once you realize you can do that.

A

Yeah right so can we talk more more about the structure of the guild and the the members, um so how big? How big is the guild today.

B

um It's um it's a it's a hard question to answer, um because uh um so um I think uh full-time people that are like actively working on open source without interruptions, we're around eight or nine right and then there's all kinds of, uh but we also support the ideas we support anyone, who's who's, doing open source. That is interesting, so there's a lot of people that are doing really cool stuff uh in open source and we support them with whatever they need. So you know they. We help contribute to their libraries.

B

We help publish their work. We give them feedback, it's kind of like the an extended group. um So you know a lot of the library the new library is now in the ecosystem, like there's a new websocket library, there's a new um transports library. So we we try to support these individuals because they just do great work. So we reach out- and we say how could we help?

B

um Because that's the goal like the goal of the the the guild is not necessarily um you know it's funny, because our goal wasn't to be the biggest open source group in graphql. The goal was um to actually just unstuck things that are stuck, and somehow there were a lot of things that were stuck so wow.

A

And and grab you all in particular, like, I think, the biggest thing that was like the the hardest thing. I just watched a uh conversation this morning with james q, quick about super bass and their decision to go with the rest, and I think the the hardest thing that folks to adopt graphql was the tooling, and that was the conversation for the past like the first couple years and it's still a conversation right now.

A

It's like yeah graphql is great, but like the tooling's, not there and when you look at rest, rest has got so many more years ahead of graphql. So.

B

A

You say unstuck: are you building general tools and ecosystem thing like pieces and legos to help people get involved in graphql.

B

Yeah exactly it's like we, we just um you know every day we work with like companies and and that's also part of this open source cycle yeah. You can see here a bunch of them like this is an outdated list, but there's like more now and the idea was also yeah. It's not bad. It's like um it's even there's even more impressive stuff there now, but the idea is like we thought we built this kind of cycle where um everything we do is then triggering the next step. So let's say every work.

B

We do with a company that feeds back to open source that feeds back to the community that feeds back more clients and it's kind of like it's this cycle um and everything we're doing we're thinking about that cycle and um so that cycle. Also, when you work on something and you get stuck, um you know, then we try to solve it and then open source it and then build a community around it.

B

So I think the idea that um what we saw by the way a lot of people are saying, like you said like the toolings, are behind, but I think also- and maybe that's also a bit our fault because, like all the tools are scattered around, um we saw a lot of people that are just unaware of the toolings that are there.

B

So that's why we created the website actually, because all the tools were under individual people, but then we saw someone who's using the kojan and they were super happy, uh but they didn't even know that I don't know, graphql inspector exists or graphql mesh exists.

B

So one of the things we've done when we joined the foundation like if you can even go to the let's say to the graphql.org back to the graphical.org uh website, then there's uh a code section on the top. So it completely rebuilt that because um and also the design was by orta from microsoft, yeah.

A

Yeah so now it's.

B

Much easier yeah, so now it's much easier to find the two links. We also added here kind of like a sorting algorithm to make sure you know. If libraries are unmaintained, they would go down on the list.

B

You know um so there's all kinds of cool stuff there um and the idea is to just make it easier for people uh to find the right tools um so yeah so and if they're not existing, then well, if some of the toolings, um but by the way, also all that our work started from pr's, but some of our pr's just got an ex haven't accepted because the libraries were like, let's say unmaintained anymore, so we kind of like were forced to create uh the libraries um yeah so um yeah, maybe hopefully now.

B

I think um the ecosystem is more for. Like there's, I I feel like today, if you're starting um it's a completely different world than even like six months or a year. You know like there's so many new things, and not only by us, but so by so many people, and I I feel like we did like a step change recently like now. The unblocking became like much different, um so yeah- I I hope, maybe even through that today, like people would find um like if you get unblock if you get blocked by graphql toolings.

B

You have someone to reach out to let's say like that:.

A

Yeah yeah and it's good to see like there I mean there's a couple of them that I've never heard of, I didn't realize near form, had a graphql um hooks library, mm-hmm, yeah, yeah, yeah yeah, but yeah. This is something that I would actually look into that I wasn't aware of and like a lot of stuff like we.

B

A

Off off before we hit stream or start on the stream about conferences and how like streaming, has now taken the place of a lot of the big conferences, because we would run into each other quite a bit at conferences yeah. um But I would have to wait for a conference talk to hear about this thing, but now on a stream, I'm now educated, and I, if I want a small little hooks library to use my with my graphql. I might give this a try.

A

I can also see that there was a release four days ago, so I can have some confidence.

B

A

That whoever shipped this release um perhaps can answer my questions or at least there's somebody who I can sort of jump in and have that conversation with.

B

Yeah and- and I think what we're now start, this is by the way, maybe one of the biggest reasons we joined the foundation because we just thought about the journey of uh like when I mentioned that cycle.

B

We thought about the journey of a beginner uh into graphql, and until then we just met people when they already knew about, let's say the graphql code, generator that total wrote or something. But at that point they're already have some experience. um We thought well, where are they starting from like? Where is the the first place?

B

Someone gets into graphql in in that context, and let's try to look from those uh from this starter and I think that's also kind of like how individual contributors are sometimes more powerful than companies, because companies will invest a lot on their website and they will invest a lot to make the beginners starting on their own websites for us and that's not necessarily the right place for them to start right. They will fight over google results. You know who's going to be first or something like that.

B

But for us we just we don't care, so um we can just contribute it to the graphql.org, but we have no stake in it. We have no ownership of it, but that's where people start yeah.

A

Going back to like our your example of uh like the tan, like I haven't heard, of the thing.

B

He's a real avenger yeah he's.

A

A real avenger.

B

I mean I mean if you go to the guild org on github, um so um you see there's like just individual people, and we have some stuff here that we could get to on the org itself. But those are not libraries. Those are like, let's say the web. Our website is open source like so, for example, our website is not beautiful as you've seen. So if someone wants to help us like it's open source and you can help, um but those are like just. I know some random stuff that we have there, but.

A

B

The libraries all the famous libraries are under the individual, so you, but you can go to the org and then you see the individuals. So that's the most important most thing to see them. And if you go to the website, you can see like the list of libraries, yeah.

A

Yeah now I wonder.

B

A

The because, like I I'm kind of like dumbfounded, because I've known about the guild for a while I've looked at the website, I look at the org every now when it comes in conversation. I'm like I, don't know what they're working on there's like nothing really here, and that was my assumption.

B

A

The last year, the last year and a half until we now are having.

B

This conversation.

A

um But you kept coming up as like a thing in graphql and that you all were working on stuff, and I would hear about your contributions to other projects, um but not realizing that when I look at the tons like I didn't realize this was part of the the guild. I thought this was just like one off a person. um You know me.

B

A

Reading the uh the entire readme or anything like that to try to eventually.

B

A

B

It's a testament.

A

About being an avenger and you're, all just trying to level up space fill in the pieces.

B

A

B

I I think, there's a there's a you know, sometimes now we're kind of like trying.

B

You know if some, if people want to follow us or see everything, that's happening because there's a lot of things, that's happening, then you know we we put on the website and on the website, there's like a blog and the blog is also open source. So we get tons of contributions to the blog. So if someone wants to blog about something, um they could actually contribute to the blog, but you know so now you can actually follow us and you can.

B

You know, um there's even like a twitter account now and there's like you could actually like follow the group, but the twitter account, for example, just retweets the other members yeah, because just to keep the the focus on the people.

A

Yeah, it's a little more decentralized. So that way, there's like if.

B

You decide to take.

A

A role at another company and remove yourself from the guild, it can still proceed and operate without your your guidance and uh contribution.

B

Exactly and and it's very powerful because we can also do things that I guess for a company would wouldn't be beneficial, but then they won't have the freedom to do it, for example, contributing to graphql.org, which will maybe take away.

B

um I don't know visitors to our website or to our products or whatever, or that we would um support libraries and and other contributors that are, um I don't know that are not under the guild, so we could do those things and it gives us a lot of freedom um yeah, it's just more fun. I think, for the day to day also.

A

Yeah yeah, and I just want to take a quick pause too as well uh folks, are in the chat and having conversation.

A

It seems like a lot of people who are interested are aware of the guild as well uh and are.

B

Quite knowledgeable.

A

More knowledgeable than than I am at the moment so.

B

A

If you have questions about different projects,.

B

A

uh About the guild, please drop in the chat. um I have no idea about your private repos, but if you want to invite me on b dougie on github by the way, um I'm happy to take a look um but the going back to the um so going back to the guild. What are some notable projects? We can sort of talk about and share and sort of highlight what we we have you on the chat.

B

Yeah, so um I think, uh let's see, maybe you can go to yeah our services and then you can just um so. um Let's see so I think first thing that um so there's a couple of different projects here by the way, a lot of projects. Here we actually some projects here we took over from other companies and maintainers because they stepped down or they just died down yeah, but there.

A

Was a need for someone to step up yeah, one of them that I I folks if you've been in graphql for a while, we know that prisma used to be graph cool.

A

They haven't stepped away from all graphql stuff, but they've definitely have pivoted into like a broader database operation, operation toolset in uh best tooling. So like did you absorb some of their projects and that sort of transition.

B

Yeah so so, um let's see so we took from pris, we helped prisma transition and kind of rebuild. um uh That was uh a graphql config.

B

So one of the things that were very annoying at the beginning- and we were talking about tooling, is that- is that um you, you have a lot of different toolings for graphql, but then each of the two links has their own um configurations. Yeah.

A

B

Share the same thing, so we just grab with graphical config. We just we took all of our libraries and then also other people's libraries and we created a shared config. Okay. um So now you can, you know, configure all your projects into one uh all your configurations into one file and that file could be. You know some. We prefer yaml, but people hate yaml. So you could basically there's like a library you could use whatever you want.

B

You know like json or you know, whatever format you want um and now it's all in one place, so the cogen config and like the inspector graphical inspector config and the mesh config and all of those things could be in one place um and also it's kind of like it became we white labeled it so like gatsby could take it and then call it gatsby country, you know or something like that: they're not necessarily doing that sometimes, but they could, if they they wanted to so um or you could share the same conflict between apollo and gatsby.

B

Let's say or something like that, uh so that's one thing of course: there's graphql code generator um that I think um some people heard of uh I mean like it's probably one of our most popular library, but the idea there is that um you could generate uh things. um So we we said, maybe we could create it's as a system as a plug-in like swagger.

B

Cogen was our inspiration, so you could generate front-end code like you can generate uh ready-made uh react hooks um like if it's pro hill- and you see here- you can by the way you have a whole uh playground. So you can actually like just you have the schema and then you have you can have some operations and you have config. So what you see here is like a very basic example.

B

You have the schema and you basically get all the typescript types are equivalent for the schema now um and that's all you need to do now. If you scroll a bit up um on the web on the on the on the website, yeah um yeah a bit down, so you can choose a live example. Yeah.

A

B

A bit down yeah, I think it's in delay for me, but then um you can see you have like for the front and you can actually generate uh react hooks for uh or react query hooks yeah. So you know, or you could generate for the back end, uh the resolver signature or you could use the same and you can reuse the plugin. So you can do the same for like so everything you see here is typescript and frontend, but you know you know you can keep scrolling down.

B

You have you know things for the backend and for node um java flow. You know whatever you want, there's like a whole net like there's a whole ecosystem of generators. So um the idea here is it started very simple, but we just kept listening to the community. I think our core um things that every day we are thinking about is are we in a position where all the feedback that we could get? We could do something with yeah. um Otherwise it's it becomes unhealthy, open source project.

A

B

Extremely valuable.

A

Because, like even just the one example about prisma and how you absorb the projects, but I could see other graphical companies projects uh folks, who were sort of out there in the open and have their name attached to things, could leverage the guild to sort of consult. I guess I assume you're all doing some consulting as well uh to help push other projects and ecosystem forward because, like.

B

I guess what I'm getting at it's like. I see.

A

React queer, I see apollo, I see like a bunch of other libraries all within this one. Yeah code generator.

B

Yeah exactly because you know we just we had projects where we needed to work with with apollo. We had projects where we needed to work with oracle. uh We had projects we needed to work with relay, um you know so, um and that's the idea here is that instead of um I don't know coming up with like this visionary product roadmap, which is cool um instead, we just solved our day-to-day issues, and I feel like this is something that I I learned from. Facebook like facebook is open sourcing tools that they use in production.

B

uh If they have really fancy ideas but they're not using it in production, it doesn't solve a problem in the day day-to-day they won't open-source it, and I think this is a very valuable. um It's a very powerful concept, not saying like we need those visionary projects, but um I think they need to be marked as visionary.

B

So when people start playing with them, maybe it's not actually answering their needs. um Maybe I I can also talk about two. I guess um big projects. I there's a list of a lot of them. I think it's worth for people to read and there's also our blog, but I think there's two projects that are very interesting, that maybe I can like just take a couple of minutes to talk about yeah um one is called graphql mesh, um I think yeah, you can click here and then.

A

B

Yeah exactly um the idea here: yeah sorry for the all the crazy stuff we're working on improving the.

A

Improv, improving the design.

B

This is yeah, it's a bit incense, so um so so the idea here um yeah, maybe we will move to the docks or something because oh yeah scroll down, so people won't get that like like an attack or something so um the idea here was.

B

You know, I saw a lot of you know when you, let's say you're a graphql startup and you want to push graphql for everyone. So now you're you have this obligation to say that graphql is better than anything right like.

A

B

Is the the best thing ever all the previous, um like all the previous protocols are, are crap. You know the holographic.

A

B

The rest is that yeah so, but when you don't have to do that, then you can just you know freely, think and then you could say well, what's actually the difference between, let's say rest with some kind of schema and graphql, and once you start thinking about it, you know we started thinking.

B

Well, maybe we could actually take um let's say and also again we had a problem I'll get to the problem in a second, but once we started thinking freely we said well, maybe we could take all the existing sources out there, because when you come to a real company, you can't just delete all the code. You need to work with. You know open api and swagger json schema, maybe soap, apis kafka streams, all kinds of things, so we thought. Well.

B

Maybe we could take all these different sources and we could automatically convert them into graphql schemas. So as the consumer, you could query them as if they were graphql also like. If we would want let's say we would want to make a product out of it. We would make it like this central centralized thing like you. Take all the sources you merge them together into one big graphical schema, and now all your organization and your code needs to run through that and now you can charge tons of money for it. Yeah.

A

B

Is okay, but if you think about the right architecture, you think maybe that's not necessarily how you want it to be like what, if we'll take all these sources, let's say generate or compile them into graphql and now, let's generate um an sdk, and that sdk could sit on the client it conceived on a on another service.

B

So now each you get as as the consumer um or is the front end. You get to query one graph, you get to send just graphql queries and you feel like everything is in one place and it's like it's centralized, but that sdk will actually call the individual distributed services in their own original protocols.

A

B

So it's grpc for one and like rest for another, um but but you get the cool stuff from graphql and you get maybe the benefits of some performance from grpc or you get to reuse all your existing services. um So that's graphql! Imagine it could also run as a as a as a gateway, but it could also run as an sdk now.

B

Another thing here is that um once you actually uh get all these sources and you want to merge them into one graph- um we didn't want at the beginning to take any opinion about that, because there was apollo federation, which is a way to like merge together all kinds of different sources, but we saw that it's not necessarily answers for all the different use cases, so we actually took over graphql tools and schema stitching from apollo, which was deprecated.

B

We revived it. And now, if, let's say, if you use graphql mesh, you can actually choose between apple, you could use a poly federation with graphql mesh or you could use schema sticking with graphical mesh. um Is federation.

A

B

A

Product is that something I have to pay to use, because I can.

B

Take it no okay, so so it's it's a kind of like a spec um um on top of graphql, it's kind of like an extension on top of graphql. That apollo uh is is, uh but you have actually a lot of different implementation for that specs implementations for that spec, for example, mere form um has their own implementation of federation um and there's also netflix, says a library called dsg. I think uh for java uh that understands the federation spec, um but yeah, for example, that there's net hot chocolate.

B

It's like the main framework in the network and their approach is much more similar to graphql tools, schema stitching like with mesh. You can choose basically all those different strategies, but maybe I'll just share one more library that I think is very very interesting. I don't.

A

B

There's a bunch of stuff, like uh you see, there's a federation example grpc soap, um all kinds of let's say sources that existing sources, like um let's say uh stripe like stripe, doesn't have a graphql api. But now you could just you know with two lines of code. You can get a graphql api from stripe. Okay,.

A

B

A

B

Yeah exactly and they're open api schemas, um so it's really cool. uh We worked actually with microsoft on we're still working with them on all data. uh So old data is very interesting protocol. There's a lot of similarities between all data and graphql.

B

So we took actually the microsoft graph, which is probably the largest um data api out there, and we they had projects to try to rebuild it with graphql manually didn't work. It was two three years ago and we just converted it completely into graphql uh pretty quickly. Arda did that, but then we reached out to them and today we're collaborating with them a lot like it's. We actually there's now like a lot of debate within us and them about.

B

You know what cool things there are on all data that we can carry to graphql and the other way around.

B

um So this whole graphql mesh project became this center of talking around differences and similarities between different protocols, yeah um yeah, so just people can go to that and um yeah uh there's a there's a I have a. I did a workshop somewhere online on like stepped on step, one step like step by step. How do you can integrate graphql mesh into like an existing project, um but yeah? We can talk about it later.

B

There's one thing I do want to present today that I think is cool uh last thing that I'm just gonna talk about like that.

A

B

uh And that's uh it's called envelope. This is a new thing, it's open source, but we didn't launch it yet. Okay, so um envelope today, there's like um the most famous uh it's basically. What we're gonna see here is is a new. You can call it or a new server framework or basically, a graphql plugin system.

B

Okay, so the reason we created that was that we kind of like became stuck between graphql gs, the reference implementation, not moving fast enough for good reasons and the most open source, the most popular server, libraries that are basically or unmen, are not maintained or are not being up-to-date um and are not being flexible enough.

B

So what happened is that, let's say there was huge fragmentation in the graphql server uh ecosystem, because, let's say you created your schema through uh schema first uh using graphql tools, then you had to build your own full server like with apollo server.

B

um You couldn't reuse things and share code between people who were, let's say using graphql nexus from prisma yeah, and you couldn't share things from people who use typed, graphql um and also when you you actually mentioned at the beginning of your talk. They said things are hard all kinds of toolings like caching tracing um um uh all these things became very hard um rate. Limiting there were whole. There was a lot of people. Gay gave really interesting talks about it. You know in conferences and stuff like that, but it shouldn't be that hard.

A

B

What we've done was we actually thought? Could we hook create a plugin system where we could hook into graphql gs itself, and then you can just install a couple of plugins um and you can share those plugins between whatever way you want to build your server. You want to use nexus. You want to use graphql tools, it's all fine. You want to use festify, you want to use express it's also fine, it doesn't need to matter about how you do graphql.

B

So this is we're going to do a very big launch for this thing soon, um but we already open sources and a lot of people are using it. Okay, what what? What what you can think about is kind of like a framework for building graphql frameworks. So um what we've? What we're now in the process is actually we created prs for basically every graphical framework out there, so we created a pr for apollo server. We created a pr for parse.

B

We created a pr for nest, gis we created a pr for loopback and we created a pr for redwood. um So um redwood is gonna.

B

We basically finished the pr, so it's gonna be verge very soon, um and now all of these frameworks, you can choose whatever frequent framework you want, but then the way you're doing tracing the way you're doing um like if you scroll a bit down, there's a list of uh all the different, so the way uh you're doing vlogging uh all these things, you could use sentry data dog, open, telemetry like off the way you do off. All these things could be shared between all these different frameworks.

A

B

um So this is like a really big deal for us, um and just this is another cool thing that happened. We spoke with prisma because he said we talked earlier about the fact that we're taking uh libraries from them and people from the graphical world probably know about graphical yoga. It used to be like a very powerful or a very famous server, and it's got a bit left behind. um Prisma just transferred that to us. So what we're gonna do is we're gonna, basically rebuild um graphql yoga using uh envelope and we're gonna.

B

Do it in the open. So we open the new issue on graphic yoga and people could just ask whatever they want and we will create it uh on this new infrastructure.

A

B

A

B

A comment that just got dropped too.

A

As well and also ajc webdev, who hangs out the redwood team a lot, so he dropped a link earlier in the chat so folks who want to scroll up and grab that link for that. I guess that pr whatever's happening there, um but there was also just a comment from n1 ru 4l uh about envelope.

A

Allowing future graph kill features too as well is that oh.

B

A

B

Yeah, that's a good point, so I I everything I mentioned now is basically us is um how we solve the problem of the the uh let's say, the the the frameworks like that we couldn't hook in we couldn't do some things. We couldn't use the new stuff here.

B

There's two cool things here: one is that we could use the latest graphql. Also the experimental features I I'll mention some of them. One is defer and stream. So when you send a graphical query until today, you will get back one single response and that's cool yeah. But what happens if some of those fields are calling, let's say a different service and that service is very slow?

B

What if you want to get partial results already and then later get the rest of the results? So that's a really powerful feature of graphql and other features you can build on top of this capability is live queries you could send a query and that query will just every time there's a new result. You will just get pushed in your result.

B

So all those features you could use today uh with envelope um and the libraries that actually uh are all building it. Some of them are being built by guild members and some of them are being built by um people we support, um so we kind of like um uh and, like I said, we kind of like unblocked the the ecosystem for using all these new things and there's another cool thing that this maybe uh I'll mention which is, um as I said, there's all kinds of our in the graphql world.

B

There's a lot of uh new features and it comes in the form of an rfc you're starting an rfc there's a very long debate um for good reasons, um and then it needs to go to the reference implementation. That also uh can be very slow here with those plugin systems. We basically hooked into each function that graphql js exposes.

B

That means that any rfc we could implement already now as a plugin.

B

So what and we are already using it today- basically, you can look at envelope- is kind of like babel for graphql, okay, so with babel you, you could use new graphql new javascript features before they were. You know implemented in browsers or before they were standardized.

B

We are now using envelope for the same uh with the same capability like there's, there's a lot of cool stuff like one-off directives and all kinds of rfcs, and we just this week, we actually created uh lauren from the guild, created the um basically a working plugin for envelope that, basically, you could use it today. Already. Okay,.

A

Very cool yeah that is uh like fascinating stuff stuff that I'm actually going to uh I'll give that a star, because I want to watch uh what's happening uh with this project. Actually, I've been, we have to actually.

B

Maybe at the custom.

A

Yeah yeah yeah, so I want to I want to see what's going on in this project, so I'm going to watch the uh releases and discussions.

B

Yeah, I'm also a like a guitar notification fanatic I, like uh I, have to have inbox zero at the end of every day, the real, the real inbox zero that I care about is they get them notifications? Everything else is boring.

A

Yeah, I I do that with mentions. If I have, I I make sure I get through.

B

A

Mentions uh during the week, uh but.

B

I can't I can't.

A

Capture every issue, mainly because I'm on at github, we use github to build github so like I've.

B

A

I've got stuff or notifications because my team was mentioned on a completely different other repo, and so I can't ever get down to the inbox here I could. I guess, if I was super super focused.

B

A

It but yeah just I've given up on that yeah.

B

Yeah, I'm I'm a fanatic for it, but but it's like uh it's hard, it's very very hard, but but it's the most interesting f like it's. For me, it's more interesting than I don't know. Twitter, feed or facebook like this is the fit for me.

A

I mean we're already monetizing a bunch of other stuff, but um but.

B

Hearing that I.

A

I do know we have some features coming down the road to make a little more intriguing to uh you know, check.

B

A

Feed on a regular basis, um so I'm looking forward to that in the future which I'll be in touch on that when we uh start beta yeah um awesome, we do have ten minutes left. um I don't know if you wanted to show off anything else. um We.

B

Talked about a lot of these projects, yeah, so maybe some projects, I think people could like look in and they could check out now. They know where their website is, and hopefully now also graphql.org is better, and mostly um you know now. Graphql.Org is uh contributed ball so like. If you anything you see, you can contribute so and if there's an issue, if you're scared of contributing, if you're stuck uh just let us know like we are on the website, there's like. If you go to our website, um you can see like we have.

B

uh You can just reach out like um you. Could you have if you scroll down, um you have yeah. First of all, you have this thing that looks like uh it looks like a an annoying bot that we're gonna sell you something yeah.

A

B

You're gonna write there. Then one of the guild members is going to write back and we get a lot of.

A

Yeah, let's just go to a um like: how are you managing.

B

A

B

A

B

Crisp, uh unfortunately, no, even though we had an inter we, I talked with the ceo and tried to uh convince him to open source it, uh or if someone has like an alternative that is open, source would love to switch, but it's just what that gave us is we got. We have it on all across our websites.

B

We got to um get feedback or get people that got stuck where they got stuck in yeah. So at the places there got stuck- and so this is like was very good complementary for us for github issues, uh our discord channel, um the graphql slack. uh This was like another complimentary thing, so you can just write to us.

A

B

One of those channels and, of course, github discussions is like really really cool, so we're we're entering everything.

A

Yeah, so actually the very first one of these streams I did. I actually did on my personal twitch, and it was with this.

B

A

um Which, essentially it's the same bot, um but it's open source and it's built and react? I don't know if this.

B

A

It gives you the same features and bells and whistles, um but it was like some sort of like project their their team did is like a little fun hack project. That's become something bigger, so.

B

A

Open source intercom or help scout.

B

Yeah yeah yeah, all this other stuff. This is awesome and because it's open source, maybe we could like merge that with all the other feeds you know that we get everything into one place and then it's just like you know it will be easy for us to answer, but it will be. Anyone could reach out from anywhere yeah.

A

B

Could be really like whether it's github discussions or discord or whatever.

A

Yeah and why this is top of mind is because I have an issue in this project where I wanted. I wanted to also have a chat bubble, and I wanted this to also.

B

A

In our discord, so that way, anybody in the discord could answer a question about somebody getting stuck uh in the.

B

Project because this.

A

Well, I don't know if you know about this project, but this project's, I'm pretty sure. I talked about this project last week as well, but it's a project to find your next open source contribution. So if you just want to get.

B

A

Source having a little chat bubble of like, where do I start or how can I, how can I figure out how to do this or whatever um that's what I was trying to read this.

B

A

Awesome I did a punching bubble, though.

B

I I need to look at it because um the last thing I wanted to mention um is actually our um let's say: contribution um support system.

A

Okay, yeah, which it might be interesting yeah. I wanted to ask about that too, because I think vortex actually dropped in the chat about. What's the pipeline, look like for contributors, the these projects.

B

A

Community as well, so you mentioned the graphql foundation but yeah. If you want to talk about that sort of pipeline and how that works.

B

Yeah, so for me, I always think about these things as, like I mentioned the beginning, a cycle or kind of like a journey um so, like I said, like everything um that we get, we try to think what's the origin of this thing so and what's the path so for that's, why we got into graphql.org because we realized like yes, every issue that people have on our libraries. They started from graphql.org there's some, maybe just a lack of knowledge and that knowledge could, you know, be beneficial for beginners.

B

So that's why we cared about that now same things. Go for this contribution pipeline, so um we actually vignesh from uh uh from the guild. Just wrote a blog post about it like if you click on the blog um there's a link there.

B

Now, if you scroll a bit down the making open, so this is basically we open source how we do open source, okay, so um yeah, there's a just a link there uh contributor uh the guilds contributor guard yeah, so this thing lives under the guild, org, actually yeah, and this dock uh I'll just run through it very quickly. If you go all the way up.

B

um First of all, uh there's a bit about our philosophy and everything we talked about, but then uh vignesha click actually created video tutorials on how to contribute nice, which I think is extremely valuable. Like sometimes you just want to look at. You know there should be more tutorials on contributions are not the only usage.

B

So that's a starting point for that, but then we also imagined a path. So if you scroll a bit down, um there's the typical workflow, so you know it starts with like you have an issue so we're basically what we're describing here and we want to visualize it um somehow. So if someone wants to help us visualize it as a map or something, I think it would be really cool, because a lot of text is always hard.

B

But basically um you have. um You know you start with an issue and then maybe it's just a lack of knowledge or lack of docs. So we just want to always highlight where you are in the process and how um you could get to the next step and then the next person that will come after you will also push that for the next step. So you know it's opening an issue creating a reproduction um creating a failing test.

B

um You know solving it locally, um then submitting a pr like all these things, and it's all with um you know just what what we saw is that it's not necessarily that people are just it's not that it's hard, it's just that they don't know. What's the next step or what to do um so.

B

We first of all build the path which we described here and then we created video tutorials to help you go through each of these paths, and we published that I think a month ago and it's crazy how many more contributions we got. Yeah like it feels like all the issues just started, um advancing themselves, yeah like it. It's been amazing and this thing shouldn't be, in my opinion, under the guild's org. It should be.

B

I know under github or, like you know, we we want others to like.

B

Actually, look at that and you know improve it because there's nothing guilt, specific um I'll, just mention a bit if you scroll a bit more down, that's like the most important part, the contribution, but then um there's stuff here about how to contribute, even if you're not contributing to code because that's another, you know, I think it's very similar to probably the project that you've seen and you've seen it way more than us, like yeah people want to get into open source, but it's just you know they don't know where to start, and then the last part which I think you know other library maintainers could use, is basically how we maintain and what toolings we use to uh to create our own.

B

um uh The shared tools we're using across all our libraries, because we have a lot of libraries, the future. Oh, we would love to contribute for whatever you're working on.

A

Yeah, so this this needs a bit of a refresh. We haven't touched this in about a couple years like two or three years, um but yeah I'd love to get like some of that context. You have in your your one, your standoff contributor, workflow guide, because a lot of this is more about finding good, first issues, finding projects finding this sort of this finding community in general. Exactly but there's not a lot of what's the next step, and what your file looks like is that next step of like okay yeah, the path yeah.

B

Yeah and- and I think um yeah this path is very similar like I actually started this path, uh trying to contribute that to the angular projects, many angular projects many years ago, yeah.

A

B

They didn't take it, but for us we finally got, and I then I tried to contribute it to the meteor project and then to apollo and now I can finally decide on things. So I can I we just created it, and um it's been really successful more than I expected. We, we actually not only that we got more contributions.

B

Actually people stepped up, and now we have in a bunch of libraries, for example, for example, graphql eslint, someone from the community is now an official maintainer because they just followed those steps, and then they just took over the project. Basically, which is the best thing that could happen.

A

Yeah, um no, that's that's great, that's that's! The goal is like you want to get. You want to pick the interest of that way, you're able to level up folks that can then level up more folks. um So.

B

A

Their experience are now engaging in like even just your the doc that we're just scrolling through, like the the individual that contributed this to your project. um Yes, is now leveling up other folks, because they were able to sort of lay out the sort of the guideline and like whether the path is is static or maybe like step. Zero turns to zero, a or zero b or whatever it is like.

A

This is all learning that I think you're developing, because, like even uh we didn't cover it, but like your one of your projects had a discussions. It's like this is part of the contributor path that didn't exist. You know last year, uh perhaps.

B

A

Stack overflow or somewhere else or a discourse, um but this is part of the the the contributor.

B

A

Where I'll, what I try to do is try to get people involved in projects in a capacity of like if I've got 72 issues in my project can I can I encourage someone to contribute in a way of like? Can you just label these issues, or could you just like you know, add request more context because, like there's so much, you can do with the bot, but when you have a human, that's exactly.

B

A

Gets you to inbox zero? I guess per se uh going back to your earlier comment: um yeah. They were able to level up and sort of exactly what you're doing it's funny, because we now like we're broadening our conversation where, before you were just leveling up graphql um yeah.

B

A

That open source, but it seems like this project, your projects have the ability to level up more of open source outside of graphql too, as well.

B

It's amazing to hear from you saying it I mean it's uh and, to be honest, graphql was also it's also just a tool for us yeah. So it's not like, like I said it's not like we're, um you know it's just it's. It's been a good tool for us, so it's been valuable to create things, but um you know like bob. The bundler has nothing to do with uh with graphql. For example, yeah.

A

B

I think I can- and I agree with you also on the bot stuff, like there's a lot of projects, with a lot of contribute with a lot of traffic that try to use bots, which is, I think, it's cool, but I think, having someone actually reading the issues and you know looking at the path and directing people.

B

This is the most valuable thing and there's I think enough. People who want and can do it just like you said, and they just need a clear path um and they could level up each other. Like you said, this is really inspiring. I think.

A

B

Yeah, I'm happy you're. You.

A

B

A

Look forward to keeping in touch I'll have some more questions. I definitely want to invite you in some other conversations in the future, uh so you always welcome back here, but also, if you want to help beta some new features, we have a github uh I'll, be I'll, be in touch for sure.