AI can significantly reduce repetitive development tasks by automating architecture creation, data entry, and media management, allowing developers to focus on creative and strategic aspects of their work rather than manual grunt work.
Installez notre extension pour rechercher instantanément dans n'importe quelle vidéo
Etch Live: May 2026Ajouté :
Hey, what's up everybody? Welcome back to another Etch Live. We'll give the uh algorithm a minute to get everybody in here. We'll say hi to some people. Uh we got Zach in the house. Pontis, Andrea, uh let's see. Morates, Northern Fish 22.
I wish these were slightly later in the day so I didn't feel guilty having a few beers while I w there's nothing. I mean, nobody around here is going to make you feel guilty. Like, crack one open, you know?
>> Uh, just do what you want.
>> Yeah, do what you want. There's no rules. We ever done we all work for ourselves, you know? There's no rules.
Uh, let's see here. Sylvia's in the house. Emanuel, welcome.
Uh, Jer, dude, some of these names I feel like I just uh butcher them completely. Justin is here. Graham, Carl, uh Allan, ready and raring to go. Love it, love it, love it. Okay, so what we're doing here is we are Oh, Sweden representing. Nice. Welcome, welcome.
Uh, so what we're doing is we're taking a look at Etch latest and greatest stuff that came out in May. So, this is for May 2026.
Um, if you were at my live stream earlier this week, my uh Giri Co live stream, WDD live, uh, say yes. Otherwise, say no because we are going to cover some of the stuff that we covered in that live stream because not everybody was able to attend, but I kind of want to just get a, you know, a feel for who was here previously, who was not here previously.
Uh we're going to be looking at the new media library, media manager, asset manager area of Etch. We're going to be looking at some of the AI improvements that we have done related to architecture creation like automated architecture creation and data entry.
You know, some people mo I think most people are very excited about the AI features. Um some people are a little critical of them. I do want people to understand that the AI stuff that we're currently working on is not like people make assumptions about it, right?
They're like, "Well, we I don't want AI to build my pages for blah blah." Okay, well, we're not even working on we're not even having it do that right now.
Okay. Um, what we are doing is having AI handle things like architecture and data entry, which I've pulled a lot of people, nobody really wants to do.
Nobody really enjoys, you know, uh, putting data into custom post types, for example, um, as you're going to see here. So, having this do that kind of heavy lifting for you, the grunt work of our projects, it's not try, it's not designing for you. It's not even building for you. I mean, you can have it do that, right? But if you just wanted to create a custom post type for you and put data entry in, it can handle that, right? That's kind of what we're working on right now.
uh some of these like tangentially related skills to what we do in our in our workflows constantly. Um and we're going to be adding to those skill sets.
And I think what you're going to see is pretty impressive and it's very very very helpful. It's insanely helpful. Um I you know I had the team work on this exact flow because it's a flow that I always run into and I'm like I I always get stuck doing it manually and I absolutely hate it and so I wanted to be able to automate this side of things. Um and then that's leading us down the road of course of creating more custom field types which we're working on right now.
And um we are also we had a talk with the team about heavily focusing on automatic CSS skill integration. So, when you do ask it to build for you, if you're an automatic CSS user, it's going to detect automatic CSS is installed and it's going to swap over to its ACSS skill set and everything that it gives you back in terms of CSS is going to be uh optimized for automatic CSS and follow the guidelines that we have laid out with using automatic CSS. Now, that is not a small task. It's it's it's it's going to be released in pieces. Um and the training is going to get refined and the like the skill uh is going to get refined probably over the course of weeks because there's just so many things to account for uh in a workflow like that. But we have talked about it and we are going to be uh drilling down and getting that done. Okay. Um if you have questions hashtag Q on your questions. If you have a demo request, #demo, okay? And we will get to those in uh later in the stream. Uh let's see. It's going to be all it's going to be all yes. Well, it's you know, there were a lot of people on that stream. There were a lot of people on that stream, but I do see some nos and some nos.
>> Yeah, that's good. So, for the nos, you know, they're going to they're going to like it. Um, and then for the yeses, you know, we did have a couple hiccups on that stream, you know, with with regard to processing a large uh CSV file, for example. We're going to try it again today and see if there's been any improvements in that area.
>> Uh, I think I think the the team did a little fine-tuning behind the scenes, so we will check that out.
>> Uh, >> they did. Plus, I think that we couldn't uh do we couldn't actually field the custom uh fields, and now we can. Oh, that's a great point. Okay, so we talked about Yeah, the limitation on the stream earlier in the week was, "All right, guys, this is going to create the custom post type and it's going to create the fields, but we're not expecting it to put the data in the fields because that wasn't part of the skill yet." Uh, now it is or it should be. So, we're going to see if it can actually put the data in uh as we do this. Okay. Um, if everybody's ready to go, let's put an LFG in the chat. Uh we do have Matteo in the house as well. So any super technical questions that people have, he'll be answering. Of course, we answer a lot in the chat, so keep your eyes on the chat as well. Um but if there's anything super interesting, then we will answer it on the stream. Okay, I'm going to go ahead and I got to find my two guest plus screen share here. Here we go. Uh nope, this one right here. There we go. Okay.
>> Yeah. So, I I was talking to Grock here about like some data structure stuff.
I'm gonna I'm gonna essentially on the fly have Grock uh generate a CSV for us, and that's what we're going to be using to import. Um I'm going to kill these things here. I don't need all this stuff. Okay, so I've got I think I think we're good on local. What do I have in here? Team members, vehicles, pet. Just some like nonsense stuff going on in this local. Um also, we we improved the design. Uh, let me show you guys this real quick. We improved the design of these two actions that you can have or two modes of the AI chat I should say.
And I think a lot of you I'm very happy about this. I think a lot of people will be happy about this. Build is the default mode. So for all of you who were trying to get it to do things not realizing you were in ask mode instead of build mode, build mode because ask was the default. uh and then it w it was acting like it was incapable of doing the things you were requesting. That will no longer be the case. It defaults to build mode and you have to opt into ask mode. Okay. And then we just improve the uh design a little bit here. Okay.
So, what we're going to do uh let's why don't we do media first? Why don't we do media first? Um so, what I'm going to do actually is go to static.wp.com hwp.com because this only works right for us. Is that correct, Matteo?
>> Yep.
>> It's only special to special people.
>> Yeah.
>> Okay. Um and I am going to Let's pull this over here. We don't need this all on the stream. All right. Um let's go ahead and get to my email here.
All right. Um let's see. Where is Did I Did I type in my email correctly? Okay. Hopefully it arrives. Oh, there it is. Good. Okay.
Perfect. Perfect. Perfect. Is this this if Can I put this in or do I have to hide this?
>> No, I would hide it just to be sure because I can't remember if uh it actually does it as a password or not.
>> Okay.
>> I think it's one time only, but you know, >> can I get these tabs? There we go. It's the hardest the hardest job in the world is getting a window into a tab. Okay.
Um, so, oh, by the way, we had AI build this in about 8 seconds on the last stream. Um, and oh, wait, this isn't the one it built. Okay, good. Oh, we we could build it again. I don't know.
Okay. Um, let's go ahead and go into media asset manager. So, I'll direct everybody's attention right here to asset manager. Uh, this used to just pull up, as you saw in the WordPress version, the WordPress media library, the modal, uh, that you guys are all used to seeing. Obviously, a terrible experience. What this now pulls up is the Etch native asset management experience. Um, so I'll direct your attention to the top left, which is where you can define collections. Uh, and things can live in multiple collections. So there's a folder icon here, but these are not treated like folders. They're treated more like tags.
You can put the same image in multiple different folders. You can create new ones. Um, all the stuff that you would typically expect here, which means once this is released for WordPress on the WordPress side of things, this is at static. But once it's released on the WordPress side of things, you will no longer need Happy Files or any other of those like folder-based media management plugins. Uh remember we made promises very early on that Etch would replace a lot of plugins, a lot of add-ons, uh and unify a development workflow. This is one of those things that is already doing that, right? Um I will go down to the bottom left. You're going to just have a quick look at how many assets you have in the asset manager and the storage being used. Uh this is currently using Cloudflare R2 for the storage and uh you can see here that there are three files essentially. This one is a very low res image so it's very small but this one is a higher res image so you see the large version of it. You've got the file information up top. You can copy the relative path. You have the alt text that you can assign. Uh this will show the generated sizes. We're going to be working on source set uh very soon.
source set support for this and then what collections it's in. Um, you can also drag. Okay, so if I want to just drag something into one of these collections, I can. Um, if I want to bulk select, I can bulk select. I Oh, well, there's another improvement. Okay, you couldn't do this earlier in the week. You can bulk drag into uh collections. Okay. And that's a nice little stack kind of animation that it's uh that it's using there. That's nice.
Um what else? What other actions do you have? Oh, look. Look what we have down here. We didn't have this earlier in the week as well. So, when you have multis select, uh you have a little action bar now. Okay. So, you can add all of these to one of your collections. And look what's coming soon. Hint hint hint, guys. Uh you can delete them. But look right here. This is the fun part. This is the part where I was like, "Hey, yet another yet another plug-in that doesn't have to be used or a pre-upload kind of workflow that doesn't need to happen.
You are going to be able to natively compress, natively resize. There's going to be a bunch of different settings like the aggressiveness of the compression, uh the max size, like max width or max height of images. Uh, so you know, it's a constant problem where you have a 5,000 pixel wide image that nobody needs and you're like, "Okay, I don't want to just send that to the server and start serving that. I want to I want to have like my upload process autoresize those." Uh, and you would say like I want the max to be 2600 pixels and therefore it resizes it for you, optimizes it, handles all of that. Uh, there's still some infrastructure that we have to talk about. Like for example, do we retain the full res version of the file [music] somewhere? Okay. Or do we not retain it? Or maybe that's a setting that the user has control of in the settings panel. Okay. Um and then all the other settings that come along way like conversion to WEBP perhaps, conversion to AVIF perhaps. Okay. All these things have to be thought about.
The reason these are often plugins is because there are a lot of like little switches and levers that you can pull to control this kind of workflow. And we do have to think about all those things. We do have to uh bake them in. But I think this is really exciting. The fact that you're not going to need a separate compression workflow, whether that be a plugin or pre-doing it like I often do with Squoosh. You see me on streams, I'll use Squish for individual images. I will use uh this native app uh Squash. I don't like these names. They come up squishes and the squashes. Um, these these are the workflows that people are used to. It's annoying. Honestly, it's annoying. I would just like unify development environment. I would like to upload my images. It resizes them. It optimizes them. And I can just keep going. I don't why why am I in squishes and squashes? I I don't I don't want to be there. I want to be where I'm already working, right? That is the goal. Okay.
So, that is new. Uh, you can search. I believe this did not work earlier. I don't have any like the problem is these file names that I uploaded. Uh, hero one balance. I mean, does this really like Is it going to find that? Let's Hero.
>> Let's find out.
>> One. I think I'm multi selected. So, I don't Hero one.
>> Okay, it did find it. Hero one balance.
And it has gone down to one file. Okay.
Does she have a name? What is her Oh, they they're they're all Hero One. Okay.
So, main image. There she is. Okay. So, search is working. That was not working earlier in the week. Filtering is not working yet. Uh, but we do have a grid view, which is really a um, uh, what is this called? What is this called? Uh, the word mosaic is stuck in my head, but it's not it's not mosaic.
>> Oh, you know what I'm talking about.
>> Masonry.
>> Masonry. Masonry. I I was right on the first two words. Uh, or the letters. One letter. M. Okay, got it. Uh, let's see.
I hope there's a squish to complete the trio. Uh Justin, you know, we we we may call it squish.
>> We'll call it squish.
>> I mean, I don't know. I feel like that's probably already taken. I feel like we're gonna get a lawsuit if we if we call it squish.
>> Um Nathan says the team is moving quick.
Well, I'm glad you feel so. Sometimes it feels like it's not quick enough in in this, you know, this in this this kind of AI based uh environment that we find ourselves in, you often feel like you're moving at the speed of light and also standing still simultaneously. Uh, and that produces a little bit of anxiety, >> but it's okay.
>> Yeah. With all the the the wall of AI bros saying, "I vibe coded 10 apps in my weekend." And you're like, "Okay."
>> Yes. Yes. Um, by the way, in the inner circle, I'm I'm doing a whole series on a vibecoded uh project that I've been doing just showing all of the ins and outs, all of the details, and all of the truth behind it. Because any I I I I I would just tell you anybody saying that they vibe coded anything relevant in like a day, a weekend, even a week, uh they're absolutely out of their mind.
They're absolutely out of their mind. Uh they've they've done nothing relevant is is the truth of the matter.
>> Or they just vibe coded something so small that it was probably not going to require a lot of time anyway.
>> Yeah. Yeah. Nothing substantial, you know. Nothing nothing substantial. Um, oh, Andrea did say it's a one-time code.
So, we, you know, they could have watched me, they could have watched me put it in. That's good to know for future future streams.
>> Okay. I don't think there's anything else here. Now, um, what I do want to do is save here. And I'm going to hit this button right here. If you look down here, this is different than WordPress, right? This is Edge Static. Somebody did ask uh regarding ad static. They said that was cool, but it would be cool if it was completely decoupled from WordPress.
And I was like, well, that's great to hear because it is like they they thought I was in WordPress and deploying to Cloudflare from in WordPress. And I was like, no, no, no, no, no, no, no, no. This this has nothing to this is not attached in any way, shape, or form to WordPress. So when I deployed to Cloudflare right here and let me zoom out and I get that uh this this confirmation message up top deployed successfully and I go view that on the front end this is completely 100% Cloudflare and this is free. It's literally hosted for free. Like there's so when you use a Cloudflare worker and they're so generous, right, with bandwidth and things like that, very static sites like this, brochure websites. And when we say static, let me remind everybody, it's like semi-static.
It's semi-static. Like it's it's for simpler brochure websites, but at the end of the day, uh it is database driven and it is going to support loops and it is going to support uh logic, I think.
Is that right, Mattea? Get some logic going. Yeah. Yeah.
>> Loops, logic, content types.
>> Okay. So, let's call it semi-static.
Let's let's or semi semi dynamic. I don't know. I don't know. We we may maybe need to find a better name. Uh but it's going to be very very powerful in a lot of ways. Uh so for those of you who are like, you know, for 80% of the sites I do, I don't really think I need WordPress. You would probably be building with this. And then the 20% where you're like, no, this is definitely like a WordPress project.
Then you keep using WordPress. And the benefit is it all happens in Etch in the same development environment. So you're not having to switch contexts, you're not having to switch environments, you're not switching workflows, none of that. Everything is exactly the same regardless of whether or of you are deploying to. Okay. So that is nice. Uh I will inspect here and we'll come up here. I mean, look at the look look how simple this is, you know, just absolutely crystal clear, clean as a whistle. Uh, exactly what you would expect, right? And of course, you see a lot less stuff going on here because WordPress like loads a bunch of default nonsense that really nobody needs. Uh, and none of that is here. So, insanely insanely insanely clean. Uh, completely decoupled. Really fantastic.
Okay. Um, let's see. Uh, I live in the real world. I understand how long dev work takes. Loving the progress so far.
Yes. Yeah, it's um, it's not it's not quick. It's not easy. Okay. Um, do we want to look at anything else regarding I think that's all for media, right?
That's that's pretty much it. Oh, I didn't show bulk upload.
>> Could do this.
>> Okay, let's do that. Um, let's go to So, I People are probably wondering about uh bulk upload. Let me go Let me go to uh what's it called? What's What's the What's the website I go to all the time for images? Uh, >> Unsplash.
>> Unsplash. Okay. Um, let's grab this one.
And I'm going to show you the workflow of optimizing images that that nobody likes, right? This is kind of what we're trying to eliminate. Uh, so I get the uh let's take this one here as well. Okay.
Uh, let's take this one. Let's take this one. It's better with people. Okay, just three. All we need is three. So, I can't squish because squoosh only does one image at a time. So, what I have to do is I have to squash. Um, so I'm going to come here and I'm going to go to desktop and I'm going to go to Unsplash. I'm going to grab those three right there.
I'll import them. And then this is where so imagine this just happening natively in Etch where it's asking you okay we're resizing it what are the max dimensions right um I never really do adjustments or effects or borders or watermarks or anything else but I do say hey by the way these files are not WEBP I want webp so uh convert them into WEBP it's going to do that part of the process um I have a quality control here that I can that I can do and then it may uh suffix the files as well if I wanted to which I think would also be great if the etch native one did this as well. Uh it also has a convert everything to lowercase which I think is great. Uh you can remove whites space which I also think is great. Uh retain date metadata if you want to. Uh and then I am going to choose and we are going to choose this exact same folder and I will go export images. It's going to do some thinking and then I'm going to hear this stupid sound and then we're going to be able to move on with our life. And so why couldn't that happen in Etch, right?
Like why why did that need to happen in some other app? Like and I'm switching back and forth and now see now I'm trying to find the files again. Like nobody likes to live this life, you know. Um but I can very easily just drop three new images in there and there they are. They appear, right? And I can go back to the list view and see them over here. I can open and see them uh in the uh I don't even know what this is the maybe we'll call this the inspector. I don't know. Uh but that's it. It's as easy as that. So uh then I can insert the Oh, one other thing I should show.
So if I'm here and I click on this thumbnail to add an image, I get the same workflow guys in a little modal form. Okay. So I can just come in and say, you know what, no, I want this one.
Okay. And so we select a different image or I can come in and say no I really want this one and we select a different image right there. So uh we've replaced both flows both the global asset manager and the picker we'll call it like when you're choosing uh an asset to put into the page.
I think that's it. I think that wraps it up.
>> Yeah.
>> What's up Lizzie?
>> I think so.
>> Morning Lizzy. Morning.
>> Yeah. I saved a bunch of questions. Yes.
>> Do you want me to go through them? Okay.
So, let's [clears throat] starting with start with the asset manager questions we have. Will assets still reside in the WordPress media library but accessed via the new asset manager? So, that's when we move the my we backport the my the asset manager into the WordPress plug-in. Yes, absolutely. We are still going to save the the assets as the native media. we just give you a different, better, more comfortable view uh and more powerful actions.
Um what is the estimated timeline for the asset manager to come to WordPress?
Um I haven't discussed it with the team.
Um I'm hoping we can do something uh next week like hopefully by the end of the week, but um it's I I gotta see with the team uh if if it's feasible. Uh right now we were just finishing up some other work that we that we uh wanted to squeeze into the uh uh the asset manager like compression before backporting everything to WordPress. So it might come a little later but it will come with compression as well.
>> Yeah.
>> Um then we have will this new media manager UI also be available for clients or just inside of the etch editor itself which is a good question. And I don't think we ever discussed this. I honestly don't >> you read that one again.
>> Uh if the media the asset manager UI is going to be also be available outside of the etch editor for like customers >> uh clients >> and I I don't know like we haven't talked about it and I I think for now it really is just for the editor.
If if there is enough >> if it makes sense we'll discuss it. I think >> yeah, I think what it'll probably more be like would be the uh if you guys are familiar with in WordPress the happy files UI where it's like your media and then it's got the collections off to the side like it's not the entire like branded etch native experience but you do have access to like the critical things for example uh still the client is still able to put media into collections for example um and then perhaps some way somehow uh some form of extension of the well really at that point like if you're doing the image processing if you've set up the settings in Etch any image uploaded by the client to the media library would essentially just use those settings like we don't have to give the client access to those settings it but uploading to the WordPress native media manager would hopefully be able to tap in and use the the image processor that we've put Yeah. Yeah, it it can have some side effects. I don't know that the that we're going to port the UI outside of etch. Uh but we'll we'll figure it out.
>> Yeah.
>> Um all right. So, uh last question that I have saved on asset manager. Uh will all the uh Oh, no, sorry, two questions. Uh will the media library support asset replacement swapping one image for another everywhere on the site? I think it's on our road map already. like we it's definitely something that we touch on.
>> Yeah. I mean, it's a it's been a gigantic limitation of WordPress since the existence of WordPress. And I, you know, I it's one of those things where I I know that a gazillion people I I think that's a real number, a real stat. A gazillion people have said, "Hey, shouldn't I be able to like replace media in this like 42% of the internet content management system?" and WordPress never found it in their hearts to uh actually add it. So, uh but we will because yeah, why not? Like it's it's always never made sense to me where I'm like, okay, like I just need to like replace that image in the library so that I don't have to go replace references to it or play the game. is everybody's probably familiar with the game where you try to use the same name and upload and then for the next like six months you still see the old version because of WordPress's internal you know nonsense of I I don't know what causes that problem but it it is a constant problem in WordPress um you know whether it's caching or or what have you but I'm sure everybody's familiar with that game and it's not a fun game and you know it's 2026 the idea that you still can't replace media easily in WordPress. Maybe they fixed it since then. I don't know.
I you know, but it's it's something we'll we'll add for sure.
>> Okay, last question on the asset manager. Uh will all the image uploads and compression stuff be also available via the upcoming API? Um great question.
Um, I imagine yes, but this is something that we need to discuss with Chris who is designing and developing the API, which by the way is going to come next week. The first iteration is going to come next week.
>> Okay.
>> So, we'll we'll know more soon. Um, I also have a saved a few questions about that static in general. Do you want to go through them and then move on to the AP to the AI stuff?
>> Yes, because once we get into WordPress, we'll just be focused on that. So, >> that makes sense. All right. So, first question. Will Etch static be included in the current ETH license or will be will it be a separate product? Um we've addressed this in uh in the Tuesday live stream as well. We have not worked out the details of the plans yet. So more on that soon.
Uh can you explain your current thinking on how a static would be rolled out available to current edge licenses?
Sorry, the same question. Yeah. Um, is it possible to host ETH uh on my own without a um on my own server without WP? So, Edge static on my own server?
No, it is not. At the moment, etch static is a cloud application that hosts the builder with a database behind it.
And um and you provide the um infrastructure for the live website. So, you provide where the website is hosted.
We provide how to edit the website. Uh there are no immediate plans for making this available in a self-hostable fashion. Th this may change in the future, but at the moment it is. Um this is how it is. Uh can we save our site locally so we can then uh SFTP our work to our server of choice? Um you can export the site already and you get a zip file with all of the pages and all of the assets. And so you potentially could just uh SFTP that to a server of your own uh choice. In my uh prototype of edge static, I had also baked in a an SFTP server. So you would just connect an SFTP server and deploy to that directly. So it may come in the future.
Right now for the deployment part, we're focusing entirely on uh on Cloudflare, but more deployment ways including SFTP may be coming in the in the in the future.
And I think this is all the questions that I had. Let me just take a look. Um, uh, there's a question about, um, does the new asset, uh, media asset, um, accept dynamic tags like site.escription?
I am not sure. I'm not sure. Um, >> what would that be?
>> I don't know if Andrea, who is in chat, might know.
>> They're talking about dynamic data references or something.
>> Yeah.
>> Okay.
Uh I do know that like these collections are going to be loopable. They're not loopable currently, but they are going to be loopable. Um so the ability to make uh galleries very easily, for example, and keep everything organized.
Uh that will be coming.
>> And then there was on Tuesday a question about nested collections and I think I answered that for sure one level deep.
Uh maybe two, I don't know. We haven't really discussed it. Uh but I do think for sure like at least one level it's got to be.
>> Yeah.
Um there are also a couple of questions about Tommy. I think uh we might want to address those and then move on to HAI.
>> Yeah.
>> I'm going to I'm going to get uh through that uh real quick. So yes, as you guys have seen in uh you may have seen in circle um we have parted ways with Tommy. We parted ways amicably. Uh we are very grateful for the work that he's done. His fingerprints are all over the ACSS dashboard and parts of Etch. And we have uh nothing but good will, good um good feelings and um and um and and hope nothing but the best for him in the future. It's been a an amicable split.
Um I see some people worried about um uh him leaving after Waji leaving and this looking like a brain dump. This is a team restructuring. It's fairly common in software uh in software um companies to restructure after [snorts] extended periods of time. Both Waji and uh Tommy had been with us for a long time. Uh we um I I I don't want to go into the details of the decisions because these are very personal. It's it's it's not um it I I don't want to I don't want to do it. just I hope that you understand and trust us with the decisions about the stuffing that we're making. Uh that I hope that we've earned enough of your trust to know that we know where do what we're doing. Um that's that's all I can say.
>> Yep. Well said.
Um all right, I think on to WordPress.
>> Yeah. Uh let's uh let's take a look at the the AI improvements.
>> Yep. Yep. Uh, and Karen said, "Galleries coming to is awesome. Had to do my last one with ACF." That's just to reiterate the goal. The goal is that's another plugin we don't want people to have to use. We felt >> sorry.
>> Yeah. Go ahead.
>> There was there was another question regarding uh Tommy. Yeah.
>> Uh, somebody saying uh do we need to be concerned uh for etch that it wasn't financially viable to keep Tommy? Uh, and no, I'm sorry that's imprecise. It was absolutely viable to keep finan to financially keep Tommy. The decision is entirely about the shape of the work and what kind of team does it make sense to match the shape of the work. It has nothing to do with financials.
>> Yeah.
>> Uh and now we can move on from this.
>> Cool. Um so ACF, Metabox, uh I mean there's jet engine like major major major companies have been built around this idea of proper content management.
And it's always been awkward because I I'm pretty sure most people who understand content management and who have been using custom fields and custom post types this entire time have always felt like I mean that's what a content management system should do. Like the idea that these things are have to be bolted on was always annoying. Not just awkward but annoying. Uh, and that was always one of our visions for like there's just those are plugins you should not need. Um, so in your development workflow in Etch, you should be able to create custom post types and custom fields and link them up and do advanced stuff with them. Conditional logic with the fields. Um, have advanced field types. um have relationships, birectional relationships, um onetoone relationships, whatever you need to do in in terms of architecture and and like building a truly dynamic website, you should be able to do that all natively.
And and WordPress has always had pathways to doing all of that natively.
It just never built a UI for it. And that's what ACF essentially did. That's what Metabox did. That's what Jet Engine did. And with Etch, you're just not going to need those tools anymore. Um, and it will start out with, okay, I just need basic custom post types and custom fields. Cool. Use that. You don't need ACF. If you're doing something insanely complicated and we're not there yet, then you might still need ACF or metabox for the time being. But the ultimate goal is those things are never needed at at any point in time in the future once we get to the finish line to our destination with regard to those things.
And even instatic, you will have they will be called content types. I've never liked custom post type like that's WordPress has never been able to escape its like blogging history, you know, and so the idea of like a custom post type really speaks more to uh posts with blogging. It's really custom content types, right? Uh and so that's what it will most likely be inside of Edge static, but you'll be able to do that stuff in in static as well. You won't even need WordPress for that. So, um, all of that is coming and it'll just be more stack unification, getting rid of the bolt-ons and the clutter and one unified workflow. That's always been the vision.
Um, okay. Let's go over to WordPress once again. So, I'm going to leave this.
I'm going to leave this. I'm going to leave Oh, I don't want to leave that.
Let's go reopen. Close tab. I can leave Unsplash. Okay, here we are back in WordPress. So I was having let's retry this. I was having a conversation with Grock about um creating a CSV for locations data. So very common that a small business like a local business coffee shop whatever you know they they have multiple locations and the client's like hey we got you know we got 10 locations. Maybe you're doing some sort of regional or statewide type uh business. Okay good. It's it's regenerating. Okay, it lost my last chat apparently. Okay, so I want you to create a CSV file for a small business that has 10 locations. Uh, I need now you wouldn't do this in real life like your client would provide you the CSV of this data. I'm just having AI generate data because we don't have a real client, right? So, I need um let's see.
I need columns for columns for name, street, address, city, state, zip, uh phone, and and we'll just do that. We'll we'll try to keep this nice and simple just for demo purposes. Okay, let's see what it comes back with.
And I'm going to show you guys the flow of kind of how this works practically speaking in real life and the value that it provides us.
Uh let me go back and make sure AI chat is working. Are you here?
When I'm in a sandbox, I Oh, Strawberry.
Okay, that's nice. Great. Fantastic.
It's got a little personality, I guess.
Okay.
Uh let's see. Preview. Okay, so we got Oh, it it used a coffee shop. Okay, perfect. All right. It's got rando uh addresses. I guess we're in Florida here. Okay. Uh Coastal Brew Coffee. What a name. What a name. All right. So, let's go to downloads.
And we have this table CSV. All right.
So, let's go ahead um and we can drag this in. So, downloads table CSV. Okay.
The client provided me this file of their locations.
I want you to create a post type for locations. Um, put the posts in and create the custom fields from the columns. Okay.
Name. I'll do name. N name will be the post title. And um, make sure you map the data to the custom fields as well.
Okay. I'm not really going to give it any other prompting instructions. We're just going to see what it's able to do with that.
Um, and we can answer some questions while it uh processes this. So, will there be a migration path to easily transition from ACF metabox jet engine?
Um, I mean, yeah, that would be nice, right? I I the thing is is I wouldn't say migrating is a priority and and I you have to weigh the value of migration. Uh now let's say you have an ACF subscription and you're like I'm trying to get rid of my ACF subscription. I want to I want to be even at that point like how done is your project? You know, if your project's kind of done, you could just um you know not renew your ACF subscription like the site's still going to work, right? So I think the cases where you would want to migrate data are few and far between.
This is mostly a situation where you're doing a new project. You're going to use etch and not ACF and metabox and and all this other stuff, right? Okay. So it says it created post type locations and a custom field group of locations. And it says it created these custom fields.
And it says it imported 10 posts and it mapped the name to the post title. Okay.
So, I'm going to go over here, and I do see locations, and I do see all of the data in here. Okay, I'm going to go ahead and hit save. I don't think you actually do need to hit save. Uh, but I'm going to switch over to one of these. And I'm going to hit the WordPress icon so we can go back. And we do see, look at this. We have custom fields. We have data in the custom fields. This is insanely helpful, right?
And and you know, we we released this earlier in the week. And you know, somebody made an assumption and randomly made a comment about like, I wish you guys would quit trying to get AI to build sites for us. And like that's not what that's not even what this is. Like if you can't recognize the insane value and like I didn't have to create the post type. I didn't have to create the custom fields one by one. I didn't have to go do a bunch of data entry. Like I could just take a CSV file of data and throw it in and suddenly I have architecture and suddenly I have the data in the system which you Yeah. Keep in mind like WordPress doesn't even have bulk data entry, right? So, it's it's really annoying to have to uh to do all of that manually. I'm going to go back to sample page is where I was, I think.
And notice that the chat, by the way, uh persists across um different pages and posts or wherever you go inside of Etch, which is very nice. So, what I'm going to do is I'm going to say that's fantastic. Uh, I need to show these uh locations in a section with a grid.
Okay. Um, I need you to make a loop definition for that CPT and then or I'll just say for these posts for for these posts and build the section and grid for me. All right. So now, if you are the kind of user who is like, which is a lot of users these days, yeah, I I want it to build for me. I want it to build for me, but I still want to be able to interact with what it builds, right? Um, that's what we're going to take a look at doing next. So, I'm going to ask it as a followup. Now that it's got the the architecture and the data, let's use that architecture and that data. Let's have it create a section for us and a grid and some cards and and uh all of that. Let's see what it comes up with.
Okay. Uh, why is there strawberries in the AI? Well, that's uh it's it's the LLM trying to act more human or something. I don't know.
Okay. So, I created the loop definition as locations grid. So, if I just take a trip over to the loop manager, it just created this right here. Okay. Um, which is nice because, you know, one of the things when we decided to not design a super pretty step-by-step UI for configuring loops, one of the push backs was, well, I'm not I'm not that technical, right? Even though Etch is a technical uh builder, right? I'm I'm not that technical. I don't I don't like doing that. I'm not comfortable with that. And I kind of told people at the time, I was like, I don't think you're going to have to like uh and this is what I meant, right? AI can very very easily whip up stuff like this. Um so it has access access to our loop manager it created the loop and then it also generated not just the uh DOM structure right here the section and it's keep in mind guys this is trained you know it's not it's it's an anti-slop workflow it is using BIM for example it is using the structure that that we have taught it to use it is going to be clean okay um the CSS leverages locally scoped variable variables for things to make things easy to edit and change. Um, as you can see the BIM structure here, it will use nesting. It will make things automatically responsive. Okay. Uh, and this is getting smarter and smarter and smarter and better and better as we go.
And then keep in mind, we add the automatic CSS skill instead of using random CSS down here and uh, locally scope variables where not completely necessary. it'll be using more of ACSS's tokens and workflows and utility classes and and things like that. But let's go ahead and see what it generated. So, I'll hit insert on the DOM and I'll hit insert on the CSS. And you can see that it has built a section and it has added some cards. And if we go in here, uh, one thing that we're working on with it right now is it's not using uh it's trying it's it's assuming we're using ACF still um instead of knowing that we're using etch for this. And so what I'm going to say is I'm going to say uh you used ACF in the dynamic data instead of etch. So my dynamic data isn't working right. And I will uh send that back. I don't know why. Maybe we should look at why uh I thought it I thought it answered strawberries earlier there, Matteo, but it it appears to very much like strawberries these days uh in our chat. So, uh maybe if Nuno's watching, he can jump in and figure out why it >> I don't know if the team did something >> why it keeps printing strawberries.
[laughter] >> Maybe some I am on the >> strawberries.
>> Yeah, I am on the uh main branch and not like this is an unreleased version. So there maybe some sort of test required the printing of strawberries. I don't know. Um so it says you're right for edge native custom fields inside the the name state should be item.ed. So it has given this it has refed it back to me and this is kind of good because it lets us explore other features right uh one being replace. So it has identified that this section already exists. It's not going to insert it again like as a sibling or something. It allows you to simply replace. Um, now I don't know if the replace actually worked. Um, so what we'll do is we'll just kill that and we'll just reinsert it. And there we go.
So now we have the data actually working, which is fantastic. U, the next step would be if you wanted to show an uh, images for these things. The way that this would work is you would go to media and this is the old media manager in WordPress. You would upload all of the photos for the locations. And if you make sure that the file name matches the name of the location, AI will be able to go in and grab those images and set them as featured images on the locations for you. And it does that with uh URL matching. Okay. So, if as long as you file name them properly, it would be able to go in and bulk do that. I obviously don't have photos for all these locations so I'm not going to do that but it is capable of doing that for you and and this is just very very very helpful stuff. Um I [clears throat] know that Chris is working on just to give you a glimpse of maybe the next step of this. You don't see any components right now. Uh ideally these cards would be a component. Now there's only one card because it is in a loop. Okay. So we see it it did use the loop properly and it's looping through one card.
This would ideally be a component. And so the question would be, can etch make that a component for me? Does it realize, hey, this is a card? A card in this fashion should be a component.
Would it be able to make it a component?
Would it be able to create properties for all of these things and then put the dynamic data keys into the props and and map the props to the areas in the card that it's supposed to go to? Could it do all of that? And the answer is yes. It will be able to do that. It will be able to do that. Um, based [clears throat] on what Chris is working on, we'll make it easy for an AI skill to create a component. Uh, and then I think everything else from there is stuff that we've already accomplished for the most part. So, um, yeah, I I think that's going to be another exciting step that is coming very soon. But the ability to create custom post types, create custom fields, map the custom fields to that post type, put in the data, not just for the post, but for the custom fields, link the featured image, create the loop definition, build the section using the loop, and then putting the dynamic data keys in. This is all very, very, very promising and exciting stuff and is insanely efficient to a realworld workflow. So, when you're like, I I'm trying to save time on my projects. Time is money. Uh you're not wanting to cut dev corners, but you are wanting to stop working like a chump in a lot of other ways. This is really really really going to be helpful.
Um let's go to the chat. See?
>> All right. We have the we have solved the mystery of the strawberries >> experiment running. I I see that >> it is it is an experiment uh on uh on on the uh ACSS training.
Yes. Uh, let's see. [laughter] Um, >> team forgot to feature flag it, so it's leaking.
>> Yeah, no worries. No worries. No worries. I'm just catching up on on chat as I'm talking and such. I don't get to also also read the chat.
>> Yeah. Yeah.
>> Uh, >> yeah. We also have some questions about like when is the ACSS uh training going to be uh ready, which is is going to be soon. Uh like we started doing it.
>> Yeah. Um, it's going to be soon. We don't have a a timeline just yet, but it's going to be soon.
>> And I and I said on Tuesday's stream, like you'll encounter it in pieces, right? It's not something where we're going to wait until it's perfect and done before you actually start seeing output related to it. um more more probably is the case that uh very very soon you start to see it doing some small things and then you see it doing a little bit more and a little bit more and a little bit and just kind of getting smarter and more capable over time uh as we feed it more context and more rules and things like that.
>> Yeah, exactly. As per our philosophy of small incremental changes uh that you can benefit as soon as possible.
>> Yeah. Um now also keep in mind uh that you know I did not feed it a design for this. So uh it came up with kind of some AI slopage in terms of the design. If you have missed past streams you are not aware but you can literally go to like I I could go to uh I could go to landbook for example right and I could sort through here and look through here and I could go find some other site that did locations really awesome. Okay. Um, and so if you come in here, like let's look at uh this is not this is not locations.
Uh, okay. So here's a good example. Why is this like it feels like faded? Maybe that's just the design. Um, I'll just take this. So you can come in here and you can feed it screenshots, right? So you can just say, okay, now build this testimonial section, right? Uh, boom.
Now, I don't I didn't tell it to create a CPT for testimonials and all this other stuff. Do I even have team? Okay, we don't we don't even have one. Um, I wonder if it can, you know, it probably put its own data in, can it?
Like, if I just said, hey, throw some dummy data in. I don't know if it can do that or not. Um, I wonder if we should try to get it to do that just as a Let's just do as experiment. Okay, sorry I got ahead of myself. Okay, create a testimonials CPT.
Load it up with five sample posts. Okay.
Uh the testimonial can it can it inject content into the into the block editor yet? Like the body of the the post. Do we know?
>> Yeah, I think so. I I think when it creates post, it can also create the post body.
Um team, feel free to correct me if I'm wrong.
>> All right. Sorry, I got ahead of myself.
I've created a testimonial CPT. Load it up with five sample posts. The testimonial can just be the post body.
Um the person's person's name should be the title.
Um we'll say six. I'll say six. Let's keep them even numbers. Okay. And then um and then uh create the loop definition for that and reenerate the the DOM structure based on uh using a loop. Okay. I don't know.
There's I don't there's no guarantee that this will work. There is a 100% guarantee that we see strawberries though. That is uh that is what we can guarantee.
Uh how much would it cost to upload a picture of a section then have Edge AI create it with all the CPTs inserting pictures creating component? Well, that's kind of what we're doing right now. That's kind of a little bit of what we're doing right now. Um the we we have talked about not feeding the images directly to the AI and then getting them back, right? Um >> yeah. Yeah, we talked about various ways like there's been some internal discussion on how to properly architect this so that it doesn't do a lot of back and forth with the LLM which might be expensive. Um, on the topic of of of prices, we also have on our radar the idea of using the codeex subscription instead of the API. It's on our radar.
We're not working on it yet, but it's it's on our radar.
All right. So, let's insert this. Let's insert this. Uh, and so there you go.
Now, there's a little something going on right here with the uh uh the avatar, I guess, cuz we don't have any photos in here. Um, but it Let's go back to Yeah, I mean, I don't like the section. I don't like the colors and such. It's It's kind of ugly. Uh, but it did it does appear to mostly uh built it. And if we look in here, it did use a loop and it did use one single testimonial card and it is correctly pulling in dynamic data. Let me go back and save this. I didn't save it. Okay, let's go to the front end. And there we go. Right there. So this only thing here would be fixed by convert to raw HTML I believe.
See if that fixes that. Yep. Uh all good. Okay. So we have to investigate the headsh shot and get that situation in. Um but check check what it did. So I now have a testimonial CPT. I now have testimonials in here. Even this with dummy data. Have you guys ever been working on like your about page and you're about to build the team grid and you're like, "Ah, I don't I don't have a CPT for that." Okay, stop what I'm doing. Go create a CPT. Go create some custom fields. Have to put in like three dummy data entry. Doing all this manually. Now, you could just be like, "Hey, whip up that CPT. Put some dummy data in. Build a loop definition for me.
Drop a loop in here. Get me started with a grid. And then I'll take it from there." And then it just does that in like 25 seconds, you know? So helpful.
Insanely helpful. Um and and this is what this is what you get, right? So questions on that. We'll open up for questions.
Um let's see. Uh okay, >> Justin says vapor wear.
>> Yeah, it gets more vaporary every single week, right?
>> Yeah, so much vapor.
>> Um yeah, I wonder why it made that heading black.
It it it tends to do. By the way, this is um mobile responsive if anybody was was curious.
Uh [laughter] I just I I guess it didn't add a color.
Uh that's actually just white.
There you go. But this is the real benefit of doing this stuff in etch is that um if you were doing this in uh VS code or cursor or claude code or whatever you end up having to just prompt for the rest of your life right whereas in etch it generates you can go in you can click you can interact you can change you can adjust you you could do the stuff you know how to do you're just not doing 80% of the stupid grunt work anymore. Um you're letting AI handle that aspect of it. Then you're going in and you're able to do the fine-tuning and the detail work and the move this around, move that around.
Okay, I want to use uh you know this color here, that color here. It's all class-based, by the way. It's it's Did you notice like if I can zoom in up here, it when it generates things, it's not it's not slop like it's it's labeled for you. It's not just nonsense divs over here. It's got a section called testimonials. If I look at this, it's using the class testimonials, right? Um, you go in here and you're seeing testimonials header, testimonials, title. Okay, it's using BIM. Uh, it's it's organized. It's doing the things you would expect it to do. Here's the testimonials grid. Okay, so can it get better? Like my preference is, let me go back to that grid again.
Um, I don't like grid to be the child, like the element in BIM. I like it to be testimonials dash grid. I like it to be its own uh block, right? Um, so these are things that we can continue to train it on and and refine. Uh, but this is this is still good, you know. Uh, and by the way, you can just testimonials grid. Uh, you can just rename it. Like if you don't like it, it's not like you have to do everything over again. You can literally just rename it and nothing breaks and now you're good.
This is testimonial card which is its own block which is nice. Okay, so like you know 98% hit rate for kind of getting it to to what you would expect.
Uh Stripe Goat is in the house. What's up Stripe Goat?
Can you already ask the AI to check the whole DOM for accessibility? Uh, well, I don't I don't know what you mean. Like, um, a lot of accessibility comes from CSS, not the actual markup. Um, a lot of accessibility comes from JavaScript as well. So, you know, can it do an entire accessibility check on the entire page?
No. Is that a goal in the future? Yes.
Uh, because that's another tool you shouldn't need, right? We have we have browser related tools for that. There are plug-in related tools for that and that should just kind of be part of a unified development workflow, right?
Just like SEO should be part of a unified development workflow. The idea that you're not going to need an SEO plugin in the future. Like yes, we can handle the title tags and the metad descriptions. We can handle the schema.
We can handle the sitemap generation. We can we can handle all the stuff you would uh expect to see from an SEO plugin without worrying about all the bells and whistles that actually don't need to exist. By the way, you know, I don't think everybody needs green and red little stop lightss everywhere across their their UI uh to know that they optimized their their their page properly. Um, and most of that stuff is nonsense BS anyway. Like we're still counting how many times the keyword was used in the body text. Like, okay, that's that's like 2007 workflow, you know? We don't we don't need that.
Uh but as far as clean code goes, it's already trained on generating that, right? Uh notice here it did not really generate any wrappers that were not necessary, but that's outside of that's that's more of just clean code. That's not really an accessibility thing.
Um it I believe it will use in certain cases like Arya uh properly and we we have to check it's again it's very scenario dependent on what's actually necessary.
All right. Um, what what are we how are we doing on time, by the way? Okay, we still got some time.
Um, let me switch over to hashtag. Let me make sure there's no demo request.
There are not. If you want a demo done, you need to get it in right now. Okay, we're running out of time. Hashdemo on a demo request. I'm going to go through hashtag Q here.
Will the dynamic image element work properly with the fallback image using source set and alt alt text? Um I don't know what you mean by fallback image. Uh currently there's no way to set a fallback. Uh you know it expects that if you added an image to the page that you are selecting an image.
I mean there is a placeholder like if I just do this um you know there's there's this but yeah you you you need to put in an image you know >> uh let's see I think we're pretty much caught up with the questions. There was just one at the very beginning from Zach asking if at some point we could dedicate a sprint to like quality of life and small updates to which I would say no. We probably will just squeeze small quality of life small updates into bigger things.
>> Yeah.
>> Rather than just gen dedicating an entire thing to just clean up.
>> I mean you you can feel free to send me the list >> uh so I can take a look at it.
>> Yeah. We'll we'll take a look at the list. Uh it's it's also probably going to be matching some of the tickets that we already have. And um like wherever we can squeeze some things, we we squeeze them. But we always go through prioritization from a perspective of what makes the most impact. Right now, what is the biggest roadblock that people using the app are are hitting rather than just the small quality of life stuff. But if they're small enough, they can be they can squeeze in uh into like a a bigger um a week.
>> Yeah. Yeah. Yeah. Um so just to clarify this on the static side of things, so I think this was updated recently. Um this used to be an eitheror situation like you can use a WordPress media which comes with source set information. Um and so this is obviously the preferred way. It's it's done by pulling the media ID. Uh then you could also let's say you needed to call an asset that was hosted statically or outside of WordPress. Um you obviously can't reference that by ID. So this field was for a static URL.
And it used to be that you you you put in either or like once you put in this you can see it disables this. Uh but if you set this before you set this it does apparently act as a fallback. I just don't I I don't know many situations where that's actually required or necessary or it's just kind of like a we we could do it so I guess we did it. But I I don't know what the super practical use case for that is. Um I guess if your if your media ever uh disappears from your data your database, it's you still have a a static reference, but if that's referencing the same file or a file that also disappears, it's not going to do you any good. So I don't know. Um, I guess there are like some statically hosted custom avatar type things you could potentially use in some cases.
The the the use case for it is very limited, I would say. But it does it does technically work.
>> [snorts] >> We have a demo request, but I don't think we can fulfill it because he's asking how to get this build a full from the No, he's asking how to get from the native WP media into the new etch asset library. But the problem is that the new asset library right now is just in static. It's not in the WordPress plug-in yet. So, I don't I I don't know that we can fulfill this.
>> Yeah. Yeah. Yeah.
It will be coming to WordPress relatively soon.
>> Um, source set seems to destroy image quality. No, it doesn't. I mean, you're looking at a very poor quality photo uh in the first place, I believe.
So, if we did um see one of these that I just did, which one would be able to look the best?
Oops. I don't want to rename it. I just want to upload it. So, this is 1600 uh pixels. Yeah, it looks fine.
It was just a That was a very bad photo.
It was from like a dealership that clearly, you know, the room was too dark when they took the photo and they took it with a shitty point andoot camera probably.
Wasn't it going to be etch cruise? Uh oh. Were we going to do a a live event?
>> Edge Camp.
>> Yeah, Edge Camp.
>> Doug asked Etch Camp.
>> Yeah. Second use case is for user avatars if they don't have an image than a placeholder. Yeah. Yeah. Yeah. I mean, you you could definitely do something like that.
Um, I mean, I think when you're doing a loop and a post doesn't have an image, that's not really where I would use a fallback.
um you would want to do a more intelligent card architecture where if a post doesn't have an image, you're changing the layout or you're doing something else or uh you're you're ordering all the cards that don't have images to somewhere else in the grid so that they're all grouped together or like there's a bunch of different ways to handle that. Like a placeholder where it's like, "Sorry, this one doesn't have an image." And like that's a little it's a little awkward for the user.
I think there's better ways to handle that with logic and such.
Um let's see.
Please please pretty please roll edge static into our LTDs even if it's quasi limited version. I mean the the challenge is um it it has direct costs to us like it's not a like a plugin is you you build a plugin you bundle it you release it somebody inst you like a million people could install the plugin and it it doesn't matter right um when you have a hosted solution when you have like a system that people have to log and store stuff in like there there's direct costs to that so I mean we could just be like, "Yeah, I guess it's just it's free and we'll just go bankrupt."
Like I that's doesn't seem like that never seems like a good solution for anybody. So >> yeah, >> you know, like that we got to figure it out. We got we got to figure it out.
>> Yeah, exactly. It is on our radar and and as I said in a comment, whatever we end up doing, we will try to be as fair as possible to anybody who has an LTD.
Um but you also have to understand that this is to some extent a a separate product. Um just from a technical perspective it it it shares a lot of the codebase. So it helps move things fast from one side to the other but it has some challenges that that in a way you didn't buy with the first with the first LTD. So we don't know but we'll try to be as fair as possible. this this um this is what we can promise other than that it's details that we have to figure out. Yeah, I mean you you like you can imagine, right? So for example with custom post types, well that's like a registration in WordPress uh like that this is there's underlying architecture that WordPress provides that when you try to do custom content types and static you you're you're responsible for creating that architecture. Um same thing with custom fields, same thing with uh media management, right? Um we're not just tapping into WordPress's media storage and media library native stuff. Uh we're you're having to build all that from scratch. Um but it's it's important to do this uh because a lot of projects are going in that direction that a lot of the market is going in that direction. Um a lot of clients don't need it. A lot of people that use etch tell us like you know I've this is what we we continue to hear all the time whether you like it or not right like 50% of my projects I no longer do in WordPress anymore. Uh but they also tell us like I wish I could still do those in etch you know uh the these are users the these are investors these are people who uh bought into etch and want to use etch um there's a lot of that right uh then you look at other things like the health of WordPress the outlook of WordPress the pace of WordPress development and its ability to try to keep up with what's changing in the marketplace and a lot of those signals point to hey like Maybe maybe you shouldn't be putting all your eggs in that one basket. Maybe that's not the smartest decision. Um so all of these things have to be weighed and all of these things have to be looked at.
>> Yeah. And to the people who are saying, well, can we have a self-hosted version?
Can we have a version that runs on our desktop? Um it is also something that is on our mind, but it's not a an immediate we're not planning on doing it at least initially. And one of the reasons is the immense amount of different potential uh scenarios for support just like with the WordPress plug-in like you can host this on any uh any hosting platform and you get um and you get a lot of different challenges with different WordPress uh hosting platforms uh including the ones that you roll out yourselves um that also has you know it's it's a lot of challenges is so to the very least the first version is going to be on a on an environment that we control. So our support is also focused on one environment that we know very well and that we optimize for for speed for reliability etc etc. >> Is there ever going to be a self-hosted version or desktop version? I I don't know. Maybe maybe not immediately but in the future like never say never.
>> Yeah. I mean you guys have to think about it from a um a smart intelligent development perspective, right? Do you go down the the path of there's a million variables or do you go down the path of there's five variables? Well, you should probably go down the path where there's five variables and really uh nail it, knock it out of the park and make it great. And then you can use that to go, okay, now we want to tackle an environment with 100 variables, right?
Um, there's no sense in getting yourself into a quagmire of uh complexity right out of the gate for no reason. Um, somebody said, uh, I thought I thought you said Cloudflare hosting was free.
Um, it is in a lot of ways. It is in a lot of ways. The costs I was talking about were not related to the websites you guys will be building and hosting.
Um it's the cost to us for running the infrastructure. Okay. Um that's a different story.
>> Yeah, exactly. You have to separate how you host with how we host the uh the application. And how we host the application, especially if it scales to like thousands of websites. Uh it's thousands of websites that have to save their data. They have to save their assets on this infrastructure that we run. That's going to that's going to become expensive soon.
>> Yeah. Um let's see. Were there any? Let me double check. Okay, no demo requests.
So, that's good. Let me go back down to the end of the chat.
Um, primary support in terms of the Etch for how many years going forward are you going to guarantee to primarily support in terms of active development the Ech?
Um, I don't know. There seems to be some sort of connotation to this question.
Um, we're not ending development of uh HWP.
Uh, we're not leaving WordPress. We're not uh what you have to understand is this is an expansion. Okay, it's an expansion outside of WordPress. Uh, it's as Matteo said, it shares the vast majority of the codebase, which means that things can easily be like we're doing with the media, the asset manager, right? just build it for for static and then go okay port it to the WordPress version so that it gets it as well.
Yeah.
>> Um it's very easy to do that >> them together so that we when you make changes on one side the other side gets them too.
>> Yeah.
>> Which is what will make this a lot more maintainable. Like >> the these are separate products because they run on different infrastructure and everything else. They have different challenges but the builder is largely the same thing in in irregardless of what you run it on top of.
>> And so we are structuring it in such a way that helps us make everything be available on multiple platforms at the same time.
>> That being said and and so with this we are not and like we're not getting out of WordPress. this we are still supporting WordPress and we're still planning on on making uh WordPress better through Etch as you'll see next week when we release uh the third party API when we release potentially the asset manager or parts of it and then we start you know we start deploying some of the things that we've started developing for static uh soon.
>> Yeah.
>> And keep in mind that Cloudflare is not the sole destination it's the destination right now.
>> Yeah. Uh, this gets much more interesting when you can push a website to a different CMS other than WordPress, right? Um, this gets more interesting when you can potentially uh build a Shopify site in Etch because you're able to connect it to uh the Shopify architecture for storing the data. Um, this opens the door to a lot a lot a lot of different pathways that are very very valuable. Um the idea if you think about solving like what is the problem this is solving the idea that if I want to use WordPress I use Etch but the minute and remember I'm an agency okay so the minute I I have a WordPress project in front of me I I'm going to use Etch now I have a site where they don't need WordPress they want more of a brochure site they want it statically hosted it's they want it to be super cheap you know like free free to host no maintenance no management no security implications they don't Nobody needs to log in that blah blah blah blah blah. Okay, they don't need WordPress. So, I need to what? Just use a completely different workflow.
Why can't I just still do that project in Etch? And then maybe someone comes along and they're like, I want to I want to do a a custom Shopify site. And this is in the future. This is not now, but in in the future. Now, I I got I got to go, you know, do that in a completely different workflow, in a completely different development environment, a completely different everything. And then how am I going to maintain all these projects, right?
Imagine that they're just all just done through Etch. It's just all done through Etch with the benefit of all. Even if you're using LLMs to really help you do 80% of the work, you still have the visual development aspect of Etch. You still have the selector system. You still have the loops. You still have the logic. You still have the components.
You still have all of the stuff uh that you know and love. And the destination of where that site lives doesn't have to really matter all that much.
>> Yes, exactly. The you you're no you will no longer be limited when we go down this route when we expand more into into that offering. You will no longer be limited to uh the the platform that you the only platform that you knew and the only platform that we integrated on. But sometimes there are better alternatives and and especially with AI making the choice of switching easier uh and making learning easier. Uh why wouldn't you?
But you keep the constant of okay whenever I want to touch the website it's always on the same interface that I know and that I've I've grown accustomed to to working with.
>> Yeah. Um let's see.
I think the next step is a web flow type SAS for building fully in etch. We I we well at least I I I wanted to avoid that, right? Like one of the conversations we had was and this is why at static works the way that it does. So [clears throat] just to clarify, let me go back to uh is there any way to get back in without or do I have to go through the same thing again? Okay, good. It remembers me. All right. Uh, let me go to Does it obfuscate the keys and settings? Do you know?
>> Uh, it should, but but just to be safe, >> let me go over here for a second.
>> Just to be safe, >> let me go over here.
>> I think it does. I I think it it it uh it shows >> it shows the account ID >> platform account ID, which I don't know how big of a deal it is.
>> Okay.
Um, I mean we could ask Cloudflare account ID is made public.
Any security implications?
>> Let's just ask it in the chat where it was generating CSV files. I'm sure that's a helpful context.
>> Yeah. Uh, Multilife says to check the rest of the messages for full context. Let me try and track them down.
Um, >> he said the team did a smart move to test the builder first with the large community WordPress has and the not so great infrastructure. If that works, then build your own infrastructure. And then message number two was I think the next step is a web flow type SAS for building fully in edge.
>> Yeah.
>> Um [clears throat] >> I think maybe the point that you were trying to make also is the like the the builder is in one place and it's a SAS where you host the website. It's your own infrastructure.
>> Yeah. So this is >> Yeah.
>> Yeah. This is your Cloudflare account information. your Cloudflare API token, your Cloudflare account ID, and then your worker name is the name of your website, okay? Or the client's website or whatever, right? And so with Web Flow, when you sign up for Web Flow, you're paying for their builder, but you're also paying to host the website with them. They're they're a fancy hosting company is what they are. Okay.
Um, and so with this, this is completely different. You are creating an account with Etch to build the website, but you're deploying it to infrastructure you own in control and you stay in control and ownership of that website, right? Um, so all of the files are yours. All of the data is yours. uh Web Flow could wake up in the morning and increase the price on you for that website to like if you don't want to pay that price anymore, you've got to find a way to get that website somewhere else, right? Um they could wake up and be like, you know, you guys are you're doing some you're selling some products we don't like and so we're just going to shut you down. They could do that. We can't do that to you. Okay, this is deployed to your own infrastructure that you control. Um, and right now it's Cloudflare. It could be, like I said, Shopify. It could be um another CMS, whatever, whatever you want it to be.
Okay. So, that that is the advantage here. And I I didn't want to build another like a web flow type thing. I I don't want to own and control people's data essentially. Uh I I want them to.
And that's kind of the the whole WordPress philosophy as well. like, you know, it's open source, you own your stuff, you self-host it, whatever. Um, so this is kind of a happy medium, I would say. Yeah, exactly. And at some point, you also look at like what are the chances that what are the chances that we could build a better a better builder unified development environment? Very high because we worked on the space for a lot a long time. We have a good team. We have a good knowledge of of the whole thing. What are the chances that we can build a better hosting platform for your websites than the thousands of existing uh solutions out there? Probably not as high. Should we spend time and money trying to compete in that space? It doesn't it doesn't make much sense. So, we'll just rely on that. Which also gives you the advantage of you still own the entirety of the deployed website.
It's wherever you want it. uh it as I mentioned right now you can deploy directly to Cloudflare but you can also export the files and upload them wherever you want. So you have a platform that is not directly supported you can still upload manually and eventually more and more platforms will come.
>> Yeah. Uh two more good questions just came in. This one's very interesting. In the future, is it potentially possible to have the WP version where you can build visually and then push to Cloudflare as static? Um, I would think that it would happen actually the other way around, but it would be very interesting where you start out static because the idea is remember WordPress is very large infrastructure. It's heavy, it's slow, it's like for the most part um it's uh maintenanceheavy for sure. It's securityheavy for sure. Okay.
It's hosting heavy for sure. Um, so the idea would be minimum viable approach.
So if you have a client where it's like I don't know like static seems real good for them, but maybe we run into a scenario at some point where we find like we should have gone WordPress.
Well, I think that's what's really interesting about this. you could literally start almost every client on static and push to WordPress if if it's decided that that's needed, right? Or migrate it to WordPress uh without having to rebuild everything and rearchitect everything. I think that that would be uh and almost do it in like a one-click kind of fashion be like ah we decided we need this to be on WordPress. Okay, no problem. because etch is for WordPress also and all the language is the same. We're essentially like send it send it over here. Uh and and you're able to do that. I I think that that's a uh an interesting future.
>> I don't know if you kind of see it the same way, >> Matteo.
>> No, no, I I see it that way. It's much easier. It's always easier to start small and then and then and then increase the range when you see that there is a need for it. And so with these websites, it's much easier to start with a small brochure static site and then when you realize that you need uh that you have needs for stuff that is in WordPress and it's not on static, then you can move there. That flow is very much on our mind and we'll probably at some point when when the time is right and the infrastructure is ready, we'll try to figure out a way to facilitate that transition. Um the other way around I do see it as more unlikely.
Um but again it as usual if somebody if if if enough people can bring the kind of uh use cases that make this obviously a something that we should build then we're going to take a look at that.
>> Yeah.
Uh Leeben says [clears throat] uh is there anything a competitor has done in the last few months that concerns you a good or bad way? positive development or negative development?
[snorts] Um, not really in this space.
Um, like the thing that concerns me the most right now is um the explosion of vibe coding stuff for example. So like I see the biggest threats as a cursor or a clawed code or something like that versus uh anything else any other like page builder is is trying to do. almost all of them are wildly behind and you know doing fairly boring stuff. Um claude code when when you get you know 50 emails where it's like I'm just doing everything in cursor now I'm just doing everything in in cloud and and AI does everything for me and blah blah blah blah blah. So, I started to heavily uh test that, actually do that workflow with like some real projects that I had in front of me where I was like, "Okay, I'm going to go down that path and see what everybody's like talking about."
And one of the big things I realized was that uh it's tremendously good at getting you 80 to 90% of the way. The last 10% is insanely frustrating. And what I continued to find, the situation I found myself in over and over again was I wish I had the etch environment, like the visual development environment where I can now just it it's produced some things. I'd love to click on it and and and do this kind of stuff that we've always done, right? See the isolated uh DOM structure for that thing. See the isolated styling for that thing. Uh have it done in my language. Have it done sensibly. have it done with no slop. Uh without having to define large sets of rules, um the loops that I'm used to, the um media management that I'm used to. I I don't like only having to look at code and trying to decipher what it did, right? I want to see it visually. I want to click on it. I want to isolate to it. I want to then go make the fine-tuning adjustments that I want to do without having to reprompt, reprompt, reprompt, reprompt, reprompt, repromp.
Uh, no. You're still not getting what I'm saying. Please keep trying again.
Try again. Try Oh, no. We're out of the context. So, we need a new context window now. We got to Hey, can you write me a prompt so I can start fresh in a new context? Oh, man. That is not a game I enjoy. That is not a game I want to play. I want to use it to its capability to get me 80 to 90% of the way. And then I want to go in and fine-tune everything myself. Okay. Um, and you know, like having the loop manager, having components, okay, where it's like, man, did you make that a component? Well, yeah, I did it in spelt and well, I I I mean, I'm not a I'm I'm familiar with spelt, but I'm not a spelt guy, you know? So, the idea that I can go in and fine-tune a component that it made, I that's not the life I want to live, you know, but I can do that in Etch. I can easily go in and manipulate components in Etch and build libraries of components in Etch and do all the other stuff that I'm used to doing even though I have used AI perhaps to do 80 to 90% of the grunt work for that project. Uh and maybe um I I hired a designer who did a custom design and I'm like that's fantastic. The idea that I would sit here and and actually just literally build that from scratch doesn't seem like the best use of my time when I can screenshot it, drop it into Etch and Etch does a pretty damn good job at it, and then I can fine-tune it to its finish point. That's kind of the workflow that that is like the modern workflow we're trying to accomplish that cursor and claude code don't give you because they don't have a UI, right? So, you're stuck. Do you know the language that it used really really really well?
Do you want to wade through code? Do you want to uh manipulate CSS that it wrote and didn't write very well or decide to use I I used Tailwind on one of the projects because I was like, you know, I'm just going to I'm going to bite the bullet. You know, everybody says it's magical. Everybody says it's amazing and uh so I'm I'm just going to do it and if it sucks, I'm going to live with it.
Well, guess what? It sucks and I got to live with it. uh until I can tell it to refactor the entire project, which I don't I don't really have the heart to do. Uh but I I can confirm, guys, it's it's Tailwind's a disaster. It's a nightmare. Um if people that learned Tailwind and maybe the beginning and they're experts at Tailwind and they're super comfortable with it, um they probably still like it and and obviously still use it. There's no reason.
There's no reason for it anymore, okay?
uh in 2026 especially and even if you're using LLMs there there's no reason for it. It has a lot of cons and not a lot of pros. Okay. Um so yeah, I I've I've done I've gone down that rabbit hole. Uh there's a bunch of dead rabbits at the end of it. It smells bad. I'm coming back out. Okay. Um so check that one off.
Uh let's see.
What are your thoughts on unblock copying your product? I mean it it is it's to be expected I guess but it's still not they just none of these none of these copycats have uh any vision you know so it doesn't in the in the long run it doesn't matter. Yeah, that's that's a problem. Like taking inspiration from another product, but you have your own vision and there's this one thing that you think would fit well in your vision. It makes sense.
>> Starting from another product and you know >> it is what it is.
>> Yeah. Any thoughts on Oh my Etch's release of Woo components yesterday? Um I love it. I mean more more power more power to Waji and uh you know it's solutions for the community that we haven't been able to get to yet. So uh I think it's fantastic.
>> Absolutely. It's why we're investing on third party API so that these third party tools are also more stable and more um and can even develop faster. Um we love to see it and we hope for more of that.
>> Yeah. And it's not to say that we're not going to um you know build build the Woo components, right? There will be native Woo components and I we've tried to make this clear to all of the third party developers. I think one of the big problems in the bricks ecosystem was that the waters were very muddy. You know, what will be native and what won't be native and it's almost like a sit back and let's let the third party developers do all the innovation and then we'll just steal it, right?
um which I think created a lot of like a bad taste in people's mouths. Um we've been very clear there will be X Y and Z and if a third party developer goes, you know, I I'm noticing that that's a little further down the priorities list.
I bet I can build it for people and make some sales and, you know, they'll love it, whatever. Um they do that. That's their decision. they already know that that this stuff will come natively.
Okay, we've been very clear about that.
So when it does kind of my commitment and if we look at track record as well, this only really happened one major time uh and it was with automatic CSS, right? So um Chris created u the context menus. I forgot what was the name of it. Do we remember the name?
>> Uh plaster.
>> Plaster. Uh, so Chris created Plaster and Plaster was um very very innovative, right? And people kept messaging me all the time. They're like, "Dude, this is amazing. Have you tried it? Have you tried it? Have you tried it?" And I was like, "No, I haven't tried it." And then I finally tried it. And I was like, "Okay, this is really cool." Like I actually wasn't like in the very beginning, I felt like, who needs that?
You know, once you've used it, you're like, "Man, uh, all right. How did I how did I do things without this before?"
And we had the team. We had the team sitting here and we could have gone, "Guys, that's a great concept right there. Let's just build our version of it almost exactly like like that and just release it and sell it as an add-on, right? We we could do that, too.
Uh we could we could make a bunch of money and we didn't even have to come up with a concept or really how it works or what it looks like. We we'll just take it." Uh we could have done that, right?
Because we had the team sitting there that that was capable of doing that.
What did we do instead? We knocked on Chris's door, right? Not not really in real life physically. Um but we said, "Hey, uh we would like to acquire that."
Like you did a very good job with that.
Not only would we like to acquire that, we would like to acquire you. Uh we would like to bring you on the team u because this is the kind of innovation that we're looking for. And so, you know, he got paid, came on the team, and then we built the tool and and did not release it as an add-on, just put it in as a feature. So, in the track record of how we do this kind of thing, that can be a possibility going forwards that if a third party developer just knocks something out of the park, uh we give it a quick audit and say like, man, this was really well done. It looks really good. um we don't want to compete with you on this necessarily and and we don't want to just delete you because if a thirdparty developer has done a lot of work and they're relying on those sales and we just come in and go, "Okay guys, here's the native version. By the way, it looks like pretty much exactly the same and does all the same stuff that kind of really sucks for that third party developer in a lot of ways, right?
So, we would like to find ways to reward them. Um, and it maybe they've done something where we're like, uh, okay, like it sufficed, but that really wasn't our vision, then then we're going to go build it natively and we're going to do our own thing. I I think what I can kind of promise is you're not going to and you know, we don't want to name names, but like, you know, you've seen this happen. you've seen this happen where um the native version comes out and it looks pretty damn similar to the third party one and it's just like well it sucks for them you know I I don't think you're going to see that like if it's going to look almost identical it's probably cuz we we we made some sort of deal you know um if they decide to provide something and it works for people but it's not our vision we're going to build our vision and it may trample the other thing because it's way better because our vision happened to be way better and our execution have then that's fair. You know, all is fair in third party development. But you're not going to see a replica appear natively and uh us us waving goodbye from the ship, right? Uh as it sails away. I I don't think that that's not a good way to manage third party developers in your ecosystem.
>> Yeah, that's not us. Plus, I mean, in the case of Chris, as you mentioned, not only did we acquire a a very good plug-in at the time and make it available to the community included in in what they were already paying or had paid, we acquired an actual solid developer. He like we we we used the opportunity to enhance our team. And now it's what, four years later, five years later, I've lost track of time. Chris is still with us. He's one of the seniors on our on our team. Yeah, >> he he has a very talented brother, Reinard. He brought him on the team and now we have solid, >> which is crazy. I mean, think think about being in this, you know, Chris was like, "Hey, you know, like I do have a brother and he's he's pretty good at this stuff." And I mean, when you hear that when you hear that initially, you would be like, "Ah, okay." He's just trying to get his brother a job, whatever, you know, gas me up and yada yada yada. I mean, Reinhardt's brilliant, dude. like Ryan just out of nowhere, out of left field, uh he like spawned he like spawned a uh an insanely awesome senior developer like just out of nowhere. Um was like, yeah, that's that's wild. That's wild. Um so yeah, and they've been responsible for so many uh of the pivotal features of like how how Edge works, how automatic CSS works, um you know, behind the scenes very much. Yeah. And and their technical ability was already pretty good, but the way they've also grown in all other categories around it has been has been incredible. Like I'm I'm genuinely proud of of how far they've come from when they started working here.
>> Yeah.
Um okay, I think we're over time, so this may be a good note to end on and uh we'll say goodbye to everybody. We do this every single month. So, at the end of June, sometimes it bleeds into like it's the first week of the next month because um one of the things I found too is if the month ends early in the week or so, like there's something coming out that week that we then don't really get to talk about. If we just wait until the following week, you know, the full month is actually completed and we can actually talk about everything that got released uh that month. So sometimes it's better to do it at the beginning >> just depending on how the dates work.
But we do it every single month. It's your chance to keep up with Etch development. I know that it's a problem these days where especially when we have fast release cycles, we're doing a lot.
We're releasing a lot. What we don't want is we don't want you to ever feel like, "Oh, I haven't been paying attention for two months. I can't open Etch cuz I I I feel like I'm just going to be way behind." Well, watch a live stream or two and you'll be completely caught up. You know, you'll know exactly what's been happening and you'll be able to jump right back in. Okay. So, that's happened to me with with with tools um in the past, right? I and I don't like that feeling. So, we do these live streams so that you guys can ask questions, you can get demos, you can see the new stuff in action, and you can be like up to date on what's going on.
>> Yeah.
>> All right. So, that's it from us.
>> See you guys very soon. Cheers.
Everyone.
Vidéos Similaires
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
Are AI deceiving us? | Roman Yampolsky, Gleb Solomin #AI #science
shortsGlebSolomin
1K views•2026-06-02
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
AI Doesn't Create Bias — It Inherits It
UXEvolved
176 views•2026-06-01
Distributed Inference Challenges Explained #shorts
alexa_griffith
466 views•2026-05-31
[한글자막] OpenAI @ Replay 2026 | OpenAI는 Codex로 개발 방식을 어떻게 바꾸고 있을까요?
TechBridge-KR
1K views•2026-06-03
Starting & Test Driving JAKE'S Abandoned BUS from Subway Surfers | POV Restarting
RestartGaragePOV
4K views•2026-06-04
Building the Future of Voice-First Sovereign AI: Sarvam & NVIDIA
NVIDIA
3K views•2026-06-01
