GitHub Universe 2017 - GitHub For Business, 30 Nov 2017

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Using GitHub's GraphQL API to manage open source projects - GitHub Universe 2017

Description

Presented by Brian Douglas, Netlify.

GraphQL is a legitimate way to build and consume APIs with less front end frustration. It has only been out of developer preview but truly shaking up the way front-end and mobile applications are being developed. Getting into open source projects can be a daunting process, and there are a lot of existing tools out there to assist in getting your first open source pull request including Code Triage, IssueHub, and of course firstpr.me. Open Sauced is one more tool to help by providing a platform to organize notes on potential projects the user has not yet contributed to by leveraging the now public GitHub GraphQL API. This talk will compare the GraphQL benefits over REST and walk through the implementation of live code implementing the open source GraphQL features using the Apollo project.

About GitHub Universe:
GitHub Universe is a two-day conference dedicated to the creativity and curiosity of the largest software community in the world. Sessions cover topics from team culture to open source software across industries and technologies.

For more information on GitHub Universe, check the website:
https://githubuniverse.com

A

Hey everybody guess: I might I'm gonna get started because I am the last talk so I'm like between you and your Laphroaig going home so and I also want to go home too, as well. So I've got to get started. I do have a lot of stuff to go through so I can I want to get all in without rushing too much. I am Brian Douglas.

A

If you guys could see on the program I'm going to talk about using github graphical API, mainly to manage open source projects, but think of this more, as it's got a use case for using the API. So a lot of us already use github API at a rest-pause of it. This is basically the showcasing the graph QL API.

A

All my slides are actually available on Gucci Detlef adhikam. That is how you pronounce it. So if you guys want, if I go too fast or you want to go home and maybe pour a glass of wine and go to these slides later on, feel free, they're right there. I also have that the m2 as well so again talk about the github, v4, API and there's. There was also a talk yesterday in this very room, Brooks I'm, pretty sure, he's happy to have his face on the screen.

A

He went through more details of how they got to this point to the v4 I'm gonna talk about how to use it and the cool thing about the launching the v4 is that they actually launched it almost a year ago, as beta when they first wouldn't get hub.

A

When I say they I'm talking about github github announced that github universe using graph QL as part of their API spec, which is really cool, because I got me excited about using graph Keola, something I heard about for a while, but never had a use case or a reason to stop production or go and try to figure out whatever this thing is, and github was a good playground for me to try it. Try it out so I hope we just talk to encourage you to learn graph QL and, if you've got through this conference.

A

This far and have ever heard. The word graph kill, congratulations, I'm, pretty sure, like almost every single keynote over there they've always been missing graph QL and how they're very serious about it. But if you don't regress Killa, it's a query based language to grab data from your ear api's and that alone wasn't enough to get me excited, but it was very exciting to see something different between between what I've, already known, which was arrest.

A

So let me take a brief moment to explain: Who I am and where I came from I actually write code at nullify. I do a lot of speaking and I do some tutorials and screencasts around metal. If I were a continuous deployment platform for the modern developer and we we handled a workflow on modern web projects and that web projects ash trick, what I really mean this Jam stack, so Jam stack is a term that I were like using because it means JavaScript, API is and markup so JavaScript really powers the web.

A

At this point I you can't really get away from it. Maybe in a few years will have like this will weird closure mess that we can use like in some Firefox nightly about browser, but until then JavaScript powering the web, so most front-end web sites most web sites have javascript in them, and we, like that term Jam stack to be able to explain what we do identify to further explain that this is the monolith.

A

So if you guys ever seen this this jar before on the shelf as you when you're a kid, you got to peanut butter and jelly in the same jar. It's amazing for, like the first scoop and then after dad gets like really messy. That's a monolith. You get you kickstart their whole product, the process of doing like peanut, butter and jelly. You get the MVP really quickly at 30 seconds and you're back to playing like Nintendo or whatever, but the problem is like.

A

If you want just jelly or suggests peanut butter, you kind of have to go through this huge tech debt and like reorg, in hiring and firing people and getting like your entire infrastructure difference. So jam stack really is just this and you just separate your front and your back end, and so you have a API and then you have your front end in your client. Your client could be a mobile phone. It could be all those other things.

A

The last talk ad, which is like all these 4k video screens and Alexa, and that's our clients, so we're talking about that, but mainly I'm, talking about the web. So before I can talk about graphic. Oh I, gotta talk about rest first as well. Is everybody familiar with rest at this point? You guys use it. You guys use the github API at this point, so these are all endpoints for github. You like slash, use your repositories and notifications.

A

Ideally, if I go from the github API API and go to my app I expect to get a user back right. That kind of makes sense. You got tracking with me all right. So if I go to slash repositories, what do you think I get from repositories?

A

Repositories there good you guys are following along alright.

A

So if I go to / notifications, what do you think you're gonna get you're not gonna get notifications, because it's not notifications, actually / user / notifications, but how am I supposed to know that unless I go fished or all the documentation or trial and error and figure out like what am I looking at that's a big problem that I've had I've, been on like a couple: different engineering teams and when I first come on board to a team, I have to figure out what the back end looks like and every engineer, I've always been a part of small company.

A

So every engineering team has no documentation around their API, so either I'm. Writing it or I'm, just looking through tests. If we have tests and then or I'm just trial and error in figuring how that works, the good thing is: github has great documentation. So it's not that painful. The bad thing is that if I want to do a project and I wanted to have like user notifications repositories and the one project I have to fetch it from three different points and then some so it's quite possible.

A

It's quite possible to get really lost in the docs, which is why I want to talk about graph QL and why is different from rest, so graph growth takes what you would normally do with like multiple arrests, endpoints, and you can make that into one single request tall, so I get the fetch data I'm really looking forward and going back to that same model that I had before.

A

This is the difference so, rather than straight, hitting up the API I'm, actually talking to some sort of graph fuel schema or a graph go where and the beauty of that is that I can wrap it all in this one endpoint, which is /, graphic UL, and that's exactly what you can do today. I also want to point out real quickly that graph coup is not a replacement for us.

A

I think we've kind of learned that if you went to Kyle's talk earlier marketplace, is going forward with using graph qo as new features, but you still going to need to use the REST API until they get feature parity throughout the entire graph QL spec rate, the v4 API should call it.

A

So it's not a replacement, it's actually more of an enhancer, so you're able to consume the data, and it's I personally think it's a lot nicer I, also like it, because it fits in my jam, so graph kill interactions, I'm just going to lay out some different interactions that I might talk about them. Briefly as well LaRocco you got the two things which are queries and mutations. If you think queries as gets and mutations as post, you can kind of keep it that way. That's higher level I, don't under the hood.

A

It's not really like that, but forgets in post.

A

You can do queries and mutations and again I mentioned it briefly earlier, but graph you'll, where it really shines, is being able to define the shape of the data you want and that's the data you get into your project so going back to first queries on the Left I have a query which I'm this identifying I want to get the notify owner and then I want to find the go true, repo and based on that I want to pull back just names and just the issues for that repo and on the right you see my JSON data blob, which I'm getting exactly that data I want.

A

So this is really powerful for front-end, specifically front of developers. I think that's really what the use case was really hammer on at Facebook as a front of developer. You could actually just request the data you want and if it's there I'll come back to you and if it's not, then you can go back to the API team and let them know to expose that in the schema mutations or again like post requests.

A

So this is my example: mutation, that's working in the github API, where you can add a reaction to issues so I just want to go to an issue and add a heart. I could do that with that, just by finding the subject ID, which is the the idea of that actual issue.

A

So now that's getting let's get into how you actually get started with the graph QL API and as I mentioned before, it's just as simple as pointing just to the / graphic UL. So you could do all your fetch request, just as you wanted before from that same API, there's no slash before you have to actually put on there and there's a really cool tool that was mentioned, I think earlier, Brooks's talk which is graph you'll explore. They also announced this last year and graphics, for is this another.

A

It's a GUI, an interface in the web that syncs up to the github API. So you could actually query all the data for github directly in a browser, and it looks like this. So this is the GUI. The cool thing about this is that you don't have to even find out what your token is. You don't have to off anything. Well, you do have off, but you can do it from the upper right. It's clicking the button you can log in and start doing, grackle right away today actually feel free to open up your laptop's.

A

So on the Left pane you right out, you can actually spec out the query the cool thing about this. It autocompletes. So, if you're not sure how to spell repository, it'll spell it for you, similar to like PS code or xcode, or something like that and in the next pane. That's where your data from so very similar to my query, the shape of the data and then returning the JSON.

A

That's exactly what you're getting here and graphical Explorer, so I've, I used to use graphical graphical is what this is kind of based around graphical is more of like an electron app to do the very same thing you just have to pass in your own tokens, ID. Actually that didn't start using Explorer until last week when I was preparation.

A

Preparing for this talk and I kind of kicking myself for not using this for this long, because everything that I need is already there and at this point, I've mainly only been using graph tool with the github API and I have a little gift that would kind of shows the auto completion so I'm, just like adding a name here and then adding a URL. So my repo URL and then that data actually changed changes.

A

So if I decide I want issues, I could delete those issues and that data will dis change exactly the shape that I'm looking for so that's kind of graph kill in a nutshell, so I heard about the graph cool API last year and I really wanted to use it really bad, but I couldn't actually sell my team on just stopping production and adding graphical to our API.

A

So I thought to myself: I'll just build a project to be able to test out this API, this github API and that's what I did and I did it with this thought, where github is the the place for open source, it's a place to discover open source and I. Personally, wanted to get involved in an open source, pretty badly I'm already involved, but I wanted to get involved in more of a higher degree, starting to contribute to larger projects. The problem is every time I go to start like to contribute to a new project.

A

I, don't know exactly where to start I, don't know where issues I know.

A

There are good, clever uses of labeling and talking to open-source maintainer Xand contributors, but if you're not already inside the the space, you don't know who to go to also, if you don't know how to use the actual projects, there's a lot of times, I'll go to hacker news or I'll, just find them like the github Explorer tab and I'll find the project that seems interesting and then, by the time the weekend comes around I'll forget exactly what that project was so I can't really I, don't have a really good way of tracking what projects I can contribute that I I'm ready to contribute to and I most likely will forget them eventually and I've tried other tools like using Trello boards and ever notes and I ended up just filling it up of all this data.

A

I'd never actually do anything with it. So and I felt a bit like this. You guys aren't familiar Gucci Mane, he's a he's. The Godfather of trap, music he's also a prophet and he's a health food fanatic. He did an interview about ten years ago talking about what it's like to get lost in the sauce, and this is a direct quote.

A

If you don't got sauce, then you lost and knowing that and knowing how much I love rap music I thought what if I made an open-source project to manage open source projects on top of the graph QL API and called it? Oh, it's froze ooh.

A

I just got shocked and I called it open, sauced, sorry, I, don't know. Okay now, I'm I must been rate limited with my Wi-Fi cool. That was a big flop, so I called it open. Sauced and I'll just use my keyboard from now. So it's a way for me to take notes and save github repositories for me to go ahead and contribute to these projects later and the way I built it as I built it in this way.

A

So this is like the basically the infrastructure I used to react on the front end and I used to github API as a back-end, and one thing that Kyle talked about his talk is that they're really gung-ho and using github as a platform and that's exactly what I did here: I'm, actually using github as most of my back-end I'm using graph cool, which is a it started off as a back-end as a service, a graph cool as a service, but I think it's gotten way better than just that and I'm.

A

Using that to be able to save my node data as well as well as repository IDs, so I, essentially I just took github and grack UL and I made open sauce, so I'm gonna, the next few floors, I'm gonna, show you exactly how I use to github API to get some of these features into my application. So I'm gonna start first with this fetching data. So that's exactly what I'm doing I'm using github as a platform and I'm just pulling that data into my react.

A

App, and it looks like this so I have a form where I'm this taking my github URL I'm, hitting the graph. The graph killed github ape yeah in point and I'm being unable to fetch this. The data I need for those forms so here and this this form I'm, actually saving a little more data than actually need, because I didn't really know where I was going with this project.

A

It'd be to be quite honest, so I have a form that I was collecting a lot of data and I'm saving that the graph cool so mainly like this IDs and names. I also want to point out real quick Grieco I actually do a lot of talking about graph, QL and graph. Yo is not a Java scripting, it's not a reacting. It's actually graph QL works of all modern languages and modern, tooling so frameworks. There's a lot of plugins and tools that people have built for this.

A

But all these slides are JavaScript. So just to let you know so, I don't want you to think. Oh man I need to learn JavaScript, because Java scripts of future I happen to use JavaScript but I, know I, know. Github is actually using a graph kill in there ruby, app I'm, so the most basic way to get graph QL into your app. If you're using javascript is using something like a node fetch or you can use Axios I'm a big fan of Axios.

A

You just get that JSON back right away and, as you can see, I'm just using the straight-up API that github.com slash graph QL and then from there I'm.

A

Just passing in my repo query and then you're probably curious of why I'm passing in a repo query or where that's coming from my rig per query, eiseley separate into a different function, because it's a lot of I know: there's a joke in the New York Times slide about template strings and not being able to a syntax, highlight I sort of figured it out, but so I'm just I built my entire query. The same quarter that I showed you earlier. It's the same query that you're looking at here.

A

It just has more data and I'm passing that to a function that then returns my JSON data, so that JSON data I can throw in my react, app I have code and content on the screen. So that's pretty much it for fetching data. That's the most like! That's the easiest way to do it when you fetch it and passing the queries. I'll talk about another way in a moment, but I just want to take a moment. Just talk about documentation, documentation for a graph curl is actually really nice.

A

The cool thing about this is that inside the graphical Explorer or graphical, when you build your schema and the backend for about graph QL, all the note, sorry, all the documentation you put on top of your schema. So as you build your schema, your writing exam. What these endpoints do. So, if I want to know what an issue does I can expect to find the documentation exactly what that is. If I can't spell repository correctly or repositories, then I can figure out and it will autocomplete for me and I can find it in the documentation.

A

I can also find exactly what is inside a repository. I can find the that users have notifications and notifications. I'll have users, they technically do, but I could figure out the problem that I had earlier and a couple slides ago, and the cool thing about that is all the documentation is here so I'm at the point, where I'm no longer fishing through the kid out documentation to figure out what exactly I need to do. I could actually just go into the graphical Explorer start typing and if I can think it that it can probably happened.

A

So, let's talk about pagination edges, so repositories issues, comments or all plurals. It means that it's an array of data and the idea of graph QL. There are edges so an edges. They all have nodes. So inside of those arrays, you have elements, and that was something I actually didn't, have experience until I started getting to the problem where I wanted issues inside of my project, so with issues I have this. This cool array over here that has two different colors I have an issue inside that array.

A

That's the object that I'm expecting whenever I return issues, knowing that I was able to discover that through the documentation here, which is exactly the same thing at a time, I was just talking about before, but again it's reiterating that all the discovery I did for issues. I never went to the old API and the documentation. I went straight to graphical or the graphical Explorer and was able to find that and I was able to find out koutou as well when I was trying to return like every single issue.

A

Nunda man I wasn't able to do that because inside the graph view of that did go to the documentation to figure this out, because I was a little baffled about this, but they do limit the amount of nodes that you can return as a hundred, but that's something that I was able to find out really quickly and because of that, I was able to figure out how to actually get issues and do pagination as well.

A

So here when I'm fetching my shoes on my project, I'm in fetching the first five issues and I'm able to paginate throughout through those those five issues and I'm doing that within that query right here. So again, this is the same query. I showed before it's just now we're just looking at this issues, point so I'm to pass in different elements. They actually filter that data and again I'm passing the first five, because I only want these issues show up. So this is live UI.

A

My UI is, it's still a work in progress again, this projects open source. So if you are really good at UI, come talk to me, I will give you free, github hearts and on any repo you want. So so these are. This is exactly what issues look like I just wanted to paginate it I want to be able to have some sort of discovery tool to find issues that I think I can contribute to. So you could also handle ordering and filtering so I'm ordering descending, but I'm also ordering by only open issues.

A

So if you notice in the this object that I'm actually have States pointed out that I'm actually doing open for whatever reason, if you wanted to find closed issues, you can find closed issues as well. So that's a good thing to think about too as well. If you want to find and pull requests or issues or comments that actually open and available to you and I did mention that you could actually order data too as well. So and again, this is pageant issues but I. Also I was one thing.

A

I didn't know how to do, and it's funny because Brooke she actually talked about this early in your talk, but I had to discover this the hard way. How can I paginate issues if I'm gonna get the first five? How do I get the next five and the cool thing about the graph QL API is that it gives you a cursor to be able to paginate through those as well. So that's Pageant issues you can do aliasing with edges. So when I get my react, app and I get my data in JSON.

A

If I have like data that edges or sorry, if I have like repositories got edges, it's doesn't really. My apologies issues. That edges is what I'm actually talking about. It's, not it's, not very I. Don't really know what that is. If I'm not really up to speed with graph view. Also I'm aliasing is data, so I can know issues that data or this stop props. That issues that data, as opposed to whatever edges, are I, don't have to explain that with my team.

A

So a listing is cool because, as a front-end developer, I can name anything I want in my front and app of what I think it should be names. So if my head of infrastructure, API made a mistake in naming, I can prove it to him by renaming stuff on the fly and as I mentioned before, our cursors are built-in. So cursors are something that github gives you. So if you need to paginate something you can find the cursor so that way, I can do exactly what she did.

A

What the AP, the v3 API and the previous ones, who you have offsets I, can just recreate what offsets are by taking the next five offsetting from that cursor mutations. So and this project is open-source project. I found that really quickly that there was mutations were pretty limited, so I didn't really have a use case to use in the actual project.

A

Yet so I was able to come up an example last night about mutation actually two nights ago and I just wanted to figure out how mutations work, because I thought oh wow I'm about to do, have a talk and I've never done mutations for the github API. So that's the mutation right there.

A

If you go literally to the graphical Explorer D type you like mu, it will autocomplete for mutation and then, if you have mutation, they'll show you all the open mutations that they have available for you and you can read the documentation right in graphical, explore exactly what these are a lot of them. Actually, every single one of our pretty explanatory what it is so add, reaction to issue or add reaction, is something that it's not specific. The issues you can add reactions, multiple things I believe on github.

A

My use case that I thought through that I actually do are issue reactions, but I believe we could do comment at that subject. Id can also be a common idea as well and again. I was able to discover this on the fly two nights ago by using the graphical Explorer documentation, which is great one other feature that I actually wanted to do was actually save data, so I do I, wanted to store notes based on repository data. So if I find a repository, I knew I wanted to follow and stalk.

A

This is basically a stalking pro platform if you haven't really realized yet so, if I were to the stalker repo and I wanted to save note data, there's no way for me to save notes in the github API that I know of I, don't know if it's enterprise feature but it'd be cool to have that in-house, so I'm using a tool called graph cool anybody heard of graph cool. So it almost looks like graph cool oil, but it's really graph cool and I'm.

A

Using this again, it's a database, F graph kill as a service and then some so it provides a lot of things like service functions and authentication for you and I'm using it. So I can persist, data somewhere else and I'm doing that by talking to graph cool, but also using a tool called Apollo. So Paul was a front end client that can consume graphical data. It's more than that. At this point there actually did a graphic you'll summit next week.

A

They also you can also introspect like Postgres databases and and build graph QL schemas, based on that. So we can infer that. So, if you don't want to spend all the time to build your own back in graph QL- and you don't want to spend money on graph cool, which mostly is free, you can also infer that on your existing databases as well and graph cool looks like this, where they actually have a play.

A

Call it a playground, it's very similar to the graph QL Explorer or the github graphical Explorer is what it really is, and they have that available for you, but you could also modify and build your schema on the back end. So if you don't have it back in this is like almost like, if you call it the firebase of graph QL and what the Paulo, if you guys, are familiar for JavaScript and react, there's a thing called Redux and you have a provider and Redux if you're familiar Apollo.

A

Has this very similar thing where you have a provider where I can pass it? My graph QL err on the very highest component. So in my architecture- and it looks like that- and it only really needs to take a graph kill endpoint. So here I'm passing a graph cool endpoint. If you really wanted to, you could pass in the API that github.com slash graph, you all there and you could actually get graph kill directly into your application.

A

So you can bypass the entire fetching data inside of your own functions from Axios and just go straight Apollo if you wanted to as well so and that's what it looks like the Apollo provider, I'm passing in the client there and the one reason I the biggest reason. What got me interested in Paolo is because Apollo's, actually it's a it's a company that came out of meteor, it's really it's meteor and Apollo's is another working group inside of meteor and they're building, tooling around graph QL.

A

So things like analytics based on queries, you can get that through apollo optics, which is a cool tool. You get up to a million requests is free, so highly recommend checking out it because you can. You can kind of like introspect to your queries. You could actually, if you're, using sequel, you can actually see what the sequel, what the sequel calls will look like equivalent to your graph QL call.

A

So if you want to identify anything, that's n, plus 1 queries or opportunity for batching there's a thing called batching with Apollo or sorry with crack yoooo, where instead of going to user repositories and then issues and then going somewhere else, you can actually batch all that into one query and which is really the power where graph code comes from by actually this going just to one query and fetching data at one time, Apollo I say also I'm selling Apollo pretty hard I didn't think I was gonna.

A

Do this, but Paula has a good caching layer as well. So if you want to cache queries, so imagine you, let's I'm gonna get all my repository data I already know: I want this data if I go to the next page. I need that data with me, so you can cache that through Apollo, using under the hood, it's like local storage, I, believe so.

A

My mutation, for this is no longer there going to be the graph gillip it now. I'm gonna make talking directly to graph cool using the day that I've saved previously, but here I can update our repository. So this is a mutation, I wrote, and this is what it looks like. So it's the opposite of a query: queries to give them two gifts.

A

These are the post I'm, just passing in objects into it, so I get to define that I want to string an ID and then I can pass it in to sense trait to graph cool, so I could set that database up, and this is the again. This is the UI. It's still a work in progress. So, however, if you guys are interested, but this is where I save notes. So this is a screen and within my saving note component I have my gradual query co-located in the same component.

A

So this is what another selling point that I like for the front end for react again: Greco's, not just a reacting. It just came from Facebook's and they're they're, really big in the react at Facebook.

A

Apparently so, but I can co-locate my my query right inside the same file, so everything I need to know is right there at the same file and I love this for doing react, native applications where I only have two screens I think this is a little more elegant of a solution when everything is all in the same, the same page that I need to look at and then I mentioned the queries down there, but then I have a regular function.

A

Inside of my react component, where I can then call just about props, create repo which there's a little more react, that's happening there, but just understand that I'm calling a function from A mutation that I create. So my mutation gets added to my rack component and I. Can fire it out to save data just like I would do any other JavaScript function.

A

So I went through all this and I kind of wrapped up that project, but I went through all this and I found out really quickly that there's a lot of things missing from the github v4 API I know: they're working really have hard to expand up with all the missing things like things like notifications is actually something I can't get from the graphical API I still have to go arrest for that. I highly recommend.

A

There's a it's been announced a couple times, but the github platform there's a graph QL section where you can actually request certain things or you can find out if there's certain things coming in the pipeline that you're interested it out. So if there's a business case that you want to go all gerak ul4, definitely check out the forums for that. I also have mentioned a lot about graph cool I spent a good amount of time talking about them, but just last week they announced this thing called will resolve our functions, and these are resolves.

A

Resolver functions are again it's more than just back in as a service. These are lambda functions that run alongside your database.

A

So if I want to fire different functions like authentication or providing a resolver for my graph, QL schema I could do that with a and graph cool, and what that really means is that so far, I've been talking to two different databases: I've been talking to github I've been talking to graph cool I can now just talk to one schema, which is gonna, be my schema, so I like using the term called graph, QL gateway like API gateway and I, could build something just like this, which I've already started.

A

Building this I haven't finished yet, but I could actually take in the github API enter resolver functions and expose that right next to my schema, so every API call all comes from my graphical endpoint as well, but I'm still using the github API. Alongside of it and there's also a cool thing that just was announced two weeks ago called schema stitching. So you can actually stitch like literally take two schemas, so one being the github schema and the other one being my schema and stitch them together, so in Apollo.

A

Actually so with that being said, I can now have one unified schema using github graph cool and not have to talk in my graph cool endpoint and not have to talk to different backends at this point. So that's something I'm really excited about I'm. Hopefully gonna start doing that in the next few weeks. If you want to follow along my repo is at github.com, sighs, be dougie, slash, open, sauced and I just bought a TLD this morning that I wanted an ounce which is open, sauce pizza.

A

So look for pizza gifts and stuff, like that in my app unless I get VC funding, then I go straight serious. We go Enterprise at that point and again my slides are at Gucci net Lafitte, calm, I also want to mention. I talked to the whole interest in graph. Ql actually started from one coming to get hub universe last year, but I got that chance to talk about Kyle, who kind of led the efforts for the graph QL at github. I talked to him on my jam stack podcast.

A

So if you guys are interested in hearing what, whether all the attentions came from for doing Dracula- and if you want to hear like the origin story about that- definitely check out that episode with me. So I get the close to I, get to close the entire conference. So I'm gonna close it like this, don't get lost a saucy github universe.

A

And this is me on Twitter, so if you guys want to follow my efforts in graph QL and this project definitely follow me, ask me questions I, wouldn't say I'm an expert but I've been doing it for at least a few weeks hardcore, not as hardcore for the past nine months. But you know it is what it is. So thanks guys.

A