[00:00:00] Rizel Scarlett: Yay. Thank you so much. Thank you. And those, um, events that you have coming up, look pretty interesting. I think I’ll tune in. Um, so yes, everybody. Um, so my, the title of my talk is called the truth about co co-pilot, but, um, I won’t bury the lead here.
My truth here is that it can help you to level up your. So just a couple questions, feel free to type it in the chat, true or false. Did I write that tweet with co-pilot or not? Am I just lying? Um, just put your guests in the chat or answer in your head and then also true or false. You can use GitHub co-pilot to cultivate your career.
Your engineering career to be specific at any level? Um, obviously my answer here is true, but throughout this workshop, we’ll go ahead and figure out if like my opinion is true. We’re gonna go ahead and debunk this. Um, a couple things will cover if you’re not familiar with GI hub co-pilot I’m gonna explain what GI hub co-pilot is.
I’ll talk about how to use. I’ll talk about how I believe it’s [00:01:00] leveled up your career. I’ll talk about my own personal experiences with it. And we’re gonna go ahead and create a tweet with copilot on our own. If, if the demo gods are on my side. So, um, very quickly, my name is Raelle. I’m a developer advocate at GitHub.
I was a software engineer prior to that. Um, I also helped to run a nonprofit called G code. Which teaches women of color, a non-binary people of color to code. And if you ever see me on social media, I’m pretty, pretty addicted to Twitter. So go ahead and follow me at black girl bites on Twitter, Instagram dove two and hash node.
And then for Twitch I’m black girl bites one, because I got myself locked out of my original account because I forgot my password. all right. Let’s figure out what GitHub copilot is. It is an AI peer programmer that helps you write code faster with less work. Um, so if you’re like, whoa, that sounds crazy.
It’s really not [00:02:00] magic. It can sometimes feel like it. But it’s not gonna take your jobs or anything, cuz it can’t work without an actual software engineer, um, with their hands on the keyboard. Um, from using it, I realized it really can’t read my mind at all and it doesn’t always spit out perfectly well written up to date.
Code, but it is really effective. Um, and I’ll show you how, um, and if you’re, you’re struggling with understanding or envisioning what copilot can really look like, it’s very similar. Or I like to think of it as like how Gmail has. Smart compose, which gives you predictive texts. So in the image below, we see like Jacqueline saying, Hey, I mean, or someone’s emailing Jacqueline saying, Hey, Jacqueline, hope I ha um, haven’t seen you in a while.
And then Gmail or smart compose is predicting that they wanna say, and I hope you’re doing well. So how it works is that get hook. Copilot is drawing context from commons and codes. Um, I mean code and it’s suggesting either individual lines [00:03:00] or little chunks of functions while you type, and it’s being powered by open AI’s codex.
All right. So let’s talk about the various features. So everybody knows the predictive code feature. This is literally a, a screenshot for me. Um, writing a comment that says, write a binary search algorithm, and then I write the first lines and then get how copilot starts auto completing or, or assuming what I’m going to write.
Here’s another feature that I don’t think people are aware of. And I wasn’t aware of until very recently, if it gives you a suggestion that you’re not into, there’s multiple times where I’ve been, like, that’s not what I’m trying to do. Copilot, you can see other suggestions. If you hover over, um, the, the suggested text, it’ll pop up this dialogue that you see here, where you.
Switch between the next and previous suggestions that it has, or you can go ahead and click open up GI hub co-pilot and it will open a tab for you that has, um, multiple suggestions. Mm-hmm um, [00:04:00] also. In terms of other features they have. So there’s a co-pilot extension, but there’s also a co-pilot labs extension.
And I said, I wanna translate this to Python, and this is what it spit out. I’m not a Python ISA, but y’all tell me if this looks right or. Um, also, it has an explanation feature, which I think is really awesome and useful. Um, basically you can put in snippets of code and it’ll, it’ll write out step by step what’s going on for you.
I think it’s helpful if you’re you’re new to a language. Um, so basically here it says we’re, we’re declaring two variables. Um, we’re initializing a low, [00:05:00] um, low to zero and high to the length of the array. Just giving you a little bit more of explanation in some human readable language. Okay. So these are cool features, but how is it useful for me?
How will it even help to level up my career? I don’t know about y’all, but my goals as a software engineer are always to get better and faster at coding and to write cleaner code. I also wanna write good documentation, cause I think that’s really important as a software engineer, I wanna be a better mentor and mentee as I continue to gain seniority in my career, and I wanna be able to pick up, um, new concept.
Quickly, because I know our industry is changing like one day we’re we have react hooks. Next day, we have bun JS. I wanna be able to pick those things up as I transition from job to job or as my, my company adopts new technologies. Crazy enough or coincidentally copilot has helped me to do all these things and [00:06:00] I believe it can help me to, to continue to level up in these ways.
Um, and I will, I will show you how, right. So for coding faster and cleaner, Co-pilot is really, really good at recognizing patterns. So if you’re, you’re doing API calls or you’re even designing an API or you’re writing unit test, once you write maybe the first two lines or the first two, two functions, it’s going to, it’s gonna.
Get a hang of what you’re doing. It’s gonna be like, oh, okay. Raelle was doing a get request. Um, now she wants to do a post request. I imagine she’ll wanna hit this URL and all that. It’ll go ahead and predict that for me. Um, which helps me to, to code a little bit faster also with syntax. It’s really great at things like red jacks.
Um, I don’t know about y’all, but for me, red XX is, is really. But I’m not going to memorize all those different little symbols. It’s just not worth the brain space. I understand how it works, but I would just rather [00:07:00] Google, like how to do some of these RegX simple stuff, simple stuff, but I feel like copilot, for me, removes that step of me having to go Google.
And if I write a comment that says, um, write some regs to validate, um, a phone number. Co-pilot will do that for me. I don’t think in any situation, that’s something that I’ll need to memorize. As long as I know that that’s a concept or a thing that I can do also with comments. I think this is the key to co-pilot.
Um, when I see people complain, I’m like, you’re not writing good comments. You get the best results from co-pilot. When you write clear, understandable comments, which then in turn makes your code cleaner. Cause to me, if an AI can understand. What you are trying to do based on your comment, then a human can as well.
And let me talk about my personal experience, right? I’m a messy coder. Like if you looked at my code, you would be like, girl, it works. But how I, I don’t understand how so. Um, , um, [00:08:00] A couple months ago, I’m still new at GitHub, but we, we, um, collaborated with the, the visual studio code team at Microsoft. And they were like, Hey, we want someone to build a vs code extension.
And they were like, let Raelle do it. And I was like, yeah, sure. But I was a little nervous cuz I was like, now everyone’s gonna see my really bad code. And I have to show it to people that are even beyond my team. Um, but I used copilot during it just cuz I was like, what is copilot? I wanna try it. And after I, I submitted my repository, um, and I mean, or shared my repository with the team, I was really pleasantly surprised to get an email back from a principal developer advocate at get, um, at Microsoft that SAIDAL your code is gorgeous, so easy to read and well, commented.
Love it. I’ve never gotten a, a comment like that before on any of my code. So I really think copilot. Helped me out there, especially since I’d never written a vs code extension before I think me making sure that I was writing clear comments so that co-pilot will be able to help me, um, with the [00:09:00] architecture of that vs code extension made it seem as though I wrote really clean, clear code.
Um, and so for my trick, right, I see a lot of people say, you know, I’m frustrated sometimes get help. Copilot starts suggesting things to me and I don’t. I say turned it off in that moment, right. Especially when you’re laying down the initial structure of your project. I turn my, I turn get hub co-pilot off.
And then once I have a pattern going, I know the logic, I know the things that I want it to do. I’ll turn it back on and have it autofill those moments. And of course I’ll still pay attention to what it’s writing. I’m not just gonna blindly accept it. I’m strategic about how I use COPI. Um, I found another awesome example as well.
So I wrote a, a blog post the other day, and someone commented that it was really good at helping them, um, do these translations for these strings. I think this is amazing, right? As a software engineer, you’re not losing the, the learning experience, cuz there’s no situation where he, um, this, this person or [00:10:00] individual needs to write out in Japanese manually your needs to go Google that.
Right. But. Help co-pilot it saved them some time by just auto filling the, the, the translations in French and Japanese and all the other languages. Um, I also talked about it being good for documentation and helping you level up, um, in writing good documentation as software engineers. I don’t think we’re the best at it, but we can always use improvement.
So if you’re writing documentation and markup and you turn on GI hub co-pilot it’s good. And let me show you some examples. Um, here is Mike McQuaid saying that GitHub, copilot so smart or so good that it already inferred that for his on call document. They should reach out to him at his GitHub handle.
Um, there’s another example here where Hayden Barnes is writing some install instructions, and it has the command on how to install this project on Sue. [00:11:00] Um, there’s also here. Um, this person is an advocate at Microsoft and he admits that half of the documentation for the static web app CLI. Has been written with copilot, which I thought was amazing.
Um, so yeah, so it’s great for documentation, as we can see, I haven’t gotten a chance to write too much documentation, but just seeing these, these testimonials about it just makes me know, like they’re, it’s doing a great job. Um, also in being a better mentor and mentee. So I like to be vulnerable. I’m gonna admit to Y.
I’m not great at peer programming. When I, I, I, I find it useful cause I learn a lot, but I tend to get really nervous when I’m peering with someone who I believe is like way more advanced than me. All of a sudden I lose everything outta my mind and I can no longer type. I found that having GitHub co-pilot.
Activated helps me because now I don’t have to focus on the syntax and worry that, oh my gosh, this person’s watching me not be able to [00:12:00] spell the word function all of a sudden or not be able to, to map an array, but now get help cots auto-populating or filling auto, completing that for me. And I’m still able to focus on the logic.
um, so here’s a, here’s another example of me gaining some confidence while peering. Um, I paired with my manager. Um, there was like this [00:13:00] text file with, with, um, some user names in alphabetical order. And what I wanted to do was every time we added a new username, it’ll get. Inserted in the correct alphabetical position.
I got nervous at that point. I was like, oh my gosh, sounds scary. And I knew the, I knew the, the logic, but I just started to forget the syntax. But once I wrote out some of these comments, it filled out the, the, the, the lines of code for me. And I just double checked as I wrote, I was like, okay. Yes. I think that is exactly what I do wanna do.
And I did a step by step, and then I went ahead and tested it. I didn’t just accept it for what it. And then also last, but not least, least gaining that context for new concepts. I think it’s great for studying for interviews there’s moments where I’m Le doing lead code problems or code wars, and I get a little bit stuck.
Um, and sometimes it’s nice to work backwards, like maybe using co-pilot to, to auto complete, or, or suggest how you can probably solve the problem. [00:14:00] Um, and then looking at that and being like, okay, I’m going to, now that I’ve seen. The solution. I’m gonna go ahead and test myself, and then you can also use that explanation feature to, to help you understand what it suggested this way.
You’re not just stuck and it keeps you moving throughout your, your studying process. Also, it’s helpful if you’re navigating a new code base, you’re new to a project or you’re new to, and, um, a company you can go ahead and take little snippets of code. If you of code, if you something that’s. What are they doing here?
This looks wild. I’ve never seen, um, people use the spread operator this way or something like that. You can take that snippet of code, use the explained feature and gain a little bit more context. Also, it’s been helpful for me as with demos as an advocate. As I mentioned before, with the vs code extension, I had to quickly learn how to write a vs code extension using type script, and copilot helped me out there.
But there’s another example as well, where I paired with the Twitter DEFRA team and they wanted me to. Um, [00:15:00] to show myself using copilot with Twitter’s API V two. And I was like, okay, cool. I could do that until I realized it’s they wanted me to use Tweety, which, um, is a Python library. And I was like, oh no, it’s in a couple of days.
And I don’t know, Python. Um, so using get hook co-pilot really helped me out. And here’s just a screenshot of me celebrating on. Twitter that it helped me. It helped me to write Python. Yeah. Python’s not like a crazy difficult to understand language, but with a few days, um, and peop strangers watching me, it helped me get the job done.
Okay. So we’re gonna go ahead and use copilot to write a tweet. I’m hopefully my screen sharing. Doesn’t get messed up here. Uh, let me go back and let me share.
This window. [00:16:00] Okay, perfect. So I already have some of it written in just for time’s sake and just because demoing copilot can be a little tricky , but as we can see, um, I’ve already used it to help me report the twee library. I used it to help me import a con a config file where I have like my, my secrets and stuff like that.
And then I, I have this, um, passing in the, the parameters or arguments for each of. My, um, my tokens. So what I’m gonna do here is I let I left this empty for a reason, cuz I want us to see get hub copilot going ahead and suggesting that already. It already knows because I’ve been using this pattern of config dot consumer key config dot, um, consumer secret.
It knows now that I wanna use config. Thought access token secret. But we can also say if we didn’t like the suggestion, we could open up, get hub co-pilot and it’ll open up this tab where we can see multiple solutions. It may take a bit. Okay, cool. Um, so [00:17:00] here it’s it’s saying, Hey, maybe you didn’t wanna do this and you wanted to do something else.
Here’s a few suggestions. Um, And none of these are really what I want, but it can also help you to just brainstorm a little bit and think, but we’re gonna accept this first suggestion that it’s given me. And we’re gonna go to the next line. Now. It’s already knows if you, if you notice, I didn’t write that comment.
It knows that I wanna send a tweet. So we’re gonna go ahead. Oh, and accept. And now it has client to update status slash hello world. Now this is nice, but it’s actually using, um, Twitter’s API V one, not Twitter, API V two, which is why I said get hub profile is not perfect. Right. Um, so I wanna prompt it a little bit more to, to send the right thing.
And actually before this, I think I’m going to, um, create a variable instead and pass in that, that variable into here. So we’re gonna create a string. [00:18:00] So I’m gonna say creative variable, um, called tweet. Like this is how specific that sometimes I wanna get with copilot, um, with a string that says, mm, I wrote this tweet with copilot and I’m in CFE do dev right now.
Okay. So let’s go to the next line and see what happens. Bang. It’s already doing it. And it also even handled the, the escaping for me right here, because I used single quotes in both places. So we have that all set. And now for sending a tweet, this is not necessarily how I wanna do it. So maybe if I say with client dot create tweet method, because that is the, the, um, The Twitter API V two, that they’re using now.
And then I say, I wanna pass in tweet as an argument. [00:19:00] And here we go. It did it for me. Um, made a little bit of a mistake. I wanna, I wanna pass in text here cuz that’s exactly what it’s supposed to look like. And then let’s say, oh, it already knows that I wanna print a message that says tweet sent. So I’m gonna go ahead and do that print tweet sent and let’s run this file in my terminal and we’ll see if I tweet it out on Twitter.
All right. So I’m gonna run. Python three sent tweet. It says that the tweet was sent successfully. And let me see if I can, now I have to go back to my, my browser so much switching that I have to do. Um, and let me pull up Twitter to see if it actually showed. Yes, it did. I’m gonna show y’all. I’m so excited.
Share screen Chrome tab. And here we go. My Twitter I’ll zoom in a bit more, but here [00:20:00] it is 31 seconds ago. I wrote this tweet with copilot and I’m in CF, uh, C F E dev right now. Um, so that worked pretty exciting stuff. And we’re almost done here and thank you for dealing with the, the back and forth switches.
So I’m going to go back to my slide deck. And I’m just gonna talk about like, okay. I, I wrote, I used copilot to write a tweet and I was like, Hmm, how could this really be practical? And I, I thought about something, right. If we take it a step further, we can send, um, I, I figured that open source maintainers can use this feature to send a tweet every time they label a good first issue.
New contributors always struggle to find good first issues. Um, and they just like don’t know how to, how to get started. So I think it’s easier to, to make it more discoverable. If your, your open source project has a Twitter account, or you as a maintainer has one, you can go ahead and set up an action or workflow that [00:21:00] every time you label a good first issue, it’ll tweet it out.
And I actually tried it. I’m not gonna demo it straight here just for time’s sake, but here is, um, some of like a screenshot of the workflow I made, right? So I say on every issue, every time it’s labeled, I want this to happen. And if the issue is labeled good first issue, then we’ll run the code below on basically here, I’m grabbing the issue that the event happens on.
I set up Python, I check out the code in my repository, and then I run. The lines that can get this going. So I run PIP three, install Tweety, and then I go ahead and run that file along with the, the API tokens that I have coming in. And here’s the screenshot of the results I was able to, to, to tweet out, um, That this is a good first issue and it was just a test issue that I put out there.
But I think that’s like a cool way to make something like this practical for you, um, who went [00:22:00] backwards. Um, but that’s just all to say and wrap up that GI hub copilot is perfect for leveling up your career at. Any level, I’m happy to answer any questions you have. And I wanna thank Brian for giving me this space to talk.
Um, I’m really glad to just be able to exercise my knowledge and share with other people and to discuss with others as well. All right. Thank you.
Brian Rinaldi: That was awesome. Thank you. Um, so first of all, if anybody has any questions, I know Roelle has to get to a, she has another presentation to get to soon, but we do have a little bit of time for questions.
So we’re happy to answer some questions here. Um, so I I’m, uh, I will ask my own while we wait and see if anybody has. Um, because first of all, I, I realized watching that, that when I used copilot during the beta and, and, and loved it, um, and I found that, uh, [00:23:00] watching your presentation, I’m like, Hey, I wasn’t even, maybe I was under utilizing, or maybe these features weren’t there in the beta, because like, first of all, I didn’t realize you could pull up the tab and see other suggestions.
I didn’t realize I could just comment. Like, it just, I was just seeing it as I typed it. You know, giving me code suggestions, but I didn’t realize you could do all this stuff to kind of really help, um, You know, kind of guided. Uh, so that is super cool. And I did also see a demo of the translation thing and I’m like, wow.
I mean, I just, I don’t think it’s something I would use all the time, but it’s more like, like you said, like if I want to understand a piece of code that’s written in a language, I don’t really know or help somebody else understand my code, that’s written in language that they don’t know. It’s a great way to.
Do that. Um, so that’s really cool. Or to write Python when you don’t know Python as an
Rizel Scarlett: example. Yeah. Yeah, no, I, I have the same sentiments. I was, I was, when I, even though I would GI hub employee, I tried it out and I was like, okay. It just like auto [00:24:00] completes my code. That’s cool. And then I was like telling a product manager.
I’m like, it doesn’t even give me other suggestions. And he was like, yes, it does. And he showed me. So like, I, I think just getting the chance to play around with it more, I’m like, oh, this is cool. Yeah. I was not using this. Yeah.
Brian Rinaldi: Yeah. So there’s clearly a lot of features where we’re not where, I mean, like, I think you just have to kind of get, um, adjusted to and add them to your workflow.
Yeah. So Meg did have a question about, um, was there a learning curve it’s kind of related, I guess, to what we were talking about? What is, was there a learning curve in getting used to working with co-pilot.
Rizel Scarlett: Um, yeah, I think so. I think so. Yeah. Cuz at first I was, I was like, oh man, it’s it’s getting in my way.
It keeps making me accept suggestions that I don’t care about. But then I started like, just especially, because I think now that I’m a developer advocate and I’m not like, oh, I have to get this, this ticket out. Now I have more time. Floor and try out things with features. So I started being like, oh, what if I built [00:25:00] this with it?
Let me see if I change the way I work with it. What if I like deactivated for this moment, but turn it on for this moment. So I think, yeah, there was a little bit of a learning curve, but nothing too crazy. Yeah,
Brian Rinaldi: I would, I would agree. I think I’ve even, I’ll admit I’ve tried other AI coding tools as well.
There’s actually quite, there’s not a lot, but there’s a few options now. Um, and they’re all a little bit different and I think, you know, at first, if you not used to it, it feels like the biggest thing for me, obviously I was under utilizing it, first of all, but second of all is, is trying to. Um, where it feels like sometimes it would get in your way.
Um, Because it’s suggesting things when you really, like, I know what I’m doing here, stop, you know? Yeah, yeah. Um, I think there’s also ways you can kind of configure some of that stuff. Um, and I’d say for me, the other learning curve was just realizing like, cuz it would gimme whole code blocks and just recognizing when it was, um, you [00:26:00] know, there were always kind of like yours where it’s like, that’s good, but that’s.
Perfect. That’s not right. I have to go in and adjust that, right. Yeah, definitely. Um, so, uh, let me get to some more questions. So, um, I know Meg had another question, but, um, I will say like, uh, so just to kind of let you all know. Brazil, obviously not a lawyer. And I know there are questions about the, uh, the kind of licensing, um, and stuff with co-pilot.
So I think, you know, I, I’ve kind of. It’s it’s kind of, uh, I say out of the, our area here, um, in terms of those answering those kind of questions. So I know people have those kind of questions. I know there are people within GitHub who we probably could direct those to. Um, a result could more talk to the usage of it, then the.[00:27:00]
The, um, licensing of it. I know. Yes, I was. I stumbled through that, but that’s,
Rizel Scarlett: that’s that’s exactly. Yeah. I, I’m not a lawyer, so I don’t know the, the legal things. I know the technical parts and I’m happy to, to direct you to someone who, who will know more right.
Brian Rinaldi: So you can, you can reach out to Roelle if you want to get like an email or who, who does contact with GitHub to answer those kind of questions, but, um, And I I’m, I’m sorry if I butcher your name.
Uh, no step asks, uh, can copilot suggest me things like suppose I’ve written a function slash code. Can it suggest me the best approach to follow in in terms of improving time space complexity? So I can Excel at competitive programming, I guess. Oh,
Rizel Scarlett: interesting. You wanna like, you want it to suggest you performant things, you know, I haven’t, I haven’t tried that out, but I think that would be [00:28:00] interesting to like, um, experiment with like maybe what I would do is like maybe in my comment, write something about performance and be like performant way to do.
I don’t know what a good example is right now. but yeah. .
Brian Rinaldi: Yeah, but I’m wondering, like you said, competitive programming, so I’m wondering if, yeah. Would that be cheated? Like like, is there, I didn’t even know there was such a thing as competitive programming. Maybe it’s like a hackathon thing. If that I think, I mean, this.
I think copilot would be hugely useful in a hackathon environment, right? Yeah. I mean, unless they tell you, Hey, you can’t use this cuz cuz we’re competing and using copilots cheating. I don’t know. But
Rizel Scarlett: yeah. Double check with the rules first before you, you turn on copilot, don’t turn it on in an interview.
Brian Rinaldi: yeah. We’ve talked about this at, at, at, at work. It’s just like, do we allow people like to leave co-pilot on, if they’re doing some kind of code test or something and it’s like, mm . [00:29:00] Maybe not. it’s yeah, those gets complicated, cuz it’s like, well, I mean, I wanna know that, you know, stuff on the one hand, but I also wanna, I also feel like.
I mean, doing your job efficiently is like, Hey, that’s a skill. Right. That’s true.
Rizel Scarlett: That’s true. That’s a good perspective.
Brian Rinaldi: Yeah. I don’t know. I think that, I think they voted no,
Rizel Scarlett: up to the company.
Brian Rinaldi: yeah. So, um, yeah, so yeah. Would guess I could imagine it would be helpful in competitive program. I just don’t know.
Is there a way you can configure it? I, your comment thing would. that like adding the comments to tell it to kind of guide it. I would, I suppose be useful in that kind of scenario, right?
Rizel Scarlett: Yeah, I think so. As long, cuz I’ve never done competitive programming as long as they wouldn’t get in trouble, but I think you can, um, write a comment, say that you want the most performance solution or something and uh, experiment, play around with it.
See if it gives you back the most performance solution.
Brian Rinaldi: And so, [00:30:00] um, I guess I, I had one more question. So, so the explain thing you. Tell, tell me more about like, you just basically highlight the code and it will tell you what it does.
Rizel Scarlett: Yeah. I can show
Brian Rinaldi: y’all. Oh, that would be great. Like, I didn’t even know it did that.
Rizel Scarlett: Yeah. It’s actually a separate extension and I forgot to show everybody this part. Um, but there’s two extensions. One is called regular co-pilot and then let me make this wider. The other one is called co-pilot labs. And so what that does, when you have GitHub co-pilot labs installed is on the side.
You’ll have like these two little panels and if you highlight the code, uh, let’s see if I highlight this. Hopefully it works well, cuz this is more of an experimental, like in beta feature and you’ll have like the explain code feature. And if you press. As copilot it’ll [00:31:00] come out with results of like, what’s going on.
This one is, yeah. Uh, this one’s kind of just writing out my comments um, but let me see if code does the following is better. Yeah, it, um, I think it’s easier or better to. Demo with something that’s more of like an algorithm or something like this. Like this is a little bit more self explanatory. Um, but that’s how this feature works.
This is not for this, um, for this demonstration right here. Not the best
Via code extensions can make my ID wonky when too many are running.
Rizel Scarlett: Interesting. I haven’t had that experience. I really haven’t. Yeah. but maybe, uh
Brian Rinaldi: yeah, I I’d say like my limited experience using it. I, there was no noticeable, like slowdown in my computer. Um, honestly the thing it ran so fast, um, that it was often, like I said, the tricky part was getting used to cuz getting ahead of me and it was like, oh wait, it’s got, I was already got all my code written, what am I doing?
Yeah. So um, Okay, well, excellent.