Matthew: Welcome to restock podcast by no code supply Co where we highlight people and the things they build. I'm Matthew. Corey Moen: Cory. Matthew: And today, joining us, we have Matt. How's it going, Matt? Matt: Very good, thanks Matthew. Very good. Nice to be here, join you guys. Matthew: Yeah, happy to have you. ⁓ So what brings you here today, man? What have you built? Matt: What have I built? ⁓ gee, good question. What am I not building? Matthew: Yeah. You describe yourself as a builder, right? I think I've heard you describe yourself that way, like as a builder. Corey Moen: The world. Matt: Yeah, I think I think I'm allowed to say that now because I can actually do it more officially than I used to be. yeah, I am. Yeah, yeah, I feel I've been doing it since I was since I was at school. But now, yeah, now, obviously, we're all in this world of AI and working out how it works. And it's the biggest thing that's facilitated me is I can build whatever's inside my head and my my my head is a Matthew: You're a card carrying member now. Mm-hmm. Matt: is a minefield of ideas that come from observing the world. And now there seems to be a way, there's no barriers. There's no, I'm sure people will say, be careful, but there's no barriers to building what you think. The word, can you do it? And you go, yes, we can. It feels like a, not like a false promise anymore, which is, which is exciting. Cause yeah, I've been able to build things that sat on my ideas list for. Matthew: Yeah, yeah. Matt: many a day, is very cool, you know. Matthew: Right? Yeah. There's never a shortage of ideas. There's always, like I said, barriers or maybe a lack of time. And both those have been shrunk down recently with the efficiencies we've gained. Matt: ⁓ it's, ⁓ yeah, I think we're all trying to work out what, what, what does that mean? Cause it was always, you know, ideas are worth nothing. It's more in the execution. And now I feel like some, some, a lot of people are going, ⁓ now you can execute infinitely execute. I think it, yeah, Yeah, that's true. Exactly. It's it. Matthew: So lots of bad ideas are getting executed on right now. Corey Moen: this. Matt: a good idea and a well, I think it just means that the execution is now getting a bit of a spotlight in like, what does it mean to execute? It's not just, you know, put a brick cement, brick cement, brick cement, you know, there's a lot, you know, like, use the analysis a lot sometimes in the wrong way or metaphors, I always get confused between between the two, I need to work out which one is which. And, know, the idea of building a house is it looks simple. if you think, ⁓ I can build a wall. But there's so much other things as the plumbing, the electrics and you know, I renovated the house of my dad years ago and I remember telling my mates, said, look, I've done the I've done electrics, you know, so they said, ⁓ you can help us out. can save a few quid. And I did it. ⁓ It wasn't far from the truth, to be honest, but I thought I thought I could. And ⁓ before, you know, I got them into lots of trouble with Matthew: Change the light bulb. you Matt: you know, the wrong way to lay the cables, they had to rip down plasterboard because it had been put in the wrong way. And you start to get an appreciation that being able to build is not the same thing as being able to build the right thing the right way. And I think we're in that world right now where it's easy enough to build, but building the right thing in the right way for the right audience that delivers the right value. know, all those things are coming up to the surface and becoming more important, which means that I think humans will still have jobs. Corey Moen: Yes. Matthew: I definitely, yeah, software is becoming cheaper on lots of fronts. Maybe not necessarily cheaper in the amount of credits we're spending, but cheaper in the sense that yeah, anybody can do it. And it takes a lot less time. So yeah, a more bad ideas out there. But I don't know, there's one thing I wanted to key in on that you're talking about there. It's that we want to build things that last, right? I think most of us probably want to that. But with the tools that we have now, you can build things that are a lot more ephemeral or just like temporary, right? Throwaway software is kind of like coming up. Like you can build like on demand app, you know, to solve something that you just need to do for the day, right? But for those of us who like to build and share things, you know, with others, we have to find that right balance of building something. you fast and iterating upon it and sharing it, but building something that's going to last however long I guess you have to, I guess you need to know that going in, right? Like, do you want to build something that is, you know, ⁓ I just need to solve this thing that I need to do today, or do I need to solve something that my team, you know, needs to do for this quarter? Or is this something that I want to ship and sell and support for years to come? Corey Moen: Yeah. I think part of the reason that that question itself has actually shifted though is because in the past, you really had to get serious about that question because it was such a commitment to building something. You know what I mean? Like, hey, are we going to spend the next six months of our time building X feature or product or whatever it is? ⁓ Whereas like that might now be condensed to weeks. And so it's like, ⁓ you know, like I think the question is still valid. You know what I mean? But it's like the pressure on it is so much less. I don't know, sorry. had to jump in. I, I, I'm, I'm curious too, as far as how's that, that relates, Matt, with like, I think the other side of all this is so much of the hype out there is like, go build an entire app versus like, you can do this. You can build plugins and things that fill gaps that you've always had and needed to save time. Matthew: Yeah. Corey Moen: and even do it as an experiment. And then if it goes well, keep doing it. And so was that even top of mind with this app and some of that? Yeah. Matt: Yeah, think sometimes when you're building, especially these days where we're all in different offices, i.e. we're in our own office and we're super connected but we don't have anyone to talk to in our own room. You sometimes need to validate that what you've built is, you get to an inflection point where it's like, this works great on my machine. I can share a few Loom videos. can maybe boot it up for my client and send it to them and it works. But what about when other people get their hands on it? And a lot of the stuff, well, I'm in middle of a huge project now, which when it's released, we'll have two, 3000 eyeballs on it on a regular basis. There's a lot of pressure baked into that. a lot of underlying, it's an iceberg. What you see on the surface is not actually what's being built. Matthew: Mm-hmm. Matt: which means it's a longer term project. You've got to go through a lot more boring stuff. And then on the other hand, you want to build stuff where you go, ⁓ I need this to be out quickly because I just want a bit of a, you want a bit of a dopamine hit, don't you? You want a bit of a, ⁓ I think I'm a builder. I feel like a builder, but I'm what I'm building. Is it worth anything to anyone else? And is it going to help? And that's, you know, that's where it came with the app for Webflowers. I had a problem. I thought I found a solution. And then I thought, well, how do I validate that this is the right solution and I can build something? Because I've always talked to people, loads of people, like Tim who did SVG imports. He's good sort of, he's a friend of mine. I've never met the guy, but we talk all the time. yeah, he's like, just build something. He's like, just build something. I said, I can't, I don't know how to. Matthew: Mm-hmm. Corey Moen: He's legend. Yeah, yeah. Matt: And I haven't had the time or the inclination to look under the hood of how to build an app and go through tutorials. And it was only recently where I thought, do you know what, maybe I just scratch that itch because Claude can help me here. Claude or Gemini, whoever can help me do this. And let's see how far I go. And that's kind of just, I had a coffee one morning. I think I woke up early. The kids were having a bit of a lie in, which is. Matthew: Thank Matt: very rare and I thought, ⁓ I'll do it, I'll do it. can't, I'm not gonna do real work today. I'm just gonna, I know I've only got anywhere between 20 minutes and an hour and a half. Somewhere in between there, I'll have to stop. And I thought, and before you know it, I've got something that works. like, there's something here. ⁓ And then yeah, then I got. Matthew: So you thought, let's see how far we can go. And then you were kind of surprised by how far you were able to get in that. Matt: Yeah, well, I would say I was surprised. I think I was just a little bit proud that, you know, that I managed to follow through on it because that's one of the biggest, I think one of the one of my biggest achilles heels is ideas starting and not finishing. It doesn't. I've you know, I always tell people it's a funny story with I've started two Ironman triathlons and I only finished one of them. And that kind of summarizes my building journey, except I probably started 200 projects and finished three of them. You have to have a level of commitment, consistency. You have to have something that gets you to the end. At least I do, at least. My driver is value. My driver is like, will this be useful for someone else? And if it is, then... that might get me over the line. If no one's interested in it, no one talks to me about it, if I mention it to people, and they don't reply or it doesn't stick, then I probably lose enthusiasm. So this one was the one where like, oh, that'd be awesome. If you could give me an open graph management system in Webflow, then yeah, I'll buy it. Kind of that kind of way. yeah. Matthew: Yeah, you got that kind of, and that's a good thing when you're building something is to get kind of that initial feedback just to gauge like, is there value is there interest here? And like, okay, you know, maybe I should keep pushing here and explore it, you have a tool here, I think you want to share this help you on your journey to build here? we jump into this? Matt: Yeah, It's called Solo. I think it's called Solo term, but it's got the brand name Solo. But essentially, I guess it's a terminal, an app stroke terminal for your agents and your dev, the new dev environment, I suppose. I think I started off with, like we all did with a bit of VS code and trying to put plugins. And then you start going, and then I opened up cursor and cursor feels familiar, but also really, really complicated. You know, a lot of this stuff I don't need to need. And then, then you were, you know, bringing in code into cursor and it was opening all these files and, you know, showing me the code. And I was like, I don't actually need to see any of this code. Like I trust what you're doing right now. And then Aaron Francis, who's, I think he's, He's pretty well known in the Laravel world. And I follow him on Twitter and a few podcasts that he goes. He came up and built this again, I think probably I say it's not vibe coded, but he knows what he's doing in terms of the code element, but he used, you know, crawl code to help him. And, ⁓ and he built this sort of little nice package where I've got my agent, I've got my terminal and I got my commands. You can, you know, Matthew: you Matt: You can do light dark mode, can add your MCPs, add your skills, all these random stuff that the world has started to evolve into and it kind of makes sense and it just gets out of the way. It just means I've got my Claude code stream going on or I've got like four agents going on on one project and then I've got my terminal where I can do the more dev centric stuff, I guess. And it just, yeah, I just. I've just kind of just it lives on my screen now like the browser does. It's it's a same. I grab a coffee in the morning. If the kitchen is quiet, I'll open my laptop and it will be there. And I'll think, ah, what about this? And I'll log a few ideas or things into a milestone that I've got. And then it will say, hey, do you want to join me to clear context and go and press yes. And then I'll open up another agent and it'll be on another project. And this thing is just. It's very well designed in a way that the memory doesn't seem to crash my little MacBook Air and all that kind of stuff. it's just worth, know another shout out for something called Ship Studio, which is very similar. And it's kind of this way, have a little, it kind of just gets out of the way and lets you, lets you claw code if that's a verb these days. what I built ⁓ OG sorted Webflow app into. Matthew: ⁓ did start ⁓ OG in solo here, or did you find it along the way? Matt: yeah, I think the ⁓ first set me or read these docs ⁓ and create the skeleton for Webflow app must have been, yeah, in an here. actually, if I, is it? Here we go. Matthew: So it sounds like it's really, you know, we're going from this world of, you know, an IDE and CLIs where you're just working directly in code to where it's now we have these apps here like Solo, which it's like a workbench for building and you have all the tools and materials that you need in one place. Matt: Exactly, yeah, it is. It is like that workbench. is almost like your command center. Where, know, like, you know, tell me about the app, you know, and then I'm, that's in an agent here. And then I could go, I've got like, I've got three, four projects on the go here. Yeah, well, these are the docs, actually. just, see, I put it in the wrong, the wrong agent. There we go. Matthew: Mm-hmm. Uh-huh. Project, Matt: And then you've got the apps here, and then I've got a few more here. So I've got some Claude agents running here. This one is I was working on recently. I've got the dev server is booted up here. Here's I had to install. I I installed a new dependency. But it just feels a lot nicer, a lot cleaner. It works in my mind. You have a sidebar. I just go and add an agent here, add another Cloud agent, and off you go. Sometimes I get a bit confused with Git work trees and that kind of, and might get a few, but it seems to do all the stuff in the background. So yeah, it's a nice transition. Oddly, it's a non-visual transition from my Webflow days, but it kind of works. I have it. I have the dev server and the local running in another browser. Or like yesterday, it was all in the stream. was looking at email logic, email send logic and rules, and none of it was in a browser. None of the visualization of it was in a browser. So I was just doing it here and chatting away. And yeah, like I used Dex. Dex is like a to-do list thing. I would do like, add another to-do. about, I don't know, contents to Dex. it will, it might tell me now that Dex isn't installed in this project. Probably, that's how demos go, isn't it? It's, but if I did it in here now, I see it, you know, that's running. As we know, like a lot of projects kind of, know, sometimes Claw just does whatever it wants to do. Right, you know, there we go. Matthew: Yeah. Matt: do that and then I'll go back to my next one and go, all right, what's happening? Okay, okay, I hasn't found anything yet. And then I go back here and it's still philosophizing as it does. I mean, it fits with my, it's weird, it fits with my like, I guess my ADHD brain, which is always looking for that, the new tab or the new thing to do. And it means that my productivity, although it's spread out a little bit more, it's... Matthew: Yeah. Matt: a bit more useful, you know, there we go. No decks in there. So, yeah, that's solo, solo term. Yeah, more solo, yeah. And actually going back to that, anyone can build anything. think Aaron has a, he's done a lot of courses on Laravel. He speaks a lot of Laravel conferences and stuff. And I think Matthew: There you go. Solo. Cool. Matt: The fact that he's made something like this, ⁓ I put a lot of, he's already validated himself in my eyes. I already trust him. He's a coherent builder. People have bought stuff off him before, used his software, and I feel that that still has a lot of merit in this world. It's not a case of everyone can build, so everything's good. I think the brand, as it were, is still alive and kicking. Matthew: Yeah. Matt: especially more from a builder's point of view than the actual software itself, I'd say. Matthew: Yeah, so he was already a voice that you trusted as an expert, you know. And so then when he built it, you know, there's a lot, there's, kind of have that trust already built in and it's not somebody who's now on social and being like, hey, I just built this app. You don't know how long it took Matt: Yeah. Matthew: and they're trying to sell you on it and ⁓ the ⁓ may good and it's nice and flashy and interesting but you know ⁓ is gonna fall apart when you actually go ⁓ and it you know ⁓ whereas with you know you know this is built by an expert knows what they're doing you know and I can trust what they're selling Matt: Exactly, yeah. And just longevity of ⁓ support, I guess, to a certain extent. It's not, you know, there's so many things and I love the world of GitHub where, know, it's the first thing you look at as you scan and say, well, when was the last update? And, know, that gives you a bit of a gauge of what's going on. You know, like, Corey, you shared a GitHub thing with me a few months ago and I was like, ⁓ my God, this is amazing. I'm going to check it out, put it in my bookmarks. Matthew: Mm-hmm. Yeah. Matt: about three weeks later I checked it out and thought, ⁓ Corey hasn't been working on this. wonder what's, has he found something new up his sleeve I need to tap him up on, you know? Corey Moen: No, it's the latter of lack of time. Matt: Yeah, well, yeah, but that's the thing though, isn't it? When you see that, it's just all these, yeah, all these signals just give you pause for thought. And, you know, there's so many out there that don't need to be maintained because as you like, I've been at, you know, the OpenGraph app now, it's like, will I do anything with it for ever again? You know, like it's that kind of stuff. If someone looked at that get every, they might go. Matthew: Is it maintained? Right, is it feature complete? And it doesn't have any bugs? And maybe, yeah, maybe it can. Well, that's a good transition here. Why don't you tell us what kind of the initial impulse or need that kind of drove you to building OG Sorted? And then we can talk about what it actually does. Matt: Yeah. Yeah, yeah. Yeah, so I guess for many a year now, I have been building in Webflow and it's been this platform that it's a magical platform. know, it's still a magical platform, but I remember first opening it and thinking, wow, this is doing everything that I don't need to set anything up. I don't need to configure anything. I can just build. And obviously from a visual point of view, that is very powerful. And as you build more projects and as you build bigger projects, I suppose, ⁓ you get your little list, don't you? And Webflow have their list as well of to-dos of what to fix. And that list, your own one gets a bit more particular and stuff you do, and then you work out a different way to get around it. And one of the things that kept appearing on my list and also probably on other people's was a better way to manage my Open Graph images. I don't manage is probably the wrong word, but add an Open Graph image to all of the pages all at once, because that's often the thing that clients go and I do it all the time is, right guys, so this is how you do it Open Graph. Sometimes I'll send them maybe a Figma frame that they can say, look, use this as your sizing. And off you go. And a day later, they come back and goes like, ⁓ can we just add, like, we're just going to add the same one to all the pages first. And then later on when we launch, we'll change them depending on what, you know, and they never, they never do. I, I went around, I actually went to the site that I was my first enterprise site that I built probably three or four years ago now. Matthew: Sure they will. Matt: And I went to check their open graph. I did an audit on their open graph images and they were all of the same apart from maybe five blog posts that had been recently posted. So someone in their marketing team has got OG images as one of their to-do lists, but the rest of the 450 million pages had the same open graph image. And I thought, ⁓ man, this is, I want to get in touch with these guys and tell them they can. And anyway, I, Matthew: Yeah. Matt: It doesn't, people aren't doing it they don't want to do it. It's a big job and it, don't think it has the, um, it's not high on the priority list because it's not, you know, you can't see it on the web flow site. You can't see it on your own site. You often, certain people in the social media team probably see it more often than anyone when they share stuff on LinkedIn or Facebook or Twitter. But yeah, so I thought, as this is, give us any problem or itch to scratch. And so I thought, right, let's see what Webflow have in their API and give it a go. And that's, yeah, I got a decent way through the solution. Matthew: Right. So you had the need and then you kind of said like, okay, is this feasible? Like, can this actually be solved? And that's a benefit of a platform that's extensible is, you know, people in the community, customers like yourself can build plugins and apps that kind of scratch that certain itch that hopefully you have and other people have. Matt: interesting because I think even if I go back a year, I think I opened the designer API docs ⁓ and remember Tim telling me, said, but Lynn, you could do this, you could do this, you know? he ⁓ like, well, to a certain extent because I'm pretty inspired by ⁓ any you watch on ⁓ YouTube. doesn't matter who it is. I guarantee you when they open up their Matthew: You had your mentor, Coach. Matt: their app panel, SVG import is there. Tim will give us some stats about how many people have downloaded it and it's like, you realize how many people's workflows you're touching on a daily basis, which is just, he doesn't make any money out of it, but it's just that, that's the developer satisfaction, guess. But yeah, I'd say a year ago, Matthew: Mm-hmm. Matt: you would have had to go into the docks and had a boat boot around and, you know, trying to understand what's going on and work out where ⁓ you'd have to do a lot of legwork to get to a base camp where then you go, right, now, how do I build this thing? Like, what's the UI going to be like? Like, do I do, like Tim sent me a notion document. We were dumping stuff in notion about, you know, ideas and ways of building and all this kind of stuff. I got bored very quickly, like reading Ocean documents. yeah, like I told him many times, like, dude, we just need to, like, one of us is going to build this. We just need to just build it, you know? Yeah. Yeah. And that's, that's what happened is, you know, I, I opened Claude code and asked, like, you know, you know, you know the way, hello world. Matthew: Yeah, just want to get going. Did you feed it like the developer docs or kind of how did you get it context to get you. Corey Moen: I'm curious about that. like the starting, isn't there a boilerplate like starting app or something from the docs? Did you start from that or did start from scratch? Matthew: Yeah. Matt: Yeah, think I and actually the docs have a really good facility, which as tiny as is, is you can copy it. So I wasn't even sending URLs. I was copying pages and I started at the start of docs, which then divides into, you know, the hybrid app and the designer app. And so I had to start doing a bit of choosing about what, what I had, what I fed. And I basically said, look, just explore it, please explore, explore what's possible. ⁓ And I've got this. Matthew: Yeah, you define the outcome and here's like, you gave it like, here's what you have to work with, you know, So it helped you architect, ⁓ like a plan for how to, how to build the app. Matt: Yeah. Do you know what? think saying there was a plan is, is, yeah, it's, it's over generous. There was no plan. It was a, it was a, it was a chat down the pub with, with Claude code basically. And that's, that's the honest truth of it. But as you said, like there were, there were guard rails and frameworks and I guess books that I sent Claude code off to read, to give them a, to narrow the context down. ⁓ And it just, it kept me there. you know, I would initially, like Tim laughs at this when I tell him how I did it, I was zipping up the zip and uploading it to Webflow and then opening on my project, refreshing my screen. And until he told me there's a way, a developer, you can do it locally and you can refresh. But initially that's what I was doing. I was like, right, let's, can I find out. what the open graph image is for a page. Yes, great. Now, can I change that image? Yeah, okay. Okay, well, how can I change that image? And I'd have to go back to my experience of, right, I could use a URL or I could upload an asset. ⁓ And then I'd say, ⁓ can you build me a facility to upload an asset? And then it would build me a UI for that. And then... Matthew: Mm-hmm. Matt: You just keep layering on those ideas of, what if, what if, what if, what if, what if. And that's essentially what you're doing every time you prompt. I know there are tools out there that give you a visual frame of, oh, could you change this? you just screenshot, boom, can you move this? Could you create two columns instead of one column? I would then maybe do a loom video to say, Aline, or to someone else on my Slack. network and go, what do you think about this? And they'd give me a few little pointers like, wish I could do this and I wish I could do that. I'd even I'd even go over and open up another claw chat and put in my essentially look at look at this code base, what other ideas can you give me? And it would give me a load of new feature ideas. And I'd have to go through and go, no, that's not a good idea. That's a good idea. That's not a good idea. But Matthew: Mm-hmm. Matt: I didn't look at the docs really ever again, don't think. don't think I even, I might have done, say, I've done the find the page, copy the page. If you came to a roadblock and it's like, ⁓ we can't do that. Like, no, you can do that. I'm sure you can do that. It's, and I might go, yeah, I might even go to perplexity then. And, because I find perplexity goes a little bit more into the, the depth of stuff like that. And it goes, okay. Matthew: Try again. Matt: And I'd read it a little bit, try and make sense of it and then say, oh no, go and have a look at this page and oh yeah, of course you do that. And that's how I created it. it was, the UI, I had no design. didn't tell it how to design, what to design. I think I might've given it a couple of color palettes that I was fond of. In fact, I know what I did. I took my portfolio site and I gave it, it's a web flow site. I gave it. the URL and it took the variables out of that site and used those colors and then I just refined the colors that way. Matthew: So this is the App Store page for the app. Look at all these features. There you go. And yeah, so here's what it looks like. Matt: This was actually the trickiest part, I think, because everything else no one sees. So this is your, all of that stuff, all of that decision making, all of that value prop. And again, will laugh because the amount of feedback I've given him on, Tim, you should do this and you should do this and you should say this. And when I got to it, I realized it's really hard. How do you say it? What's the USP when you've got something so Corey Moen: ⁓ yeah. Matt: involved. And again, was going through with Matthew: You have a limited surface area here too, right, to work with for the UI. So you have to be very calculated in your content, in the words that you use to make sure you're conveying the right meaning in the right context without taking up too much space. Matt: Yeah, it's actually something that Jack Redley messaged me yesterday. He goes, because you're talking a lot about the features, but you're not talking about the problem. And I would talk a lot about the features and not about the problem. yeah, and sometimes you have to tee up the problem to then kind of meet it with that solution. And it's easier said than done because Matthew: Yeah, and then how you're the problem. Matt: You don't want to talk about problems. You want to talk about solutions and exciting things. it's, yeah, it's trying to get the fine balance. I suppose that's, can say, we're going back to that thing, we can all build, but there's a lot more to building a business or building a successful thing than just putting stuff together in code. ⁓ Yeah, it just really makes you think and you get a lot of imposter syndrome of like, I can spout this really well and I can criticize and feed back. But when it comes to doing it myself, it's hard. Corey Moen: Yeah. Matthew: What, when you look at kind of introspectively at having now built and released this and it's out there for people to use, what do you feel is kind of ⁓ maybe the biggest learning that you've had ⁓ from doing this project? Matt: Other than pushing it live when it had a fundamental bug, then I had to pull it down again. It's actually good to dwell on that because we all get in a flow, don't we? We all get in a build flow. And that flow, as we were talking about before, it's like the yes flow. You can do it. Anything's possible. And the friction has... there's a lot of friction has gone. the friction of, know, like, when you're making a big decision in life, you know, you often ruminate on it, don't you? You'll have a coffee or you might, you might have a discussion with your partner about stuff. And it's not until you maybe both go out for a walk or you go and have a game of paddle or tennis or whatever it might be. And then something just, it clicks and it settles and the settling is where the decision make sense. And I think we were able to build so quickly that a lot of the settling of decisions, and in this case, a lot of the testing and critique of things, straight up, you just ignore it. haven't got the time. Like I said, you've got an agent working over here, an agent working over there, and it's like, fill in the gaps. Don't sit in it. Don't sit in this thing for a more than five minutes, otherwise you're wasting time. And I had that. ⁓ was waiting for the Webflow team to do their due diligence when I submitted it. ⁓ And ⁓ I waited and they gave me, boom, we're live. I'm like, ⁓ yes, right, LinkedIn, Twitter, Reddit. And then I sent it to a WhatsApp group I was on. ⁓ Matthew, I think I probably sent it to you on Slack. And then I actually loaded it into a project that I was working on. thought, right, I'm better, you know, I better dog food this. And it didn't work. Yeah, but. Matthew: you Corey Moen: Hey, that's what learning for, right? And feedback is a gift. I do think you're onto something though in the realm of like, I have at least personally also experienced this and seen a few others, ⁓ the more building you're building with AI and some of that. my own, I guess, take on why this has been harder is like testing in Webflow. ⁓ I think there's a lot of things I took for granted. Matthew: happens to all of us. Corey Moen: that like, especially if you're building certain things in a certain way, responsiveness or whatever is kind of like it's handled for you, you know, or like CSS prefixing is maybe like a good example of like, we used to really have to sweat those things and test QA and cross browser check and all that in the past. And some of that was just taken care of. And then especially the more you build with like frameworks or something that you're familiar with and you like, I can just set this stuff. I don't even need to double check that it actually looks right on mobile. I know it will. Matthew: Mm-hmm. Corey Moen: ⁓ Whereas then you move into this space of just like AI helping you just code stuff, whether it's an app, a site, whatever. And like, I sometimes forget to like, ⁓ I should go test all those nooks and crannies and corners. ⁓ And I don't know. Yeah, my own lived experiences maybe in that same realm in the last two months of like, ⁓ I didn't even actually test all these things before I like shipped that whatever PR or something. ⁓ So it's very humbling, right? Like, I don't know. Matt: Yeah. Corey Moen: In a lot of ways, do you think that's changed your perspective on shipping stuff? Did it reinstill some fear of, this is why it's hard for me to finally ship it? Or did it have the adverse effect of maybe making it realize, even though it's live and there's a bug, it's not that big a deal. You can fix it, it's out, it's updated. Maybe people won't even remember in the long run. So did that help? Matt: Yeah. Matthew: You'll remember, but will other people? Probably not. Corey Moen: Right, right, Did that change your perspective at all on like getting past that fear of shipping something? You know what I mean? Matt: Yeah, good question. think if you'd asked me an hour after I realized that the fear was real, know, the fear was like, delete this post, delete this post. Like, you know, I was, you know, Corey Moen: ⁓ yeah. down Matthew: You're backtracking as fast as you could. Matt: Yeah, you know on WhatsApp, like I know you guys don't use WhatsApp maybe as much over in the US, but you have a delete message thing and when it comes up with delete for everyone and or just delete for me and when that delete for everyone is not an option, you're like, ⁓ no, someone's and I felt like that, you know, when people were liking posts and but yeah, I think the day after probably, I think then you go through a period of Corey Moen: Yeah. Matt: look, no one's got hurt here. My pride has got hurt. And then you roll on through and the next time you do it, yeah, I did open up 50 different projects and test it 50 different times. So I guess, yeah, the fear, it's not so much the fear now, it's more the, it's instilled that reminder, as you said, of responsibility that when I'm on my screen, it's okay to fail. when you're pushing it out to someone who's relying on that. Again, the OGE image thing, people aren't relying on that for big things, but with other apps, they will be. They'll be working on, it's got a learning center, they'll be working on passing exams. And so if this crashes or it doesn't work or the video, doesn't, like if the exam results are... Corey Moen: Yeah. Matt: or the test answers are wrong or they misaligned or miss marked or it has a material impact. So you have to really think hard about, you know, test and test and test again, or even just, I suppose in our world now is go and look for testing frameworks where they've been validated and they are not just, you know, not just a little plugin that you put in and it goes, ⁓ they'll tell you all the bugs. Well, who knows if that's been tested, you know. Corey Moen: Right. That's totally and testing. I've heard this as well recently on like testing is so contextual is what's hard is like you might see on social, know, some engineers out there being like, I'm, you know, building and testing full end to end with agents only. But then if you actually get the context that there may be like building backend stuff where there is no visual interface, everything that's getting tested. Matthew: Yeah. Corey Moen: is very much easily testable by an agent because they don't have to look at anything. They don't have to click anything. It's all just confirmations from configuration and lack of errors and whatever. But I think for us and the majority of people that I think we in the webflow community and beyond that are building marketing sites and things that are visual, whether it's like tools like this, apps on top of them or the little site. That is a whole other realm of that's actually like, guess I'm stoked for that to be solved. I have yet to see something that really has an ironclad way to solve end end testing, both from like visual, you know, things, accessibility things, but also like in this case of an app, the context of like, it needs to be loaded in Webflow, have the context of your project. which could be different from one project to the next. Like, it handle a project with a thousand pages? Does it handle one with 10? You know, like that whole spectrum is really, really hard to nail, especially just solely with agents doing it. You know what I mean? But I think it's a topic that's important, you know? Matthew: Mm-hmm. Matt: Yeah, I was going to say the context of the bug that I found was actually related to the environment. in Webflow, can upload an OGE image with a URL. You can, I think you used to be able to upload it. ⁓ yeah, can upload and then you can pull it in from assets as well. And I've been testing my project, my test projects were all with the URL. Corey Moen: Right. Mm-hmm. Matt: because that's the way I did it before and that worked. But then when I opened up this other project, it was all brought through from assets and I like, it doesn't work. it's to your point is that especially with a platform like Webflow where on the surface, looks like relatively straightforward, everything happens logically and the same way you quickly realize that no, people build things in lots of different ways. And that's, I guess that's the beauty of Webflow from where it was is like, it's HTML, CSS and JavaScript and there's lots of variables coming in and things come from different directions. There's little hacks that people do and not being able to really test that is something that's always going to catch you out, I suppose. And the environment, I suppose, that you're building in. As you said, development teams have test suites, they have test approaches, they write tests. And even when I first... built my first app with Crawlcode. was like, someone told me years ago to write tests. You should always do test-driven development. So thought, I'm going to do this. And then a day later, all I've been doing is writing tests. And then you realize that the AI writes tests that can be passed by the AI that is designed to pass tests, which means you don't actually get what you need anyway. yeah. yeah, hopefully, yeah, hopefully. Corey Moen: you Right. Matt: Hopefully there's not any other bugs in there, I'm pretty pleased with the result. Matthew: Okay. I have a question for you then. So, you know, webflow has a new, what is it like 1 million CMS item limit. Have you, have you tested OG image on 1 million CMS items in a project? There's your edge case. Matt: Hehehehehe There's my edge case. Well, if they're using 1 million CMS items, they're probably an enterprise which they can pay me to solve that problem when it comes to it. Yeah. Matthew: Yeah, they can hire a whole team of people just for OGs. Corey Moen: Yeah, that's a great idea. One other, I think, use case I'm wondering, because I think looking at the app, what this mainly solves for is static pages. then, to your point, Matthew, on CMS pages, if you want a general OG that's used on all the items of a template page. But I don't know about you all, but something I still struggle having a really solid solution for. is automatically generating a unique ⁓ Open Graph image per CMS item based on fields that are set, right? So like a project example ⁓ I work on is like there's two option fields they can pick. They can pick an illustration, or I guess that's a single reference field, and then they can pick an option field that's a background color of like, I don't know, six, eight options, whatever. And then those two things are what combine to make the Open Graph image. And so like my hacky solve, Matthew: ⁓ yeah. Corey Moen: has often always been using either a published hidden page, or in this case, we just have it draft and you have to preview with custom code enabled. And it's a little form hidden with the HTML to Canvas JavaScript library. And so it's just rendering this exact size in a visually hidden way of the OpenGraph image and then using JavaScript to export that and download it or whatever. And then they upload it. that the dream for me has always been like, I just want automatic, like as soon as they either select those specific fields or maybe when they click publish, you know, a process automatically runs and goes and download, you know, upload or yeah, generates that image and uploads it and then publishes again or something like that. And I've, I've always had the thought of like, maybe this could be done with a cloud app. I don't know. ⁓ but it just came to mind in the realm of like, I don't even know if this could be done with a web flow app. Matthew: Well, I think. Corey Moen: like as a new ⁓ future iteration for you, Matt, but I don't know. Yeah. Yeah. Matt: Well, I think I sent you one, who... That's it, exactly, yeah. And Rob Hope actually sent me that, yeah. Matthew: Yeah, this, right? Corey Moen: Mmm. Matthew: Well, let's describe what this is. Yeah, so this is, you can actually use HTML ⁓ for open graph images. Corey Moen: Interesting. And it's all, you store them in a URL or something so that you can like. Matthew: So yeah, I. No, you're just pointing open graph to the HTML. Corey Moen: Interesting. Matthew: Now, I don't know, one thing I haven't really dug deep into this is how like, when OpenGraph images are shown on different social networks or different platforms, is it how well is this supported? But, I don't know, Matt, if you've looked at it, but yeah, when you shared this with me, I was like, oh, this is exciting. Like, cause this is exactly what you're saying. You can use basically templates and then just pipe in some data fields and you you've got unique OpenGraph images that are responsive. Corey Moen: Right, right. Mm-hmm Matthew: ⁓ as well. Matt: Yeah, well, so what it does, I think, you have almost like, Corey, almost like you did there, but with a step taken out, is it renders on your browser. But if you have a look there in the Twitter image content field, so it goes to a middleware there where it actually goes to that page and takes out that HTML into a JPEG. Matthew: Mm hmm. Corey Moen: Mm-hmm. What is the middleware then? Matt: It's on the OG kit. Well, I guess maybe it's Cloudflare workers or something. It's just a serverless function somewhere that's just doing the work. Corey Moen: Mmm. ⁓ yeah, there's a price for this. Interesting. Matt: Yeah, there's a price for this. Yeah. So it's kind of like, it's kind of like Banner Bear. But Banner Bear is based off a visual editor, I think, that you put in. Whereas this is, as you said, Matthew, this is based on, you create it with HTML, CSS, and then it doesn't need that visual editor because you're building it yourself. ⁓ But it still needs that conversion server, as it were, in there. Yeah. Matthew: Mm-hmm. Corey Moen: Mm-hmm. Mm-hmm. Matthew: is Okay. Corey Moen: And that's where I wonder if you could use a Webflow Cloud app to ⁓ host that middleware, right? ⁓ Instead of their service. Like if you're trying to get around the payments, in other words. Yeah, I asked about the URL thing too, because I don't know if it even worked, if this is online anymore. But there used to be this thing called tail graph. I think it's down. Matt: So. Yeah, yeah. Corey Moen: ⁓ that was like, it was this kind of concept, but you would like go to this little configure. I used it on no code supply Co. Actually, it used to be like what generated all the OG images. And then at some point somebody flagged, it wasn't working or OG images were all broken. And, ⁓ and anyways, it was like, you would configure and style your OG image in tailwind. And then it would, it would give you a URL. And so the, the, that's just the URL you would use as the source of your open graph image. So then that way you could in Webflow dynamically replace pieces of it. So like, want to replace this part of it with the actual dynamic CMS item or whatever, or the title or whatever. It was very clever. The URLs were massive. And so like, I think there was also bugs with that, like where certain social platforms like wouldn't render them maybe because it's too long or hit some limit. But I don't know, it's still like. Matthew: ⁓ because it was like a base64 encoded URL. Yeah. Corey Moen: Exactly, like yeah, yeah, yeah. And so, I don't know, thank you for sharing this one. This one's interesting, I'll look into it. I just feel like this is one of those things that I would love just a more elegant way to solve for it without, I'm cheap, so without paying for another service. Like I just want a way to, you know, stand this up reliably. ⁓ But anyways, yeah, food for thought. Matt: Yeah, well, I think Matthew, you kind of riffed on an idea on Slack of the idea that you could build a component in Webflow that then became your template, as it were. Yeah, and you can either bring in, yeah, you bring in the props and then you can link them to CMS. Now, you you still might have to hide that on your site somewhere or, you know, there's some sort of visibility stuff going on to make sure it doesn't render out on your site, but... Matthew: for that. with props, Corey Moen: Mm-hmm. Matt: Yeah, it's like, it's probably the next step in, you know, it's a... It's for web... Webflow could put this in, you know, click their fingers, you know, it's so easy to build these things. ⁓ Corey Moen: and one flow if you listen to... Matthew: Yeah, that's how building works as we've talked about for the past hour. Just snap your fingers. Matt: Yeah! Corey Moen: I mean, imagine in the page settings if there was a way to just like open the component based open graph generator or something, I don't know, that you could just see a little component and do what this is implying where you can build it all with HTML and then bind in CMS items and all that and they handle the middleware and the in-between would be so, so cool. ⁓ Yeah, I don't know. Matt: Yeah. Matthew: Let's talk about something that's real today. Matt, you want to share it? OG sorting, like let's do a little show and Matt: Yeah Yeah, let's do this little demo of it. This is one that ⁓ I've ⁓ a site up here which is relatively large in terms of, I say it's average large in terms of pages. But just to walk you through the, ⁓ ideally problem that it's solving is ⁓ at the base anyway is, ⁓ we put an open graph image ⁓ on, go and have a look at a page like private assessment fees. ⁓ we click in, but we often end up clicking in. ⁓ Click on the settings and then we'll go down and there's lots of options here for ⁓ SEO settings. So we can see a little preview at the top here, title tags, meta description, and then you go all the way down to open graph and you often say, okay, make it the same as the SEO title, which is great, the description, which is great, and then we select an image. So in this case, we wanna go, okay, let's select an image and it gives you. It's really good, it only gives you images that are OpenGraph compliant, so JPEGs and PNGs. So we might click on one of those like there and add that image. Now, this image is actually a big, if we go actually and have a look here, it's like 3,840 by 2160. So it's huge, it's completely overblown for what we needed to do on the OpenGraph image, but it's there, it works. So we save that. And then we go on to the next one. So let's go on to the next one and we go through and we just have to do it now. Yeah. Matthew: One, two, yeah, three, four clicks, five clicks each time, yeah. Matt: Exactly, you know, the URL one. So a lot of us would go in here and we would go down to one that we use already and we might, you can't do it. No, go into, you, you go into here and like, I don't know, let's go and you'd copy that link. Then you go back into here and instead of doing it like that, you would go in and add the URL and just, but you still do that and still save. So. Matthew: asset management. Matt: That was the little, you imagine we've got lots more here. And then in terms of the CMS, so we've got CMS here, we click on the settings. ⁓ Matthew: Well, and the thing is you have to go into the page settings for each page or template to even know if it has an open graph set or not, right? You have to manually inspect each one, not just to set it, to just even know if there is one. Matt: Very true, yeah. Yeah, to know it is and visually it's very good, but yeah, you have to, there's no way of telling otherwise. You can do an audit, you can publish your site and do an audit with loads of audit tools out there. But the CMS one again is sort of the only options you would have is select from any of the image fields. And in this case, we've only got a logo image, but even if we had set an image up for a field, we'd then have to go in and upload that. Matthew: with an external. Matt: image, that OpenGraph image to every single item, which again, unless you use a CSV import is a pain in the backside. So yeah, so all of those, as you said, all of those, the auditing, the uploading, I thought there must be a better way. So we have... Matthew: That was the pain point. That's the problem you're solving. Matt: That's the problem we're solving yet. So if we have a look at what OpenGraph, OpenGraph sort, OGSorted does is, so the first thing it does is almost act as that audit of your project. So again, this is probably version 25 of the UI of trying to make it clear, but the idea is that we have a little, almost a little table here. We have page, image, title, and description. The idea is we have the page name. We have the folders and the folders are very important. Initially, I couldn't get the folders working, but I got the folders working because we know that folders actually make navigation very much easier. Then we have a little sort of green, green red, like warning system, I suppose. And in this instance, the image is OK. The title is OK. But the description needs some attention. And then Matthew: Mm-hmm. Matt: It gives you a little tiny thumbnail. Now, you there's ways that maybe I can make this a bit bigger because it's sometimes it's hard to read, especially if you've got text on it. But the idea here is that you get an indication of this page has a open graph image on it. And on an individual basis, you can click on here. And what it does is give you a little preview of what it looks like. So again, like you're not far off the amount of clicks, I suppose, that we would have for a setting page. but you go in, if it's an orange, it's a little warning that may be a little bit overcooked in terms of size. Now, it doesn't make a huge difference, but it's not optimum. So that's why it's orange, not red. Then down here is we surface the OpenGraph title and the OpenGraph description. And in this instance here, they're ticked as matching the SEO titles, but we might actually want to have that change. So we could unclick that, we can change this, it's great, and then save that and that will update the page settings for the Open Graph title as well. Also have a little bit here is these are recommended maximums, so we might if we go into one of these, maybe it'll have it. A lot of these are empty actually. If you go over the 160 or if you go over the 60, it will give you a little red mark on the table saying it's over optimum. But as you can see here, we don't have an open graph image on here, so we can upload it. Again, very similar to the settings. But what we see here is we've got four complete and 27 pages needed attention. So we can actually click on this little link here and go. only view the ones that need attention and it will show us all the pages that need attention and we can go through them and we might just go, do you know what, there's 27 of them, there's a lot of pages that need attention. And if I do a quick scan, a lot of them just don't have an Open Graph image, they've got red on all of these that don't say anything visually. So I can, what I can do now is select them all, I can upload. Matthew: then you could deselect the ones that do have images, right? If you didn't want to overwrite. Matt: Exactly, yeah, I could do and actually that's a great shout for a little filter actually is to not show the ones that do have an image because although they do need it, yeah. Matthew: Right, because there are different kind of things that are being audited here that or that need attention, right? Matt: Yes. Yeah, exactly. This is where I jump over to solo term now and type a few things. And before the end of the call, we've got a new feature on the app. But yeah, say, look, do you know what? We want consistency. So we're going to have all 28 of them selected. Now we can upload from the library, upload from the image, or add an image URL. So what we would do this way, we'll just upload from the library. And we'll say we're going to use this one here. and then we click apply and off it goes. So it's applying this to all of those pages in one fell swoop. you know. Matthew: I like the feedback that it gives. You can see it going through each page. Matt: Yeah, I'm not going to take any credit for that idea. That just appeared randomly on one iteration with Claude. Yeah. Matthew: We can thank Claude for that. Well, it's nice visual feedback. It kind of gives you that confidence of what it's doing under the hood, which normally is invisible. Matt: Yeah. Corey Moen: was going to ask this, how many of the little pieces and UI decisions in here were you really hard steering it or was Claude, I think it could be handled this way, even in examples like the image title description column with the color dot, ⁓ was that kind of like you just wanted to capture it and then this was Claude's take at it or yeah. Matt: So I would say each of the features or each of the UIs, I suppose, had probably three or four iterations based off an initial stab that Claude would have had at it. So I guess the initial build was it came out as it came out. And I would jiggle it about, tell it to push columns here. Like this image thing was at the bottom to start with. And then I had to tell it to put a table in. But each time I did give it a kind top line brief, I suppose, it would come with an opinionated way of doing it. And then I would three or four steps of refining it based on my understanding of how clear that was. Often it would not be that clear. I think we had a scale here where it was like, I don't know, there's five different colors of. indicators and it was really hard. They had red as the image being good and then the title being, you know, filled in but less than 80 characters was orange or, you know, just it wasn't, it didn't pass any of my usability tests first up, you know. There's, you know, color coding of complete versus needing attention, you know, even here you go and well need attention is orange. and the intention is red here. So there's little things like that that you just have to keep an eye on. Again, I'm not a designer, so I'm not. I don't see these things sometimes in a way that maybe a very good designer would in terms of the way it hangs together. But it's almost that thing of you know there's something not quite right, and you need to somehow communicate that to the agent and then hope. the agent comes back with something that was in your head when you gave it the brief. But yeah, it's just, I wouldn't say it's throw stuff at the wall and see what sticks, but there's definitely an element of that with a lot of these little things. But I think the beauty of it, sent, yeah, and then you would step away from the nitty gritty Matthew: Yeah, ideation and iteration, right? Matt: of the UI and you would go, look, hey, look, take a pause. we just think of other ideas that we would add? I would look at a space and go down here. This wasn't here before. I would say, ⁓ look, there's a huge amount of space, wasted space here. Is that wasted space? Or maybe we could do something here. And then I started to iterate. then during those kind of conversations, it would go, hey, why don't you put an export and import? ⁓ Never thought of that. I don't work with clients that maybe are that organized with this stuff. But you know what? If it had 500 pages, being able to export what is filled in, especially for the title and description, and if they had an asset library somewhere else that they had URLs for, ⁓ wow, we could do an export and an import. And that would add a huge amount of value. Matthew: And I tested that feature, which worked really well for me. I mean, it was just a test site, but I pulled the export of the things that needed attention. I threw it into Claude and I said, here's what's missing. Fill it out for me. Now it was just a test, so didn't give it a good review. But normally you give it a good review, make sure it did good work. But had Claude wrap it up in a CSV, upload it, boom, everything was fixed. It was a super. Matt: Yeah, just, yeah. Matthew: super smooth process. So I was really happy with that. Matt: It's funny though, it's funny you say that, because when you did say that, I was thinking, okay, well, what's, is this just, is this an unrequired step in that process? know, Webflow have an MCP. When, when do the, like, how long is this going to exist as a utility that's actually valuable because it does something that nothing else does versus just a stop gap from that? You know, even now I'm thinking, oh, maybe. maybe you don't need this. Maybe there's a ⁓ UI somewhere else that allows you to interact with OGs, plus it has some AI. Or maybe you need, you know, have a little, a Claude button up here or an AI generated button up here and, you know, they can bring their own API keys and you click it and it goes, ⁓ do, do, do, do, do. And it takes in the context of the page and all of that kind of stuff. yeah, it's, that's the, I guess that's the beauty of it is it's, it's because it's, You mentioned right at the start, it's not disposable, but it kind of is disposable, if that makes sense. Matthew: I would kind of make the point that all software has an expiration date. Sometimes you get to decide when the expiration date. Sometimes it's whoever the platform or the operating system decides for you. And sometimes it's just lived its life and its usefulness is done. But all software is impermanent. Matt: Yeah, but I guess the world has changed in a way like, Corey, you mentioned at the start, which was, or near the start was the idea that ⁓ the effort stroke costs that you have to put into something to get to a certain point was often, often the blocker for so many valid ideas. And now that barrier has load, you know, that barrier has gone down so much, you can still get to a point where you know, Matthew: website. Matt: you launch something and then, yeah, two weeks later, something better has superseded it. But you don't feel so bad now. You're not going to worry about going home and going, ⁓ yeah, that thing that I've been working on late nights for six months, it's just been built by Google. The old saying of Google will build it, Amazon will build it. Now, Claude, Anthropic might build it, but the high chance you wake up the next morning and Anthropic have built something. Matthew: Good Matt: more than likely won't be an OG sorted type of app for Webflow. But yeah, the MCP will do it. The Webflow MCP will get to a point now where it will be smarter and more efficient at doing this for certain sites. yeah. Corey Moen: Yeah. Matthew: Yeah, but it's about workflows. I think having visual UI and interfaces that help you be more efficient. And this is a very specific workflow for open graph images. But I still think there's a lot of value there. And some people will prefer, yeah, to use the MCP and just tell it what to do and walk through that process. But this is a very structured way. ⁓ to get that big picture of your site and be able to address specific issues directly. Corey Moen: And in a visual way too. I think there's something to be said about like, I don't know about y'all, but the thing I still struggle with, with especially MCPs, not just specifically Webflow, but like anything is, know, Claude, did you actually do that? What you just said? Like I, maybe it's still a trust thing I'm getting over personally, but it's just one of those things that something like this that to Matthew's point earlier of like, Matthew: Yeah. Matt: haha Corey Moen: you don't have to click in every page. You can see everything in a scroll and very quickly get confidence that it's good to go, you're ready to ship, you don't solely rely, well, Claude said it was updated, sorry client. Or that, yeah, yeah, Matthew: Well, yeah, what if Claude ran out of context and it compressed and only looked at the first so many pages but said, I completed it, which it did, but it didn't tell you that, ⁓ I didn't look at all 1,000. I just looked at the first 100. Matt: Yeah, yeah. I guess you're right. The context of this is it's a visual, you're dealing with a visual task that has a visual output. So maybe it does make sense to have that visual canvas to, yeah. Yeah. Matthew: Especially the image part of it, Like actually, you know, to see and verify that the image is the correct one. You need to be able to see it, not just see it. Matt: Yeah, but like even when you think about that, it's like, as I said, like a previous site, it's just the same image. And that's when you get into that programmatic style of, right, the image doesn't matter so much, but what it says on the image or what it says as a headline on that OG is what matters. And therefore you programmatically do it with a CMS field or whatever it is. Matthew: next time. Matt: But yeah, but just like just as an edge case that the one the one thing that really troubled me was this. ⁓ Say, for example, we had this, you know, CMS item and I said, make it the logo image. There you go. NHS Health Service. So you you would save that and then actually recently put in there after two seconds, it will refresh to just make sure it validates what you've changed is true and it will. The idea is that you can't do anything with this because it's already linked to an open CMS field. So again, it's just those little edge cases where you only really pick those things up when you've worked in Webflow for such a long time that you realize that this could be an absolute nightmare for a client who's just been and just updated all their CMS items with all their images. And yes, you're not getting rid of those, but you could override them all very quickly. No one would know about them and off you go and you would publish and two days later, all those landing page, know, CMS items that they wanted to go live don't have the right OGE image. So, yeah. Matthew: Yeah, this makes it accessible for those less technical people who, as we've talked about in experience in sites and projects, they just don't go back in and update things. Or they might hear there, and then they forget, or they get too busy. ⁓ This gives them an easy way to actually do the task in a short amount of time. Matt: Yeah, so yeah, so it's, that's, that's it now. And I say, I told him, said, like, I'm gonna, I want to get as many downloads as SVP import has. And I'd say, I'd say it has the potential because every site has open, open graphs. Yeah. ⁓ but as it like, you know, when I did my coffee business, it was, there's a lot of pressure, like sales, sales, sales, sales, you know, but whereas when you can do this and Matthew: you It's a very universal aid. Matt: you're not as attached to it as we would have used to have been in terms of time investment. It's nice to just put it out there and not worry about it forever and a day, which is quite a nice thing to be able to do because you built something, it makes a difference, but you don't have to worry about it. It's quite refreshing. Corey Moen: remember as well that at least the peace of mind I tell myself for like Mast is like if there's an issue or problem, am I charging you for it? No. Matt: No. Corey Moen: Like, what am I really beholden to? Like, you can get mad if you want, but like, I'm doing this out of my own time and, you know, passion for the community. And I think apps like this are very much the same. It's like, try not to sweat the pressure if people are extra crucial about like, why doesn't it do this? Or I need this. Or it's like, cool. Like, go build it yourself. Anyways. Matt: Yeah, well, there is, but you have it with Mars is that there will still be this underlying ⁓ responsibility that you feel, I'm sure that yeah, it would take time to get past, but you'll still have that, you know, do I just disappear? you know, Mars is, you've put so much love and attention into Mars, but also there's a community around Mars now where Corey Moen: Hmm Matt: it appears on I don't know how many hundreds of thousands of sites. So do you feel an obligation to keep evolving that? do you are you as free as you just said, we're like, hey, sort it out yourself. Because you you don't get where you are, because you're not like an empathetic, diligent person who cares about what you do. So Matthew: Yeah. Corey Moen: Mm-hmm. Matt: interesting to see what you what's your take on that because you've you're there now anyway so Corey Moen: Yeah, I think this, the thing that helps me sleep, I will say, is that ever since we, know, ever since like MAST was like given a name and a thing, because it was always just kind of, I don't know, the underlyings was always how I built for a long time, ⁓ is that it's very opinionated on purpose. Like it's intentionally not having every possible utility class under the sun. It's intentionally not using. I don't know, like just this morning I got feedback. I'm like, should all the form fields be components with custom elements? And I'm like, maybe it's a consideration. Like, you know, I have reasons why. ⁓ And so I don't know. It's one of those things where I'm like, it's opinionated out of my own experience, I guess, and the feedback from others around how it's used and can be most generally used across as many sites as possible. But that is at the same time, a guarantee it will not fit your site out of the box. And that's also the point. You know, and so I'm like, if no one uses it or they don't find it useful, there's a lot of other options out there. I don't know. ⁓ It's, yeah, I don't know, man. It's one of those things that I think maybe another way to look at it is there can be grace in working on tools like this in a developer type community where like, if you're in this community and considering this tool anyways, you hopefully are at least to the point of like, knowing what it is or realizing that you could go build your own version or modify it after the fact and if you aren't then like honestly there's a lot of other options out there even beyond web flow like Squarespace or whatever like if this isn't your thing out of the box then you know that realize there's other options and I'm happy to help educate that I don't know Matthew: Yeah, we're not talking about building consumer products, right? Where you actually are kind of. Corey Moen: Totally. Yeah. Yeah. I think to be clear though where I still sweat a lot is is where money becomes involved like as soon as you know, cuz I yeah whether it's been people externally or ⁓ Man then then I think I take that maybe too seriously like we very briefly no good supply code We were selling like merch for example and like a few people had bad got bad prints, you know, cuz it's all just through like printful or whatever Matthew: Mm-hmm. That's a commitment, right? Once you start taking people's money, yeah. Corey Moen: And I'm like, just, I'm like, yep, I can't take it. I'm taking it down. Like, cause I just have too much internal pride maybe around like, I am so utterly passionate about like making sure that if I am offering some of the world, is a value or a certain bar of quality. And if I, that, if any realm of that is out of my control with something like Printful or whatever, I just can't, I don't know. I gotta, I just gotta, I gotta pause on that. But when it's something like math, that again, that passion still comes through. Matt: You Corey Moen: It's free. You have a choice. I am not charging you anything. You are not beholden to this in any other way other than if you go build a site on it and then there's some bug that's discovered or something like, of course, I'm going to try to help you fix that or whatever. ⁓ But that's just another point of learning. I don't know. Yeah. Matt: Yeah, it's, what I like is that you can still feel the pressure because, you know, there's been lots of discussions, you know, there's lots of, say, webflow frameworks out there and it's, you still feel like you need to step up and, you know, ⁓ not defend, but maybe it probably comes across as defending, but it's like, yeah, yeah, it's like, no, no, this is the way we do it for a reason. Corey Moen: Mm-hmm. ⁓ yes. Matt: And know, fair play, there's never any, ⁓ know, sling your hook if you don't like it. There's never any of that chat, but it's, you have, you still have that, you know, it feels like you can never disappear and let everyone squabble over the, over the mass versus client versus Lumos versus small, medium, large frame. You can never, and then just, I'll let everyone, all the pigeons, they can go over there and I'll just, I'm, I'm, I don't, I'm not going to pay you for this. It's, you're always in there to go. Corey Moen: Mmm Matt: No, this is guys give me give me a bit nostalgia. This is how we're that's it works. This is this is nice. Corey Moen: Yeah, yeah. I think- I think tools get to a place of stability though. Like I was on, I was so passionate about Mast the last, like the last half of last year, I was just like excited about feature after feature after feature. But a lot of that passion came from like rebuilding webflow.com in it. And we learned a ton of things. ⁓ And then just more and more feedback I got from people also like always have had inspiration from things Timothy puts out on Lumos or Marcel was small, large. And so like cross considering those things. Matthew: Mm-hmm. Corey Moen: But here's the thing I think that it's at least at this moment in time, it's kind of in a place of stability that yeah, there's some nice to have things, but like, I don't know, I'm in an in-between state. And then the other side that gives me peace is like thinking back in past careers or then, or seeing how other tools use libraries and things that are also in a place of stability, but maybe haven't been touched in a while. To give you a very literal example is Webflow itself. ⁓ uses, you know, certain JavaScript libraries under the hood to do stuff. Like a good example that I discovered a few months ago, ⁓ partially because of, ⁓ I'm drawing, is it, ⁓ I'm drawing a blank on her name. ⁓ somebody that like posted something about Mast on Twitter. And then I was like, Hey, come hang with our Mast contributors group and all that. ⁓ and it was related to like, ⁓ light, dark functions. Anyways, she discovered Webflow is using lightning CSS is a library under the hood for per page CSS for that function. And this is only one of many others I've seen that I won't name that Webflow is just using under the hood to do things. They're not, everything's not built from scratch and things like lightning CSS don't quote me, but the site looks fricking old. And so I'm like, a lot of these are these like not new and maybe not even like actively maintained libraries that are at in fact, used in big products and just we would never know. You know what mean? And so if that makes sense in the point of like even if I stop working on Mast now for the next two years, I think it could still be useful to a lot of people because it's still in a place of stability. you know, like that's cool. As long as Webflow doesn't ship something that like majorly alters things, which I also have a lot of faith wouldn't happen. Like I just saw ⁓ Merrick who's like one of the OG principle engineers at Webflow. I'm a huge fan of Merrick. He is the brain behind things like the custom element and stuff that I'm like, ⁓ my gosh, Webflow wouldn't be the same without Merrick. He posted on Twitter the other day, just as an example, among some AI hype talk that one of his biggest points of pride in Webflow is that there are sites that were published 10, 11 years ago on Webflow, still stable, still live, no issues. And his point he made was like, next, Jess didn't even exist then. There is something to be said about building sites in a way that is so stable and secure that you can have that peace of mind. And that's part of the major value Webflow brings that people don't always recognize. And so the point is that I think that ties into libraries and frameworks and things that you choose to use and know that they're in a place of stability that you can rely on them long-term and hope they won't just detonate like classic WordPress plugins or whatever back in the day. So I don't know. Matt: Yeah. Corey Moen: Sorry, ranty, but yeah, shame to mine. Matthew: Yeah, there's definitely something to be said about when you are building upon something else, like how mature is it? How stable is it? And I think that should be as much a consideration as, you know, what features does it have? Yeah. Let's ⁓ wrap this up. Yeah. Matt: No, it's it. Corey Moen: love it. Well, y'all, let's wrap it up. Matt: Yeah. Matthew: Otherwise, it'll never end. Corey Moen: You can talk all day. I love it. Matthew: Yeah, yeah, no, this has been great. Thank you, Matt, for joining us and sharing what you've been building. Everybody go check out OG Sorted. Matt: No, thanks. Thanks for having me. It's nice chatting to you guys. I think, yeah, you guys have been a stable in my Webflow journey for a long time. And now the world's opening up to other things. It's good to have you guys on that journey at the same time. So, yeah. Corey Moen: Love it. Likewise, we all figured out together. That's half the fun. Matthew: Yeah. Matt: Yeah Matthew: Absolutely. All right. Thanks everyone for watching. See you next time. Corey Moen: Thank you.