638: Hop, Hop, Hop
02:14:02 Share

So as mentioned a few times in the past, my kid for his schooling uses Chromebooks like many American children. And everything's on the Chromebook as we discussed a little bit in our school special. Obviously because he is my kid, he is going to pay attention in school to the minimum degree required.

And the rest of the time is goofing off on his Chromebook. Hmm. Who does that sound like? Yeah, right? And it's fine because he gets decent grades and doesn't cause problems in the class. So the teachers don't care. I'm sure the teachers all know he's goofing off on the Chromebook. And they just have bigger fish to fry. So they don't bother him and he doesn't bother them. So anyway, the Chromebooks, though, have various school...

content blocking and filtering software installed. Now, what do middle schoolers want to see on a computer? It's nothing harmful. They're not looking up like bad adult stuff. They just want to like watch YouTube and play games. Like that's it. It's all about YouTube and games. That is like the holy grail is I want to goof off and watch YouTube. I want to play some stupid web game that, you know, I can make this Chromebook less horrible for myself. That's all they want to do.

So my kid has throughout his middle school career so far done all sorts of things to try to evade the blocks to the point where even he was running a spreadsheet that was shared between 40 other kids that would link out to unblocked games. And as they would get blocked, they would update them. And he was...

basically running an open source project like he he was like he was like the moderator and he was hiring like other moderators trading them like lollipops and stuff he at one point was selling ad space in the spreadsheet like it was pretty impressive wait ad space what was being advertised other spreadsheets by other kids

That's incredible. Like a podcast app that runs ads for podcasts on it. Right? Yeah. Can you believe? God, he really is your kid. Oh, yeah. I know. And I told him right from the start. He doesn't do this anymore. He's moved on to other distractions. But right from the start from that one, I told him, like, listen, I know at some point I'm going to get called into the principal's office to talk about this. And I'm like, I just want you to be just promise me two things.

don't let people post bad things about other people, and don't enable cheating. If you don't do those things, I got your back. So fortunately... Oh my God, he was so clever with it. The teachers at the front of the room, they have some kind of overview of the screens of everyone's Chromebooks in the class. He figured this out. He could kind of see how much detail they had. And so he would...

create the layout and give it like a big title on the top that would be something relevant to what they were doing. Like if they were doing a project on dinosaurs, it would be a big title on the top that says Brontosaurus, you know? And then like all the links to the games and everything would all be smaller down below. So like they wouldn't be as visible on the teacher's screen. Too small for old person teachers to see or when they're scaled in the thumbnail, you can't read it anymore. Exactly. Or below the fold because the previews didn't track scroll position. Right.

So anyway, among the other things, I forget what the filtering software is called. It might be GoGuardian or something. There's some filtering software, and it seems like it's mostly like a browser extension that runs on the Chromebook. I don't know anything about what is available for Chromebooks and how this stuff typically works, but...

they've he's found various workarounds to block web urls like just like spam clicking the bookmark but often like eventually it would just work because i think it seems like it's doing javascript client side blocking so like the way he described and he showed me a couple times it's like

you can view the page, it loads, and then it seems like there's an onload handler or something that runs. It's like after it loads, if it's supposed to be banned, the extension kicks you over to a block page or something. So there's all sorts of weird hacks. And the news tonight was he found out how to get an alternative Google login on the Chromebook. Because normally, you can't log in with your own home Google account so you can maybe watch YouTube. No, normally you can't log in. But he found this was so perfect. He found...

he found that if you go to a certain like YouTube page, it tells you you can't log in and there's a link that says learn more. You should put him to work in app review. Right, exactly. That's what I'm thinking of. Anyone who's linking to their alternate payment methods on their website. So that's exactly what he did was what Apple does, which is like he clicked on learn more and just poked around until he found a

a different way to log in with a different login link that was on some header or some footer of some like Google support page deep within God knows what. And he found a different way to log in. And so now he has his, he has two Google logins and he can watch stuff. It's like,

This is, like, I love this stuff. Like, I love, like, you know, when I was a kid, I was mostly just, like, goofing off on my TI-83 and, you know, slowly infecting the school with good assembly games I downloaded off the internet. But, like, now, like, I mean, I love that, like, so much about school has changed.

but this one thing hasn't. This one thing of like, you give kids technology and they will find ways to goof off and play games. And whatever you put in place to try to prevent that, they'll find a way around it.

We should do some follow-up. And I'm not going to name this person, but somebody tweeted, mastodoned us, whatever you want to call it, and said, listening to episode 637 made me realize that I did, in fact, miss the merchandise sale this year. Oops.

Every time. It's always at least one. So I appreciate, I appreciate your honesty, uh, in as thanks for your honesty. I'm not going to link to your toots and I'm not going to name you by name, but you know who you are and yep. Next year, let's, or next sale, let's work a little harder. Uh,

We made passing mention of this last episode, and we should talk about it at least briefly now. Stageplay will be bringing live performances to mixed reality headsets. So this is stageplay.com. I'll put a link in the show notes. Stageplay's performance capture process allows producers, performers, and creators to capture their performances in stereoscopic 3D and distribute them live or on demand.

The StagePlay app, developed for mixed reality headsets like Apple Vision Pro and MetaQuest, enables remote audiences to participate in live entertainment experiences virtually from their homes. At launch, StagePlay will feature a performance from Blue Man Group, which was captured in the show's original theater in New York City earlier this year. More details, including consumer launch and pricing, will be announced later. Let me tell you, I've seen Blue Man Group live a couple of times in New York, and they are phenomenal. Actually, I've seen them once in Charlottesville, too. But anyways, they are phenomenal, put out on a great show. I

I'm really tentatively excited about this. I'm sure the cost to participate in one of these shows will be prohibitive, which is going to make me a lot less excited. But sitting here in blissful ignorance, this sounds great. I don't know. The pricing might not be that bad. I mean, I like that they're launching it not just on Vision Pro, but also on the Quest. So it seems like they have a much broader addressable audience.

On the other hand, people who can afford a Vision Pro can probably also afford more expensive tickets. So we'll see how it goes for them. But we mentioned this so many times, just put a camera in the audience. And we mentioned so many times about how things are sometimes shot like plays. This is straightforward. It's not music concerts, but it's, hey, there's a performance on a stage.

like a Broadway play or show or whatever, and we'll put a camera in the audience, and I hope they just put it in a good position, and that'll be fun to try out. I've also seen Blue Man Group in New York. If I had one, I would definitely try it out to compare the experience. Marco, have you seen Blue Man Group in New York? No, but I did see them in Las Vegas. I actually just saw a play in New York last week. Yeah, you always go to Las Vegas for your big shows like The Sphere or Blue Man Group. Yeah.

Yeah, it was like – I think that's – this is exactly what I hope to see on the Vision Pro. So I hope this pans out. I hope it's good. I hope things are available, and I hope this continues. We'll see. Those are all pretty big conditions. But look, someone's got to do it eventually. So hopefully this pans out. Oh, and I should also mention you and I have –

going back and forth in a, I believe a friendly way about whether or not this was mostly around Metallica was the most recent time that we, we had this argument, but about whether or not just sitting a camera stationary on a stage, is that going to be good or bad? And I,

I don't know if that's what stage play is going to be doing, but I think it stands to reason that that might be what they're doing. Just a static stationary camera that you can use to like look around. And I am super curious and excited to see, is that trash? My, my assertion is that's not going to be that great because you're going to be far away from things you want to see, especially in the context of a, of a band, but perhaps as well in the context of a stage play and your assertion, Marco, and jump in if I'm

misleading anyone. Your assertion was, no, let's stop moving about. Let's just be stationary and give the control, so to speak, to the user to just look around if they want to change their perspective. And I don't know which one of us is right, but this might be the way in which we find out. And maybe I'll be eating my words. Who knows? Well, the thing about stage plays and music concerts is that

they are designed to be viewed by people in stationary seats in the audience. Yeah, that's a good point. They are literally designed for that and optimized for that. So if somebody places a fixed camera in a really good seat,

you know, a few inches up so it's not going to look in the back of someone's head who happens to be tall. Like, you know, just put a camera in a really good seat or above a really good seat and you will see the performance the same way people in the audience are seeing it. But, you know, somebody like that, if you put it in a good enough seat...

That seat might cost $1,000 or whatever. It might be a hundreds of dollars seat. And so you're actually seeing it better than most of the people who are seeing it live and paying $100 or $200 for their tickets. The potential for this to be good, I think, is quite high. If you imagine a fixed camera in many other...

you know, media, it might not be as good. But these productions are specifically designed to be viewed from a fixed position for the entire time. It should theoretically be really good and really fun. And it should be very similar to actually being there. Obviously, you know, you will. The play that I saw last week, I made the mistake of purchasing a drink during intermission. That drink was $40 worth.

Whoa, whoa, whoa, whoa. I'm sorry. A single presumably alcoholic beverage was $40. Four zero. It still feels better than buying the $15 bottle of water.

Holy smokes. $40. Because you don't go to a lot of shows, Casey. Captive audience. I know. I know. So if your other market for seeing a professional play is – and the tickets were a few hundred bucks because we got good seats, so it was a few hundred bucks. So if that's your alternative of that kind of price scale –

you can sell one of these tickets in a vision pro for like i don't know 30 bucks like it might like that might work like who knows i think the business case could be there and it could be really good we won't really know until we get some of these made and we get and we see like how are they shooting it what are the what are the realities of sitting in a fixed camera for two hours in the vision pro like we don't actually know that yet but

But the only way we're going to find that out is by people starting to do it. So I look forward to this. Yeah, me too. Yeah, this is not unprecedented in terms of experience. Like if you think of all the things that we've named, setting aside VR entirely, and as people are pointing out in the chat room, there's been existing headsets with similar content available for them for ages. Just the actual real events. So sitting in a play, like Margo said, you know, you're sitting in a seat. Sometimes your seats are crappy, you don't get a good view, but you can still enjoy the performance. And of course, they probably put the camera in a good seat,

But then there's also things like, I don't know what you would call these, but like when they did the Hamilton thing on Netflix or whatever, it's,

that's not shot as if you're sitting in the audience. Like there are multiple cameras. And if you watch that Hamilton, I don't remember what streaming service on, it was Disney plus Disney plus. Anyway, if you watch that, uh, it has better views of the performance of Hamilton than any person who was sitting in that theater because they have multiple cameras and they move around similarly with music. While there are ones where they have a camera in the audience that are recorded, uh,

Also, there's what is known as concert films, back when you used film or, you know, concert videos or whatever.

that use multiple cameras, multiple performances over multiple nights that might have music overdubbed on them or whatever. And that's an established genre that does not look like you're viewing from the audience. Sporting events. If you watch an NFL game on television today, we have things like amazing camera angles with a flying camera and the thing like just and all the things painted on the field, like views you would never get in any seat in that stadium.

If you go to a football game, you may be up in the nosebleed seats. It looks like a bunch of little ants. And you may be looking at the big jumbotron to see what's going on. And all of those experiences, I think, are...

is potentially applicable to the headset. It just depends on what you want. Some people prefer, they love the NFL, but they prefer to watch it at home on TV because they feel like they get a better view. They want to be on their own couch with their own snacks. They don't want to pay $40 for a beer to Marco's point. But there's also something to be said about the atmosphere being in the stadium. And that is another place where VR can help because

For sporting events, maybe not so much for plays where you're supposed to just be quiet and not be annoying your neighbors, but for sporting events, being able to turn to your left and turn to your right and see your buddies and feel the energy of the crowd is part of being there live. And guess what? Through the magic of networking and VR headsets, that is something that someone could do if it hasn't already been done. Someone in the chat room will tell me if some other headset has already done it.

of quote unquote virtually attending a game with your buddies and you know a million other people in like instant stadiums like using you know technology from MMOs and stuff where you're in a stadium with 30,000 other people and there's a bunch of these stadiums you don't know which one anyway

You could simulate that in a headset and no one of those experiences any worse or better than the other. Like there's something to be said for being in the stadium seats, watching an NFL game. There's something to be said for watching it at home. There's something to be said for being in the theater. There's something to be said for concert movies, etc.

versus sitting in the seats and watching a rock concert. So I think all of this, like there's no, I don't think there's any tension between them. It just depends on what you're in the mood for and what you want. Or like, I don't know anything about F1, but you've talked before about the million ways you can visualize a Formula One race because there's so much going on at the same time. Or golf, someone mentioned that too, is the golf thing where you can follow different people on the course. Like,

Technology makes all of this possible, including things that simulate, you know, being in person at the Masters where you get a stakeout in front of one hole and just hope something dramatic happens versus the concert video version where you can fly around and see everybody. So I think there is only going to be more of this as time goes on once we get the viewing hardware sorted out.

We are sponsored this episode by Squarespace, the all-in-one website platform for entrepreneurs like you to stand out and succeed. Whether you're just starting out or managing a growing brand, Squarespace makes it easy to create a beautiful website for your business, engage with your audience, and sell anything from products to content to time, all in one place and all on your terms. No matter what you sell for your business, maybe it's physical products, digital products,

It could be something like a PDF, e-books. It could be a newsletter. It could be a private podcast. You could even sell time slots if you're like a coach or something. There is so much you can do with Squarespace for your business. And also, honestly, it's all great for personal sites too, but they've been doing that for years. The business stuff, they're really amping up in recent years, and it is just so capable and so good. They now have this incredible design intelligence system. This combines their two decades of industry-leading design expertise with

with cutting-edge AI to unlock your strongest creative potential. You can empower to build a beautiful, more personalized website tailored to your unique needs and craft a bespoke digital identity to use across your entire online presence. And of course, for your business, Squarespace Payments is the easiest way to receive and manage your payments in one place. Onboarding is fast and simple. You get started in just a few clicks. You start receiving payments right away.

Your customers have more ways to pay than ever with popular payment methods like Klarna, ACA Direct Debit, Apple Pay, Afterpay, Clearpay, and whatever else comes along. They're always really fast to support it. You can try it for yourself by building your entire site in trial mode at squarespace.com. See for yourself without even making a purchase how well Squarespace works for you. I strongly encourage you to try this. You might think, oh, maybe it can or can't do this, or maybe it's going to be hard, and you're going to be very surprised.

how much it can do, and how easy it is, even if you're not technical or you're helping somebody who's not technical do it themselves. Go to squarespace.com slash ATP when you're ready to purchase to save 10% off your first purchase of a website or domain. So once again, squarespace.com slash ATP for 10% off. Thank you so much to Squarespace for sponsoring our show.

All right. We have a bunch of follow-up with regard to Epic and Apple. And I'm just going to kind of go through this piece by piece. And please, gentlemen, interrupt when you're ready. Reading from The Verge. Who wrote this article? This is Jacob Kastranakis. In the end...

Apple sought to maintain a revenue stream worth billions in direct defiance of this court's injunction, Gonzalez-Rogers says. She notes that inside Apple App Store chief Phil Schiller advocated for the company to comply with the injunction, but that CEO Tim Cook chose poorly by ignoring Schiller and letting CFO Luca Maestri convince him otherwise.

Big yikes. Yeah, this was, by the way, like, we hadn't had much time to, like, read into much of it before we had last week's show, but Phil still looks pretty good in this. Like, I gotta say, like, you know, like, I've gotten up his butt a little bit over time with App Store policy decisions, but he looks like a pretty strong voice of reason here compared to the CFO team and Tim Cook, who...

basically ignored his very valid concerns. Although the bar is really low here. Like, the sentence basically said, you know, Schiller advocated that the company should comply with the injunction. Yeah. I mean, it's like, Schiller advocated that they follow the law.

what a rebel he is. The bar is really low. He's like, they kind of told us to do this. Shouldn't we kind of do it? And then, you know, Tim Cook is like, but do we have to really? Yeah, pretty much. Yeah. So it's, he comes out looking better than his siblings, but it's,

it's kind of, I mean, in the end, it's not his call. Like he can, he can suggest, Hey, I think we shouldn't do this because I don't think it will go well for us, but it's not his call. So I get, I mean, and he is, as Marco has noted many times has been what promoted to the roof as you're saying or whatever, like he's got one foot out the door. He cares about this stuff, but he's not running the company. And so, yeah, he's in there saying, why don't,

you know, like the court case is over. Why don't we just do what they said? And by the way, a lot of people wrote in to point this out and has been emphasized in future right and subsequent writing about this topic. But we didn't mention last time. So we should now this case that we're talking about. I said last time, you know, Epic won. They won this part of it. This is the anti-steering part. But the larger case was like Epic saying Apple shouldn't be able to have an app store. And the court was like, yeah, no, they can have it. But this anti-steering stuff that they're doing, yeah, that's BS. And they have to stop that. So, uh,

You know, Apple, quote unquote, won the case, except for this, you know, in all cases, it's like there was 10 counts and nine counts when Apple's. Well, I don't know if counts is the right phrase, but anyway, there's sub aspects of the case. The only thing we're interested in talking about here is the one thing that Apple lost, which is you can't stop people from telling their customers about better prices and sending them to external payment things and blah, blah, blah, blah. So that's the part that Apple lost. But they won the whole rest, which was like Epic was saying, it's no fair that Apple has an app store and can kick us out. And.

Epic did not win that part of it. Right. So I read, uh, I read most of the filing earlier today because I'm a glutton for punishment and I, you know, I skimmed what I didn't read. And for the most part, it was very readable, particularly for a legal document. Um, and I, I do a

overall agree with what Marco said that certainly all the poll quotes make Schiller look really, really good. But there's definitely other times in this document that he looks not so great, including like changing his story based on new evidence, which is to some degree fair, you know, like obviously as you get more information, you're allowed to change your mind. But the way the judge took it was not great. But still, it doesn't negate Marco's broader point that he comes off looking fine.

far and away the best of all the Apple people. And also because like if they, you know, if they had won this, this sub aspect of the case, like he would have been the naysayer who was trying to make them do something they didn't have to do. And in the general vibe is always like for all parties involved at the Apple executive level, like essentially, uh,

What can we get away with? Like, what do we really have to do? Because we saw before there was any lawsuits, what Apple chose to do in the absence of external pressure was what they were already doing. So it's clear the company was OK with that. And now it's a question of how much do we have to change what we're doing to get by? And Phil was more on the side of I think we probably shouldn't do that because we're pressing our luck in X, Y and Z, which is, again, not a very strong stride in position saying we shouldn't be doing this because it's wrong, because he could have been saying that before. But as far as we know, he wasn't.

But I did pull just a handful of quotes from the court filing. This was on page 24. Apple hired the analysis group. I'm sorry, let me back up. Apple hired a group to basically say, how much should we charge for our intellectual property?

You know, what is a reasonable commission? We clearly should be charging more than zero, right, right, right. And so what should that number be? And so with that in mind, Apple hired the analysis group purportedly to conduct a bottoms up study, which quote, estimated the value of services provided by the Apple ecosystem to developers quote. And the whole thesis of the judges, uh,

opinion about this was basically they backed themselves into 30%. And so this group concluded, and this is in the court documentation, that dev tools and services, for those efforts, for the dev tools and services that Apple provides, we developers should pay anywhere between 3% and 16%. For discovery services, which I'd like to come back to in a moment, anywhere between 5% and 14%. So hey, guess what? That's between 8% and 13%.

30%. Imagine that. And what I, and I don't want us to get wrapped around the axle about this, but like discovery services, five to 15, well, really 14% of my earnings should go to discovery services. What services is Apple providing me on a regular basis in terms of discovery? And even before you get to that, the premise, uh, being, uh, the, I mean, setting aside charging anything, the premise being, as I've said, Oracle style pricing, where, um,

You usually think of something costing a certain amount of money. How much does this car cost? How much does this house cost? But in the world of enterprise software and other situations where the pricing dynamics are a little bit different, they say, well, there's not really a price. It's more like a tax. So it's a percentage of some amount of money that you make.

which is a really strange way to price stuff. It's kind of the way that like traffic violations are priced in countries more civilized than we are, where instead of being a hundred dollar fee for speeding, it's a percentage of your income to make it equally, make it, make the deterrent equally sized for everybody who speeds as opposed to it being ruinous for people who don't have a lot of money and trivial for people who do.

So that's like dev tools. It's not X amount of money because, you know, these things they do, even though there are incremental costs for each additional developer, they're largely fixed costs for like running the app store and creating X code and all that stuff where you don't have to create X code again for every new developer. So again, there are some incremental costs here, but it's largely a fixed cost, but they don't charge for it like it's fixed costs. They say, well, if you make a lot of money, we want to make a lot of money. So we would prefer to have some percentage of,

of your income or revenue or whatever. And that's a great business if you can get it, let me tell you. Because instead of saying, we can't figure out how much to charge because some people have a lot of money, but some people have little. Can we just sell this new car? Can we sell it for...

25% of whoever buys it, it's their annual income. That's not how cars are sold. And I think it would be quite a deterrent if they were. And it's like, well, you don't understand. We have the factory and that's like 10% of your income. And then we've got the dealership and that's another 5% of your income. It's like, wait a second. I reject the premise that because you made these things and I want to use them or buy them that you deserve a percentage of my income. But yeah, so once you accept the premise, it's a percentage of income. Yeah.

And you accept that, oh, look, magically, somehow, the percentage they came up with is exactly the right amount to deter people from using this instead of just using an app purchase. What a coincidence. They nailed the percentage when they first started running it. And...

out of the goodness of their hearts they've been running the app store all this time at break even how generous of them yeah it's it's ridiculous so then i forget what page numbers these two follow the two final poke quotes were from and i apologize but reading directly from the document as the 2025 hearing revealed the analysis group's report did not materially factor into apple's decision making process it was created as a showpiece for the court the plan backfired the

All of that was the poll quote, including the plan backfired. That's from the judge's document. Incredible. And then finally, and I think we've heard this one before. We might have even read it last week, but it's so just incredible. Quote, quote, quote, quote.

Yeah, honestly, like, I can't imagine anything Apple could have done because, like, the whole point is, like, okay, we've decided to do this, and how can we defend this thing we've decided to do, which is still charge people even though they do payment outside the App Store? That's a decision we've made. Phil was against it, but I think we can get away with it. How can we defend that? Well, we need to show that, like, actually, we calculated our costs, and this is about what it is, and so you make up –

Can you imagine a set of like research that they had? Like there's, there's just nothing you could have presented to the court to say, well, we looked into our very own thing that we're doing already and we've decided it's probably going to cost the same amount as we were already charging. Like, I don't think there's any way that anything they could have done, like any outside firm they could have hired any way they could have written down. Like,

just add up the salaries of the people involved in the time they take or whatever, because it's just it's a nonsensical thing. Like you're starting from a premise that the court essentially rejected, which is that Apple should be allowed to charge anything for this, essentially, which is, you know, why they ended up with the judgment against them here. Yeah, I mean, it's like I think if you if you look back on where Apple goes the most wrong, right?

It's mostly hubris. Historically. This is true just as much in the Steve Jobs days as it is now. Apple mostly goes wrong with hubris. And this is...

Just full of so much. It's just like they really thought they could get away with totally disregarding this. First, they thought they could get away for a long time with very anti-competitive behavior as their market and influence and control over the economy grew.

They thought they could get away with this forever. They didn't give an inch. Like if there was ever an opportunity for them to give an inch or keep it all, they chose keep it all.

every single time and by the way that's another instance where phil came out looking pretty good because he did you made the joke about app store breaking even because it's something that steve job said early on before the app store was even launched i think uh but at some point very early in the app store is in uh history where it seemed like it was clear that it was going to be real popular i think phil had an email that we saw in a court case that was like hey you know after we start we break like

a billion dollars in revenue per year or something to the app store. Should we think about maybe like lowering our percentage or something? Or like, it was basically like once the app store gets really big, maybe we should change the rules. And I guess what you were saying, Marco of like, uh, as they went from a small fish to a big fish, it's easy for it to like creep up on you and say, yeah, we were, we were behaving as if we're this tiny fish, but now we're this giant power. Um,

we should maybe play by different rules. And Phil was saying that at such an astronomically low number, like they hadn't reached it yet, but it's like, when we get to a billion, we should probably have different rules. And it's like, guess what? We're going to sail way past a billion and never have a second thought about this decision, Phil. Sorry.

That's bananas. All right. So Apple's response, we'll put there's a bunch of different like Verge links and whatnot. We'll put them in the show notes. But Apple's response, Apple senior director of corporate communications, Olivia Dalton, sent a statement to the Verge that reads, we strongly disagree with the decision. We will comply with the court's order and we will appeal. And as of today, May 5th, they've appealed. So we'll see what happens.

Then additionally, I think it was the day after this broke, Apple did officially update its app review guidelines from the email that they sent to developers. The app review guidelines have been updated for compliance with the United States court decision regarding buttons, external links, and other calls to action in apps. The

These changes affect apps distributed on the United States storefront of the App Store. Yeah, they were pretty quick. Like the little statement was we disagree, we'll comply, we're going to appeal. They disagreed. They updated the things to comply and they're appealing. So they did what they said. Although the update was like in typical fashion, it was kind of like a text diff where it's like section 3.1.2a.

uh no longer says x y and z like it wasn't like a textual diff it was like a sentence description of like each section and what they it used to say that you know we no longer require this we we say this you can't like it was and people are trying to like parse it by looking at the original text and looking at this and then doing a text diff and there was lots of questions going on as we'll see in a second about okay but like what are the rules now yeah

So from NJ Size Roundup, one of the questions is, and this was raised by Ryan Jones. Ryan writes, at least how this is written right now, it seems like there's no mandate to link out. In other words, a Stripe Apple Pay button can be directly in the app or the entire paywall could be a web view with a button right there in it.

Now, with that said, Gruber disagreed. Gruber wrote in his post, this does not mean apps can now use alternative payment processing in-app. It doesn't even mean apps are no longer required to offer Apple's IAP in-app for purchases and subscriptions. All it means is that apps in the U.S. for now, but Apple really ought to make this worldwide, but I suspect Tim Cook wants to fight this on appeal in federal court, which he does, are free to inform users about offers available on the web and to link to those offers online.

And I think Gruber is right about this as far as I can tell, but honestly, with AppReview, who even knows? Who even knows, again, what AppReview will actually do? But the thing is, that's what people... Even if you have to link out, the prohibitions that were most egregious were you couldn't even tell people anything about pricing online.

You can say, hey, you know, here's an app purchase, which Apple makes us put in, but you can get it for 15 to 30% less if you click this link. And people are going to click that link because they want to save money. So even if it goes outside and goes to a web page, they're like, people don't care. People buy things on the web all the time. People buy things in Safari on the iPhone. It happens.

Happens all the time. It's a thing people are used to. So that is the real freedom that matters. It would be better if like Ryan Jones suggested, you could just do it inside the app with the web view or whatever, because then people wouldn't even know, but we'll see how the shakes out. I mean, this, the reality is like this whole thing about like the rule. I mean, this is one area that I agree with Apple, I think, uh,

The idea of drawing the distinction of like, well, you can't have payments in your app, but if you link out to the web browser and do it there, that's so different. Of course, it's the same thing. Like, who cares? You can make a one-click link now, now that you don't have one unsigned link that can't include any tokens or anything. Now you can just have the app generate a link that automatically logs in the user. So you click a button in the app.

It kicks you over to Safari with this user-specific URL. There's a payment screen there. It could even have a one-click payment option like Apple Pay or Shop Pay or Stripe's link thing. There's so many other things.

You can have a one-click payment there, and then it can redirect you back to a URL that launches you back into the app. You're basically having payments for your app to your own payment processor. You just have a couple of weird screen animations between each direction to get there. I know that we try to draw this distinction. Certainly, Apple doesn't want this, but there is no difference business-wise between

between this being in the app and having the app kick you over to the browser and kick you back into the app. Like, there is no difference. Especially if you're allowed to use web views in the app because then you literally won't leave the app. And you're not, to be clear. But I don't think that matters. I think what we've been arguing all this time is really...

that apps should be able to use their own payments if they want to. That's it. Like the idea of like, well, it's okay if it happens in the browser. I think that is a distraction and not a meaningful distinction to the actual use of these apps in practice. And I think we will very much see a lot of these like, again, like, you know, simple one-click things. You know, right now people are like, oh, everyone's going to build their own thing and you'll have to log in and enter your credit card. No.

I mean, yes, people will do that. But the more common case will be you'll be kicked over to a page that has Shop Pay or Stripe Link or something where you already probably have that logged in. Or it could literally be Apple Pay on the web. It could be an Apple Pay button. But I think what we're going to see here is if, again, if this sticks, which that's a huge if, but if this sticks...

We will see Apple compete. Maybe. I mean, because again, like Apple is at its worst when it doesn't have to compete. That's when we always see the worst behavior and typically mediocre products from Apple. Apple does great when they have a fire lit under them from competition. When they have real competition in an area, they compete very well. But when they give themselves an advantage, like, you know, some form of technical or policy lock-in,

We see them getting complacent and we see them releasing mediocre things or letting things linger forever, kind of unaddressed or non-competitive options. Apple being complacent in areas that they don't have to fairly compete in results in crap products and crap experiences. Right now, Apple Pay, in my opinion, is not the best way to pay for something on the web.

The best way to pay for something on the web, in my opinion, is ShopPay. I think it's way better than Apple Pay because in my experience, it is way faster and more reliable. I have all sorts of problems with Apple Pay. How many times have you tried to use Apple Pay and it's like, you have to update your billing address for some reason even though you put it in 10,000 times. And by the way, where do you put it? God knows. Like, there's so many weird little gotchas sometimes with Apple Pay or odd failures. I've never had ShopPay fail, not once, and it's just as fast. So, like...

Apple will actually be forced to compete in areas that like maybe they were class leading at one time but they have no longer they no longer are class leading and they've been able to be complacent because they had technical or policy lock-in that allowed them to be complacent now they'll have to compete and again I

I think that's better for Apple. Apple does not do well in any way when it locks itself in to its advantages. They do great when they have competition. They really like they put resources behind things. They update faster. They ship better quality products when they have competition. And all these areas that they don't, they stagnate. So I think one of the many ways that it will actually be better for Apple if they lose some of this control by force, I think one of the ways this will be better is

is it will drive them to improve their products to become more competitive. And that will benefit things like their hardware business. So I think overall, I mean, I've been saying for a long time, I'm glad that I'm

I'm glad that our friends in other podcasts, hey, Upgrade, I'm sub-podcasting you just like you sub-podcasted me. By the way, the Upgrade segment on this was excellent. You got to go listen to Upgrade this week. It was very, very good. I can't believe how fired up Jason got. But it's very encouraging now to see that many other people on many other podcasts and publications are now coming to admit the possibility that maybe we need a new CEO at Apple because

Time after time, we keep seeing Tim Cook making decisions that really seem to be about short-term profit and not long-term strategy. And I think the long-term strategy is when Apple makes better solutions and better products, they make more money from their products because more people buy them because they're more competitive and they're better.

So I think it is the best long-term strategy not for Apple to create more areas in which it can be complacent and get mediocre, but to actually keep making their products better. And the only way they consistently do that is when they have competition. So I actually think this will be good for Apple to have this control forcibly taken away from them. We'll see how it plays out.

All right. So with that all in mind, Tim Sweeney is grandstanding again. Tim writes, we will return Fortnite to the US iOS app store next week. This was a couple of days ago. Epic puts forth a peace proposal. If Apple extends the court's friction-free Apple tax-free framework worldwide, we'll return Fortnite to the app store worldwide and drop current and future litigation on the topic.

Tim Sweeney's got a little bit of a Trump vibe. Like, he'll start with a statement like, we will return it to the App Store next week. That's not a thing he's empowered to do, but he states it like it's going to happen. But then you get to the next sentence. This is all from him, by the way. This is a direct quote from him. It's not us summarizing or anything. If Apple does this thing that we want them to do, then we will deign to bring Fortnite back. It's like, I'm not sure you're in a position to make that statement.

demand Tim, but you know, there he is out there saying stuff. And my initial thought when I first heard this was,

Why would Apple care? Because you know they're not going to make any money from it. But then I realized, no. Again, what I was just saying. They should care because having Fortnite on the iPhone increases the value of people having iPhones. Yeah, that's what I said last week. Like, it could come out as a net win for them because now suddenly you have iOS with Fortnite back in it. But, you know, will Apple...

Will Apple see it that way? That's the thing. Will Tim Cook see it that way? I guarantee you no, because he has shown time and time again, he sees zero value in a software ecosystem for his devices. We now have documentation showing how those top Apple execs think of developers. Well, and also they really hate Epic. They really hate Epic, but they also really don't respect software. Yeah.

Like they really do not respect third-party developers developing stuff for their platforms. They see us solely as a resource to be harvested, and they do not see value that we bring to their platform. I understand. Look, all the Apple people who are out there getting mad at me right now, I understand that other people in the company think differently. But you look at what your top execs have literally said and have shown up in court and stated themselves or have now shown up in discovery and documentation, they're

What they say says clearly otherwise. They clearly see developers as a resource to be extracted, and they do not see any value that we bring to their platforms except fees. They see no value for our apps to exist on their platform except what they can extract directly from fees, which is...

scarily incompetent for the leaders of computing platforms. I think you're stating their position a little bit more extreme than I'm not. Read what they say. I feel like what you're trying to say is they don't see as much value as developers do or as most outside observers do, which I totally agree with. No, their argument is literally that developers have like a free ride if we're not paying them. They're like, we get our apps for free. Like we get to ride on their platform for free. What they're saying is that we would,

we would benefit from the services they're provided without paying what they feel like is an adequate amount of money to pay for those services because the $99 a year is not sufficient to cover them. I think you're being very generous to them. I mean, to be clear, I disagree with their position entirely, but I think you're making it a little bit more extreme than it is.

it is honestly you are making it what you hope it is but what that what they actually say is very different oh you haven't you haven't brought me you haven't brought me the the receipts as the kids say to show me the statement that is that extreme well your homework john congratulations you've brought this upon yourself hey congratulations you played yourself if you're gonna make me read the judgment i already did that yes so oh well see then i i think marco was being slightly

Extreme, but only just. Only the teeniest bit. I mean, it's a moot point. We don't need to argue this because I agree that they need a leadership change. So it doesn't really matter. When it comes down to it, they need new leadership in this area. And if this area goes all the way to the top, then they need new leadership at the top.

That's actually an idea I have for a future potential member special, but I might bring him to the main show. We'll see. All right. So back to Tim Sweeney grandstanding. He has said that, oh, or yes, he did say publicly, oh, I am going to, or we, Epic, is going to bring Fortnite back by way of the

Swedish Epic Games Apple App Store account. Somehow, Fortnite has returned. Yeah, right. So Tim wrote, we have conversed with Apple on the topic and we will use our Epic Games Sweden account to submit Fortnite to the US App Store. We created this account last year to launch the Epic Games Store in Fortnite in the European Union and Apple required an

an EU domiciled account. So since that's sitting there and that hasn't been banned from the app store, like the American one was, apparently that's their backdoor. It was. Remember, they did ban it. Then it got unbanned. There's this whole history behind it. You can follow the link to see it. But like the Apple did, they created it after their thing was banned. And then Apple said, ha, you can't create that over there. We're going to ban that too. But then they unbanned it. And here's the thing. Now he goes from being Trumpish to being more lawyerish because he says, we have conversed with Apple on this topic, which could be a true statement.

And then we will use Epic Games account to submit. Okay, that could also be true. But those two things do not combine to equal. Therefore, Fortnite will appear on the App Store because those are, yeah, that's, we'll see what happens here. But Apple is super mad at Epic for reasons that make sense. And I don't see them being particularly kind to Epic on this. It's clear Apple has already decided that they have weighed

benefits of having a fortnight on a platform or, and the detriment of Epic being annoying. And they chose not to have fortnight. I'm not sure the math has changed for them. Yeah, I agree. Uh,

All right, so trying to move quickly here. There's a litany of people that have come out of the woodwork and said, ooh, ooh, there's a business opportunity here, which there is. So very quickly, Epic Games has decided they're going to set up web shops for iOS developers. Reading from MacRumors, Epic Games today announced plans, this is on the first, for Epic Games Store web shops, a feature that will allow developers to launch digital storefronts that are hosted by the Epic Games Store. Now, this is where I think it gets interesting. The Epic Games Store will charge developers a 0% fee for the first $1.

million in revenue they collect per app per year. And after that, developers will need to pay Epic a 12% cut. The fees are applicable to all payments that are processed by the Epic Games store. Epic Games also says that players that spend in Epic web shops will be able to accrue 5% Epic rewards on all purchases.

Competition. Mm-hmm. Imagine that. Stripe, as was foretold, I'm pretty sure I brought this up last week, Stripe has said, well, let me just read from The Verge. Payment processing platform Stripe just added a way for iOS developers to accept payments by linking outside their apps, dodging Apple's commission fees.

Stripe has published a guide that shows developers how to accept transitions outside an app using Stripe Checkout. In the sample video, you can see someone preparing to buy digital fruit with in-app currency, but instead of showing a transaction page inside the app, it launches Safari and takes the user straight to a Buy Now page.

This is functionality Stripe already had, but they quickly put together a little guide and says, look, we already, this is the thing we already do. You can add this to your app. Look how easy it is. Competition. Yep. With sample code, their doc, I mean, I skimmed their docs super fast, but it looked good. Then Patreon has said they will update their iPhone app. So page reading from the Verge, Patreon is planning to submit an update to its iOS app that will let creators accept payments outside of Apple's payment system. A spokesman, a spokesperson, Adia Taylor tells the Verge.

Last year, Patreon said it was forced to switch to Apple's in-app purchase system, which applied a 30% fee to all new memberships purchased in the app or else risk, quote, being removed from the App Store, quote. I will say, though, everyone is rushing to take advantage of this. And if you already have a business that can do this, that can take advantage of this quickly, great. I would say, though...

Don't count on this long term. This is a brief, possibly brief free-for-all. But if Apple wins any kind of appeal or some other injunction, I don't know how these things work in that kind of detail, but the second Apple doesn't have to do this anymore, they will close that door shut and cut everybody off if that happens. So,

I would not rely on this. Like if I was like, you know, an investor, I wouldn't invest in a company that relies on this. If I was making a brand new app, I wouldn't make a brand new app that relied on this. Like take advantage of it while you can, but it is still in flux. It can still be, it is still being appealed. It only applies to the US, you know? So there are a bunch of asterisks on this. And so enjoy it while we can, but don't assume it will always be this way.

Yeah, you got to have conditionals anyway, because like I said, it's U.S. only. So just keep those conditionals there because you might need to, you know, turn off one of those feature flags if this thing goes the other way. Yeah, very abruptly, by the way. Then friend of the show, Riley Tested, writes, I just submitted a new Delta update without all the external payment restrictions. Goodbye, scare screen. And then on The Verge.

You can see some of the details of the scare screen, which we've talked about in the past. I mean, we can go into this more, John, if you want, but I feel like we've covered this before. Do you think I want to, Casey?

I try, let the record show. I try to avoid just sitting here and reading at you. I don't put things in here for no reason. All right, well, here we go. Buckle up. Reading from the verge. This is about, these are details about the scare screen. I just put this here instead of earlier because Riley specifically mentioned the scare screen and we've talked about it before, but we learned more about the origins of the scare screens from, I think this is from like the discovery process or something during the case.

Apple chose to iterate on the scare screen with the goal of dissuading users from continuing on the web. The pop-up included a paragraph of text and employees discussed using scary language to warn people off. Raphael Onak, a user experience writing manager at Apple, instructed an employee to add the phrase external website to the screen because it

Sounds scary, so execs will love it. I'm sorry, but that's just patently hilarious. Another employee gave a suggestion on how to make the screen even worse by using the developer's name rather than the app name. Ooh, keep going, another Apple employee responded in Slack. Even Cook got in on the action. When he finally saw the screen for approval, he asked that another warning be added to state that Apple's privacy and security promises would no longer apply out on the web.

So this is a demonstration of a bunch of things here. So first of all, I mean, this is just the way it is in big companies. If your job in the company is to make this screen and your bosses have told you, we want to discourage people from going outside of the app to make payments, then

how scary can you make this scare screen? So there they are being dutiful employees saying, well, we need to, we want it to be truthful, but we also want to phrase it in a way that is as scary as possible. So we're not going to lie or make up stuff or whatever, but saying external website instead of just website. Ooh, it's scary. Execs will love it. We feel like we're doing our job well. Uh, and then this one I thought was great because it's like, um,

using the developer's name rather than the app name it's true like the app store if you go to the app store you see any app that's available you can see what the app is called but you can also see the name of the developer but not using the app name in the message is scarier because people don't know the name of the company that makes their app like even now whenever it's like my wife is always saying what is this charge from you know boingo co and i'm like i don't know

probably software because like I bought, you know, app XYZ. I might not even know what the name of the company is, but it shows up as a charge from the company. So, I mean, just off the top of the head of the people who are listening, do you know the names of all our companies? Um, I bet you don't, right? You might know the names of our apps, but our company names are different.

So put the company name in the message, because if someone sees a message that says, you know, as the message says, we have a screenshot here. Any accounts or purchases made outside of this app will be managed by the developer. And then in quotes, call them scare quotes if you want. Test tech. It's like, what the heck is test tech? I'm using Delta. I don't know what test tech is. That sounds like someone's trying to steal my credit card. I'm not going to go to this link.

That is clearly a documented, conscious decision by Apple employees to make the message scarier while obscuring it. It annoys me in various instances of in-app purchase where you don't have enough characters to say something sensible, and that's Apple's own system. So again, competition would be better there. Yeah. But this entirely reveals that, like, you know, Apple could say, we're not trying to scare people away. We're just saying the honest truth. Like, you will be outside of it. And everything in here is true, but it, you know, the discovery...

Discovery is a B, as they say. We were totally trying to make this scary, and we were doing it in ways that essentially hurt the user experience. And they felt so...

empowered to do that that they knew they would be rewarded for it like how much effort has been spent making their products more competitive in these areas compared to how much effort they spent on this intentionally degrading it so that it would be as uncompetitive as possible of an alternative like this is what these people this is what was encouraged by their higher ups and by the company culture it goes all the way to the top my friends i hate to tell you like

They are telling you right here, this is what they think of you. Believe them. This is what they think of you. This is how they develop software. This is how they make decisions. This is what they are optimizing for. There is no greater story about privacy and security. This is what they actually care about right here, laid bare.

At least this part of the company. Developer Relations probably has a different attitude. But you know, this is the part of the company that deals with App Store. And speaking of effort, they put in so much effort in DMA compliance to essentially make it as difficult and unappealing as possible. It would have been so much less effort not to do that. But this part of the company that's responsible for complying with these judgments that they don't want to do in the first place is absolutely doing everything it can, doing a tremendous amount of additional work to make it so that all...

All the alternatives that they're being mandated to do are as unattractive as possible, which is what we all said when we saw what they were doing. But here is documented evidence from inside the company that it's not that's just not just a side effect of what happened. It is consciously what they're trying to do. Yeah, this is this is all part of why I say it's better for Apple.

If they get cut off from this drug they're addicted to because look at how much it is perverted with their culture. Look at how much this has corroded and destroyed their priorities of things like user experience. This has been so bad for them culturally.

It's been great short term. It's like a drug. It's been great short term for their feelings and their numbers short term. You keep saying short term, but that's the correction I would make before. Tim Cook made these decisions not for short term profits. He made them for long term profits because they've made years and years and years of tremendous profits. So now he's making terrible short term decisions about not complying. But when these decisions were made and policies were set down, long term.

long-term profits. Unfortunately, they're not true to the spirit of the company, which is we're supposed to make products that people love, and then from people loving them, we make money. This is sort of the opposite of that. Right. And the whole time, he has been... I don't know if it legally qualifies as misleading the shareholders, but I would certainly say BSing the shareholders in terms of the whole services category. They have been knowingly peddling a narrative that services is stuff like Ted Lasso, and the reality is the vast majority of services money comes from fees.

the Google search deal and its app store fees. That's the vast majority of services revenue. And so they've been kind of BSing the whole industry and the shareholders into being like, oh, they're making all their money from services. But it really comes down to these two things.

These two large chunks of fees that are possibly in a precarious spot right now. I mean, this whole thing, again, this has corroded their thinking. This has perverted their incentives. And this is not Apple at its best. We have seen the great things Apple can do at its best. Look at the hardware. The hardware is amazing these days. A lot of the stuff they do is great.

Then you have this other area where they're really doing a lot of harm, all for short-term numbers. It's short-term, quick-hit numbers that I think are worse for the products in the company long-term overall. So I hope we get leadership in there that cares a lot more about the products and the long-term strategy of the company because Tim Cook does not.

Then Spotify has also announced, hey, we want a piece of that action too. So reading from MacRumors, Spotify has updated its iPhone app in the U.S. with out-of-app pricing and subscription options for its premium plans. Spotify users in the U.S. can now view pricing information for its individual, duo, family, and student plans directly in the iPhone app. And there are buttons that lead to Spotify's website where users can complete the payment process.

Spotify shared the following statement, quote,

We can now give consumers lower prices, more control, and easier access to the Spotify experience. There's more work to do, but today represents a significant milestone for developers and entrepreneurs everywhere who want to build and compete on a more level playing field. This is more or less the sentiment of everybody. The reaction to this from developers and users alike is essentially celebration.

Because users want to pay less money, right? Yay, cheaper things. And developers are like, finally. And I put screenshots in here of what it looks like on Spotify. So I think it's an interesting thing that may become standard if this turns out the way it is. It's a button in the app, but it's got the little box with an arrow pointing out of it. Not the shadow pointing up, but like an arrow pointing up and to the right. To essentially say by clicking this, you know, it's the open a new window feature.

glyph essentially the open a new window icon but on a button to let you know oh you're going to be sent to a web page where you'll do some payments and we'll check you back here and i wonder if that if that will become standardized across all these apps that the people will come young people maybe if this becomes standardized and sticks they'll become they'll start associating that icon not with open a new window but instead with pay because that's how you pay for things on the phone right it's like well let me tell you a story right and

We are sponsored this episode by Delete.me. Delete.me makes it easy, quick, and safe to remove your personal data from hundreds of data broker websites at a time when surveillance and data breaches are really common. You know, the data broker industry is really creepy, I think. It's not one of my favorites because what they do is they basically compile as much personal info about as many people as they can, and they bundle it, and they buy it, and they sell it, and they make it easy for anybody to go and find.

This can be things like your home address, your phone number, your family member's names, like really creepy stuff that you don't want to be that easy to find in two seconds on the internet. So what Delete.me does is they've developed all these opt-out procedures to go get your data taken out of hundreds,

of hundreds of data broker sites out there. Then you just give it to them once, you tell them, here's the NFL I want to be removed from these sites, and they will go and file all the takedowns and all the opt-outs. And as new data brokers come up, they figure out new opt-outs for them, and they'll file those too. And they will send you a report on what they were able to remove, where they found it. And it's a great service.

to just get your stuff out of all those data broker sites and really cut down on that creepy factor. So take control of your data and keep your private life private by signing up for Delete.me. Now at a special discount for our listeners. Get 20% off your Delete.me plan when you go to joindeleteme.com slash ATP and use promo code ATP at checkout.

The only way to get 20% off is to go to joindeleteme.com slash ATP and enter code ATP at checkout. That's joindeleteme.com slash ATP, code ATP. Thanks to Delete Me for sponsoring our show.

We have some people who wrote in with regard to Ubiquity. The original person who had asked about it was Ryan Budish, who writes, I get what Marco was saying. Cloud gateways are the routers and you need switches, access points, and some kind of PoE or power over Ethernet, which may or may not be built into the switches. But I think Marco, as someone who has a hobby of putting networking equipment into homes and restaurants, was perhaps slightly underestimating the complexity of the Ubiquity offerings for someone with less experience.

For example, they offer seven, excluding the rack-mounted enterprise and large-scale, cloud gateways, 15 access points, not including the enterprise outdoor mega capacity or bridging offerings, and 14 or more switches, not counting any of the rack-mounted switches like the one that Marco was telling Casey to get. I think John was trying to ask the question I'm getting at. For a technically-minded homeowner who probably has more than the average amount of internet-connected devices but is running neither a restaurant nor an enterprise from their home, how should they evaluate the 1,470-plus different combinations of routers, switches, and access points?

I mean, this is a little ridiculous. Like, I get Ryan's point. Like, it is. There are a lot of options. They have more options than, for example, Eero or Apple used to. Yeah. Right. But again, this, I mean, this is enterprise gear. Enterprises have lots of different needs. And I think the products, I think you can actually look at them for the most part.

be pretty sure whether something is for you or not. You know, like for instance, do you need a 16 port switch, a 24 port switch, a 48 port switch? Well, how many things are you plugging in? Like, you know, those questions can be answered pretty well. Do you need to be POE? Do you need it to be 10 gig? You know, you know, the answers to those things when you're buying them. For the most part, keep in mind, again, this is enterprise gear.

most of their offerings, you can pick the lowest one that fits your physical needs. So if you have five ports, you can get the cheapest five-port switch, that kind of thing. Because for the most part, the higher-end stuff, these things scale from Casey's House to sports stadiums. They're designed to serve that entire market. The existence of the higher-end and more specialized products

doesn't necessarily make it harder to know what you can get. I would suggest the only area that I think it really, where you really don't know is when you look at the APs, the wireless APs, because some of them are like pro versus light. Sometimes they have ones called long range. They obviously they do have some called outdoor. I think obviously the outdoor question is pretty easy for most people to answer. But between like pro and light,

You can just get Lite. It's fine. You can get Pro if you really want to. The price difference usually is not like massive. It might be an extra like 60, 70 bucks. So it depends on, you know, your price sensitivity and how many you're going to be buying. Usually for home use, the Lite APs are fine.

I've used a mix of lights and long ranges and pros in my various installations over time, and they've all been fine enough for my needs. The options that seem to change the most often are the cloud gateways, aka the routers.

And again, those are mostly it mostly comes down to like how much processing grunt do you want and what form factor do you want it in? Do you like if you need rack mount or not, you probably know that. So, you know, if you need built in hard drives to be to serve as a video camera recorder, you probably know that, too. So most of it comes down to like how big do you want the internal storage, if any, and how much processing power do you want?

And that, again, that really depends. Like, do you have a gigabit connection? Then probably get one of the medium to high-end ones. Do you have, you know, just a basic thing for an office with a few users? You can get the cheap ones. Fine. So there's a lot of different options, but that's because they serve a lot of different markets. But it doesn't mean it's necessarily impossible for a home user to figure it out. But, you know, hey, if you want more help, we're

We're happy to keep getting this question every year or two of like, what ubiquity stuff should I buy? And I'll give current recommendations at any given time.

Then Brian Donovan writes,

The most important thing, networking performance was worse. Streaming to my Apple TV became less reliable, so the family noticed. The setup experience was worse because it might see the new devices as I plug them in, but then it would just sit forever on configuring. I ended up having to reconfigure the network from scratch multiple times, but the performance issues never went away. I ended up reverting to the Google Wi-Fi. Is this my fault? Probably.

but I'm not sure what I should have done differently. I hope you cover this topic again and that Marco tries a little harder to get into the beginner's mindset to the extent possible. Despite my negative experiences, I'm willing to give it another try. I do need to replace my nearly decade old wifi equipment eventually and ubiquity is still in the running. So I put this in here for a specific reason. Let me just, I'm going to give a brief, very fast, hopefully not too inaccurate overview. The people will yell at me about who know more about networking. Um, but this gets to, uh,

One of the things like I think maybe where this person went wrong is they have a Google Wi-Fi network, which I assume is a mesh network, and they were just trying to replace it with a ubiquity one, like one for one, as they said, which I'm not sure whether we're working out. But I just this is a good sort of ATP 101 brief segment here.

We talk about mesh networks and Wi-Fi all the time, and I'm not sure people, I mean, I don't know, maybe everyone listening to this podcast knows the difference. But for the people who don't, I want to explain because I think it's super important. Back in the old days, you'd have a Wi-Fi access point in your house, and if your house was really big and the Wi-Fi access point wasn't centrally located, there were rooms in your house that got bad signal.

Mesh networking was an attempt to help with that. And that would let you have a place in your house where you have Wi-Fi. And if you had bad signal that was far away from the place in your house where the Wi-Fi thing was, you could put another Wi-Fi thing. And the reason that would help is because your device, your phone, your laptop or whatever, would talk to the Wi-Fi thing that you are close to.

And like a relay race, that thing would say, oh, I just got someone over here talking to me. And it would send its signal to the next thing that's closest to it and so on and so forth until it eventually gets back down to the one router that's connected to your internet. Like sort of, you know, relay race or, you know, handing off.

You talk to this point, this thing talks to that point, that thing goes to that point. That is a mesh network. It's a series of devices that all see each other, configure each other, and then you just wander around the house and the Wi-Fi access points tell your device, oh, it looks like you're going into that room over there. You should probably talk to this access point because it's closer to you. And then your phone would say, okay, and it would talk to that access point. And whatever access point you're talking to, you talk to it and it passes on to the next one, the next one, hop, hop, hop until it gets to where it's going. That's mesh networking.

And it's better. It actually extends your signal better than like repeaters and other stuff like that, because if you need to reach some other point, you just buy another one of these things and you plug it in somewhere and then it sort of self configures. My question for both of you, which I don't actually have the answer to because I'm not researching Ubiquity like you two are, is does Ubiquity have mesh networking nodes? Yes.

And most of the access points can also support like a mesh uplink, which is basically you give them PoE, but they don't have to be connected to the network. They will wirelessly peer. However, this is where I will say if you need wireless meshing, like if you cannot run network cables to each AP from the central source, don't bother with Ubiquiti.

Those are features that they have, but they are not really optimized for that market. I think the Google Wi-Fi is fine for that. If you're going to be doing that kind of meshing where, like, you know, you have basically repeaters. And I know the mesh people get really sensitive when you say repeaters because that's something else sort of technically maybe, but, like... It absolutely is. That's the whole important distinction. Yeah. It is. But, yeah. But they...

they don't work that much better than the ones with repeaters. It's nowhere near the performance of wired. Oh no, they absolutely do. Maybe the ubiquity ones don't, but like that was my question. But it's like, when I see you guys talking about ubiquity stuff, I definitely see that the ubiquity is, I didn't know if it was just not possible or they were just leaning heavily into it. But like to clarify when ubiquity, if you buy like wireless access points, say you bought one of them, like you're either your router has a wireless access point and some ubiquity things do, or you buy a wireless router and another access point.

Oh, I can't get signal in this far part of my house. What ubiquity expects sort of like the default of ubiquity thing is, oh, well, of course that other part of your house slash office slash stadium, uh,

just take the ethernet drop that's in that room and plug in another wi-fi access point and what that means is like that wi-fi access point would be in the far part of your house where you get bad signal and your device would talk to that and you know what that wi-fi access point would do it would send your signal along the ethernet wire that's attached to it and is probably powering it back to the router it wouldn't say oh i gotta find some other access point and wirelessly talk to it and they can talk to the next one and hop hop hop no it's got a wired connection to your actual network and

And that's how like, you know, Wi-Fi and offices work. You run Ethernet to all over the place, every place you need an access point, like in hotels, in stadiums, in offices, the Wi-Fi access points have Ethernet wires connected to them. So whatever access point you're near, you go wirelessly to the access point and then you're all wired back.

The reason people love mesh networks for home is people don't have ethernet in every room in their house most of the time. You want to reach wireless to the far room that gets bad signal. And all you want to do is plug something into power in the wall and you'd be like done and done.

And ubiquity products are often so far from that that you can't even plug them into a wall unless you get a power over Ethernet adapter because they take all their power from the Ethernet cable that they expect you to be plugged into. And you're like, how do I plug this into the wall? I'll just get a PoE thing and adapter. It's like, okay, well, how does it get its signal? So I'm glad to know that they can do mesh networking, but the home focus products like Eero and apparently Google Wi-Fi and

Orbi and all those other ones are so focused on the home case. They like nobody has ethernet all over their house. Just assume that the only thing this is going to be plugged into his power. And they also include dedicated radios just for doing the, I think they call it the backhaul or whatever, just for doing the between node communications. They have radios for talking to wifi at all the different frequencies.

And then when your signal hits an access point, it's got another radio whose sole purpose is to talk to whatever is the nearest mesh node to send your signal hop, hop, hop on its way back to the one router that's actually plugged into Ethernet.

Yes. And that is like, I actually don't know if ubiquities routers have like separate, I mean, they all have, or if they're APs rather have, have separate radios for that wireless backhaul. They do all have multiple wireless radios these days. So they, they might be able to just like, you know, use one for that.

But that is not a market that Ubiquity focuses very much on. So they don't put a lot of effort into it. Yeah, I think that was one of the upgrades in Eero was like they used to have multiple radios for like 2.4 and 5 or whatever. But then they did an upgrade a couple of years back that's like actually we're going to add a whole other radio just for the backhaul instead of reusing any of the other radios and that helped the performance. Yeah, so I would say like if you are having wireless linking between APs, if you cannot wire them all back to the home switch...

Just stick with the consumer stuff. Stick with Google and Eero. Those are going to be fine for that purpose. Until Ubiquity decides they want to address that market, because they totally could. They have all the technology, and it seems like it should be possible. The reason I wanted to ask about this is because I think a lot of our listeners are just like, oh, I just have a home setup, and they might already have one of these mesh ones.

And I'm not sure Ubiquiti addresses them that well yet. I don't see any reason that they couldn't, and I hope they do, because they seem to have way more home products. Like, I just saw recently a video showing the Dream Machine, a little R2-D2 trash can thing. That's so, like, it's basically like an airport express. Yeah, but that's one of their pro routers. That's what I use at the beach. Right, it's got Wi-Fi in it, and it's a router, and it has a tiny bit of storage with an SD card slot.

But it's not, it's not a, it doesn't seem like it's a mesh network out of the box. I mean, I still expect you if you're adding APs to have other wires. Anyway, that's, I think this is an important clarifying point for people who are dying to get ubiquity stuff because it seems like they haven't quite reached that far down to the consumer realm for people who absolutely need mesh networking. Yeah.

But it's not very good, and it's not going to be meaningfully better than what Google and Eero are doing for that. It doesn't sound like their products are focused on that. No, they can do it, but they're not focused on it, no. And certainly it's not going to magically be better than Google and Eero substantially in those areas. This is designed for people who are going to be wiring things, Casey.

It's designed for people who are going to be wiring every AP back somehow to the home switch. So what I would get to answer the question out of today's lineup, what I would suggest for a typical home full of a nerd, but other people who tolerate the nerds,

I would say your router choice right now is the Cloud Gateway Max or the Cloud Gateway Fiber. The Cloud Gateway Fiber is higher end, it is more expensive, and it is out of stock. Another thing to know about Ubiquiti is that it turns out we are not the only people who like them. Many people have learned the good word about Ubiquiti. And so whenever they have a new high-demand product, oftentimes it is out of stock.

What you need to know about Ubiquity also is that if you go on Amazon to buy their products, they are usually sold for above retail price on Amazon by third parties.

You can also get them from other retailers like B&H Photo is a Ubiquiti retailer. And so sometimes I'm able to get stuff from there if Ubiquiti doesn't have it or if it's if it's, you know, it might be faster from B&H for me. I'm in New York and they're in New York. But for the most part, they don't have that many other first party retailers that sell things at MSRP. So usually if you want something from Ubiquiti, buy it directly from them. It will usually be substantially cheaper than the same thing on Amazon.

And sometimes things go out of stock when they're new and in high demand. Sometimes they're hard to get for a long time, usually not for too long. And they have alert me when it's back in stock. Anyway, so going back to the question, what I would suggest, Cloud Gateway Max or Cloud Gateway Fiber for the router right now of the current lineup.

If you want an inexpensive PoE switch, the Ultra switch, like just one word, Ultra, it's an 8-port switch that has two different options for AC adapters, 60-watt and 210-watt. If you're going to be powering a bunch of sub-switches, I would get the 210. If not, get the 60 if you're just going to be powering a few APs. And then for the APs, you can use the U7 Lite and get two or three of those for a typical American house.

So those are $100 each, so $200 or $300 for your APs. You got about a little under $200 for the Switch and $200 for the router, so you're in $600 maybe. And plan on running Ethernet to all those access points. Yes, you should run Ethernet to all of your access points. That's a really good home setup for almost anybody. Do you have any U7s installed in either the restaurant or your house? It doesn't matter where. Yes, I have my new Long Island house is all U7-based.

pro i have two pros one light and one of the little uh in wall ones the reason i ask is i've heard rumblings that the wi-fi 7 support is real spotty and in general the u7 line of wireless access points is not nearly as not physically sturdy but like um

computationally sturdy, for lack of a better way to describe it, as the U6 line. But I don't have any personal experience with this, so maybe I'm lying to you. Yeah, I know our friend Stephen Hackett had that problem with the U7s when he put them in his house. For that reason, I bought all U6s for the restaurant, even though I have U7s here in the house I'm sitting in talking to you through right now. But I figured, like, you know, just in case, let me stick with the U6s because, you know, I don't need...

all of the fancy like Wi-Fi 6, 6 gigahertz radios that the U7 added. So it's fine for my purposes in the restaurant. But I will say I've had no problems with the U7 stuff. And that is, you know, sometimes Ubiquiti will have like one product that has like a little bit flaky software. And, you know, you get a few software updates later and usually it's fixed.

So that's I think the U7 AP line is pretty new still. It's only I think a year or two old. And so I think people had issues with it early on. I don't know how many of those issues remain because I'm only using it in one place.

And in that one place, I have no issues, but who knows? But those are definitely software support. Do you happen to know just off the top of your head what the major noteworthy differences are between the light and regular and pro editions of the access points?

You have to check each generation because they vary, but typically the Pro usually has way more radios. And so the idea of the Pro is it's made for higher density environments. So that's the one you put in the stadium. You put a bunch of Pros because that's when you're going to have a lot of clients coming

on like you know talking to an ap you need you need to like you know share a bunch of bandwidth and do a lot like so if you want like a lot of radio bandwidth get the pro for most home use that's really not necessary now if you care a lot about your wi-fi transfer speeds if you want to like run a speed test from your brand new you know iphone that has wi-fi 7 or whatever you

You want to run that speed test and you want to get like, you want to, you want to max out what you can get, get like, you know, over, over a gigabit, you know, if you want to do that, yeah, you're going to want the pro probably. Cause that's going to have like the most radios. I can start using them in parallel and stuff like that. But for most people's needs, including my own, frankly, the lights are fine. Good deal.

So you had mentioned that the Cloud Gateway Fiber is new and never in stock. And let me tell you a tale. I've been talking with Stephen about doing this in fits and spurts in the house and properly wiring the house for Ethernet. The way it exists today or existed last week, spoiler alert, is that I had an Eero Pro 6, I believe, in the office running as a main router.

Um, then by way of a Mocha bridge that converts from ethernet to coax, and then another one on the other end, I had an Eero pro six in the primary bedroom, uh, working as a wireless access point again, with effectively a wired backhaul by way of a combination of coax and ethernet. And then an Eero pro in the living room as an additional wireless access point. And, um,

Stephen has said to me, you know what, I really recommend the Cloud Gateway Fiber. In Stephen's experience, he had originally bought the Cloud Gateway Max that you had recommended, and he found that he was having trouble saturating. I think he said upstream, but I don't recall. One direction of his connection to the internet, he found the throughput was not as good on the Cloud Gateway Max, and the Cloud Gateway Fiber fixed that problem. Your mileage may vary. I don't know.

But I noticed that the Cloud Gateway Fiber is apparently sold out till kingdom come. And so one of the benefits of having my apparatus, which this is a term I stole from Merlin, is that one of my 350 Docker containers is a thing called changedetection.io. That is the website. And what you can do is you can self-host a thing that will basically just ping away at other websites and tell you when they change.

And over the last several releases, it's also gotten a feature where it sees that, oh, you're looking at a product page for something and it will parse out without you having to provide regex or anything like that. It'll parse out what the product is, how much it costs and whether or not it's in stock.

And I have that hooked up. It's incredible. And they have a hosted version for the record. Like I run this myself because it's not very computationally intensive and I have Docker set up and squared away. So it's no big deal. Right. But you could, and I genuinely don't know how much it costs, but you can get a hosted version of this where you have them host all of this for you.

But I have something like 20 or 30 different web pages that I monitor on this, none of which are really critical in the grand scheme of things. But it is nice if you have a place where it will tell you when something changes. And so I put the Cloud Gateway Fiber on there.

And I tweaked it so that it'll check every like 10 minutes or something like that instead of the like once or twice a day that I would for most things. And sure enough, in combination with pushover, which I've talked about many times in the past, which is a mechanism by which you can send yourself text message or, excuse me, push notifications. And it has an API that a lot of different open source projects, you know, integrate with. So I got a push one day last week saying, oh, the Cloud Gateway Fiber is in stock. And so I ran upstairs and Insta bought one.

And late last week, I was able to find the time to install it. And I got to tell you, this has been both a blessing and a curse. So the installation process was... First of all, congrats on snagging one. That is not easy for a high-demand Ubiquiti product. So good job. Yeah, I'm telling you, change detection.io, for something like the Switch 2, it's not enough. Because that's just like...

bananas demand. This is a lot of demand. That's a whole different level. I didn't ask you to. Did you try to get a switch on switch to on launch? I did. And there's a, there's a whole story there that maybe we can talk about in the after show. It's nothing bad, but it's just not worth getting into right now. But the short answer is yes, we have, we have a reservation. How we got that as a whole thing, but we'll come back to that. Did you go on? No, I just, I tried like when I woke up that day and it was already over. Yeah, no, it's not, it's not the kind of product you wait to wake up about. Yeah.

Yeah, I didn't care that much to get it on launch day, but I'll try to get one when it comes out. Maybe it'll take me a few months, but it's fine. And John, you got one at midnight? I got one the old-fashioned way, lots of open tabs and devices. Atta boy. So anyway, I got the Cloud Gateway Fiber, and let me just jump straight to the end. The worst part about the Cloud Gateway Fiber, which you warned me about, which Marco warned me about, which Stephen warned me about,

is now I need all of my networking equipment to be ubiquity. I need all of it to be ubiquity. And I don't want to spend that kind of money all in one shot because I like being married. So I don't know what the future holds for me, but I'm like itching because I need switches and wireless access points. I need all of it. You need to run wires too? No, because I've got the coax. The mocha thing. Yeah, yeah, exactly. So you already have the wires. I mean, it's not optimal, but it is more than sufficient for my needs.

Uh, so anyways, so I go to install it and, uh, what I did, which I don't think was the right choice in retrospect, what I did was I just pulled the ethernet cord that goes into the Eero that comes out of the ONT, the fiber optic, you know, ONT. Uh, I pulled that out of the Eero and I plugged it into the 10 gigabit port on the, uh, cloud gateway fiber and basically told it go. And, uh,

It was fine. Like it did set itself up okay. And it seemed to be working. And let me tell you, there's a lot of switches to fiddle with and it is reasonably user-friendly. I am not a networking guru by any means, but it was reasonably straightforward to figure out. But then once I got it squared away and, you know, it was talking to the internet, which was also a nice surprise because my recollection of doing this with Facebook,

with Fios in years past was that you had to like go on Verizon's website and tell them to like basically kick the ONT to allow a different Mac address to connect to it. It was like a whole nightmare years ago. And now it was, it didn't care. It could go back and forth between whatever didn't, didn't matter at all. Uh, I freaking love Verizon files. Uh, you can take it from my cold dead hands, but anyways, um,

But I did notice after connecting my laptop physically to the router or to the Cloud Gateway Fiber, first of all, depending on what I was doing, my laptop was getting 100 megabits per second. And I've narrowed that down to I believe that was a bad old Ethernet cable. So that was my own fault.

but consistently I was only getting 250 megabits upstream. Now I have a symmetric gigabit connection and consistently I was topping out at 250 megabits as though I was being speed limited somewhere for life. I couldn't figure this out. And eventually I did enough tagging and Googling and duck, duck going and whatnot that somebody had said, pull the change, the WAN connection to,

from the 10 gigabit ethernet jacks, receptacle, whatever you want to call it, a port, there's what I'm looking for. Change it to one of the other ports and plug in the WAN there. Now, there's a couple of things here. First of all, how freaking cool is it that you can tell the ubiquity, I would like the internet to come in over there, please.

rather than on this port. I would like it on this other port. And it's very straightforward to do it. So before I actually committed to moving the port, because the management is mostly via the web, and it appears to me to be via Ubiquiti's website, I guess proxies or what have you, or establishes some connection between your local device. You can log into your router from anywhere and manage it from anywhere.

Which is mostly nice, but a little unnerving in a case like this where, you know, what if I lose the connection to the internet? But anyways, so I at least did that correctly. So I moved... Oh, wait. And by the way, to be clear, you can also do it locally. Like, it also works locally. And by the way, another thing you could try that you probably didn't immediately assume would be very good, the iOS app. I did try that. It's pretty good. And it can also... The iOS app can connect directly via Bluetooth to most of the new routers.

Oh, hey, you can log into this locally. Who knew? Yeah, it's great. So basically, whatever connectivity is available to get to it, it will use. Good deal. So anyway, so I move the WAN port. I tell the management interface, I would like the WAN to come in on port one, reconnect it, and everything's right as rain. Good to go.

That all is well and good. The place where I think I screwed up is I did not tell the Eero to go into bridge mode before doing all this. I think in retrospect, what I should have done is before I pulled the internet connection away from the Eero, I should have said to the Eero, go into bridge mode and then basically immediately cut off the internet and throw it into the cloud gateway.

I didn't do that. And let me tell you, that Eero was frigging pissed because the only way you manage the Eero is through their iOS app. And it's a similar thing to what we were just describing where you go to the iOS app. The iOS app is communicating with Amazon servers. Amazon servers are communicating with the Eero.

It was a nightmare getting that squared away. Eventually, I was able to convince it to know, really, I need you in bridge mode now. Eventually, with enough patience, which is not something that comes naturally to me, it did seem to figure itself out and everything was all well and good. But it took like an hour or two for the Eero to get a grip. Now, again, some of that I think was my fault by perhaps choosing the wrong order of operations, but it did eventually get a grip.

So then I start doing, you know, poking about in the management interface. And one of the things it does is deep packet inspection. So it'll tell you not only what devices are using what, you know, so much and so on bandwidth, but also what applications. And I see that Netflix is like, and this is early on, you know, right within a couple hours of installing.

Netflix is like the third most used application. The kids were either not home or like not playing with their screens. We don't even have a frigging Netflix account. What kind of trash is this? Would you like to guess what the problem was?

Remember, I'm doing a lot of diagnostics while this is going on. Was Speedtest accessing Netflix? Fast.com, baby. You know who hosts that? Netflix. Oh, that's great. Sure enough. So it took me a minute, but I was like, why is Netflix being listed here? And it was like gigs upon gigs of data. And sure enough, it's accurate. It's because this dummy over here was going to Fast.com on the regular.

That's awesome.

And you can see the little dots flying around the graph of the topology. So you can see the dots going from Verizon to the Cloud Gateway Fiber and then flowing into my MacBook or whatever the case may be. I feel like you're watching the actual audio move and audio hijack when you see the little dots move. Basically. Yeah, basically. It's about the same thing.

It's incredible. The problem, though, is that I have something like 55 devices connected to the Cloud Gateway Fiber, and they're all flat because they're all either on a non-ubiquity switch or they're hanging off the Eero's. And so as far as the Cloud Gateway Fiber is concerned, that's just one flat network.

What I got to get in my life is a switch for the office. That's a unified switch. So that'll separate things out. I got to get some unified access points. So if you're looking to get rid of somewhat modern unified equipment on the cheap, let me know. I might be. I know. Well,

I'm not kidding. I would love some semi-model. I don't want a U3 if such a thing exists. I'd like a 6 or 7 or something like that. But if you, the listener, happen to be wanting to get rid of something, do let me know. Marco, if you would like to send me something and pad it with Ubiquiti access points a la Kindles, do let me know. I will say, too, when you are looking at that kind of switches and stuff, like

I suggest getting one really good switch. Have that be like your main switch, but then like your like leaf node switches, so to speak, like the ones that you put like with your TV and stuff, those can be the cheap like little flex switches that are just PoE powered and have like, you know, maybe four or five ports on them and usually cost like $4.

40 bucks like they those can be really inexpensive switches because they you would still get all of that management and introspection features you know of ubiquity you still get those with their little tiny cheap switches too you know they don't have as much like throughput and they don't have as many features but like all of the stuff that you're talking about you just need it to be a ubiquity switch they're all like what we used to call managed switches back in the day they're all managed switches even the cheap ones

So you don't need every switch in your network to be some $300 powerhouse. You can just have one good one and kind of fueling them all, and that's kind of all you need.

Yeah. So I'll put in the Slack and if I remember, I'll put in the show notes, but to be honest, I'll probably forget. And there's not much to look at, but if you look at my topology, it's, it's a frigging mess because it's just, everything is microscopic because it's one, you know, exceedingly wide list of 50 some devices. And it's just, I cannot, I cannot have this.

But no, the equipment is incredibly cool. There's so many things to fiddle with, but I don't feel like I've had to fiddle with them for the most part outside of this WAN problem, which admittedly shouldn't be a thing. Like it should work just fine the way I initially started, but at least I had an easy solution and I could take care of that on my own.

But the deep packet inspection is cool. The graphs are cool. Not only just the internet traffic, but like there's all sorts of different graphs and things you can look at. When you go to the devices, instead of like some very good looking, but kind of funny, like line art that the, excuse me, that Eero uses, you get like full color pictures that are like accurate to all of these random devices that I have in my house. Like they have the pictures of each individual sonology and my,

mostly auto discovered the right one. Like it's incredible. It's if you're, if you're the kind of person that wants to like tweak and fiddle with knobs and just, just play around. Oh, it's so great. But now I need all the ubiquity equipment. And, and another time I'd like to talk to you about how I should lay this out. Um, which speaking of, I don't want to belabor the point,

And I do think we need to move on. But one of the things that somebody might have been Steven pointed out to me is there. If you go to design.ui.com, you have to upload something as a floor plan, but it can literally be a blank image, but

You can draw out your floor plan, which I have done, and you can draw out, if you set a scale, so this wall is 20 feet or whatever the case may be, you can draw out a floor plan of your house. And I found an app called, what is it called? It's called Magic Plan, all one word, where it's really designed for professionals. But if you want to do a single place, you know, like your house and nothing else, you

then it's free to do that. But once you start adding more places, you have to pay for it. It's pretty expensive because it's mostly a professional tool. But I use that to LIDAR scan the house. And there were a couple of small problems with it, but for the most part, it was great and took a lot less time than, you know, measuring everything and so on and so forth. And it's,

probably not perfectly accurate, but for the purposes that I'm trying to accomplish, it's more than enough. So I uploaded that to... You don't have a floor plan to your house? We never had a good one when we bought it back in 2008, and I'd like to get another one done at some point, but I'm both too cheap and too lazy. How could you buy a house without seeing a floor plan? What's going on down there, Virginia? No, there is a floor plan, but it's awful.

It's truly terrible. It's not good at all. And I have, I like did web archives of the listing for the house when it was up for sale from 2008. I can still get to them right now because guess where they are, gentlemen? On the Synology. And anyways, but the floor plans are trash. But the point is you can drop

uh, access points, wireless access points. And if you tell this design tool, what the material of your walls is, is it metal? Is it wood? Is it drywall, et cetera. It will make a colorful map of approximately what your wifi coverage will be based on where you're placing what items. Again, I'm going to put a link or I put something in Slack. I'm,

I don't think this is going to go in the show notes for sure. And in fact, neither of these links might, or pictures might go in the show notes because it's a little bit of, don't be creepy, but imagine there's a floor plan of the house and you're seeing like green and yellow and red and so on and so forth. And the same thing with cameras, you point the camera and it'll show you the approximate coverage area of the cameras. And so I don't have any cameras yet, but that's one of the things I do plan to do. Eventually this stuff is incredible. And this is all for free on the design app, but basically,

Don't ever look at this. This is my tip to you. Unless Ubiquiti wants the sponsors, never ever get their equipment because you'll send all of your money to Ubiquiti. So just be blissfully ignorant. I made just a terrible mistake by signing myself up for this. Be blissfully ignorant. Learn from my mistakes. Don't ever look at anything that Ubiquiti does because it's all so cool. That's awesome. All right, John, you've been busy. Tell me what's going on with hyperspace.

Just to remind everybody or people who are just listening to the first time, I recently launched an app called Hyperspace. The deal with that app, I need to come up with like a snappy elevator pitch or whatever, but it's basically you can get back some disk space without deleting anything. If you are a digital hoarder like me, you want to, you know, have as much free disk space as possible, but you don't want to delete anything. And there's tons of apps out there that will like

find duplicate files and then tell you to pick which ones you want to delete. But that's like picking among my children. I don't want to delete any of my files. I just want the disk space back. So hyperspace does that for you because it finds files that have the same contents and it makes them so they're sharing a single instance of those contents on disk instead of each having their own private copy.

That's what the app does. And I talked about it on the show. Before I even had a name, I talked about it on the show that I'm thinking of making this app. And the title of that episode was an incredibly dangerous app because it is, in the end, an app that modifies a bunch of people's files. And it doesn't know what those files are. It's like those files were created by other apps. Those files have your data in them. And here comes my app, this app that knows nothing about them and says, yeah, I'm going to mess with those files. So incredibly dangerous app. And when I launched it in 1.0,

I launched it with a very safe feature set. I made it not do a lot, like a lot of the most difficult things that it could possibly do. It just didn't do them in 1.0. I knew there was a potential risk in terms of like people are going to get the app and be like,

you know, I, all my files are in iCloud drive. This ignores them one star. And that happened, right? But that's, I knew that was going to happen, but my plan was always launch with a safe feature set, get the thing out there to a bunch of users, because even though I had a really good, really big beta test with, with ATP members, in fact, as part of a,

A membership promotion thing we were doing, if you're an ATP member, you get to be on the beta and a bunch of people did. There is nothing like giving the app to regular customers who are not listening to a tech podcast. You get so much better results and feedback and everything from that than you do with the beta test. Yes, that's true. Let me jump in here, though. One of the best decisions I think I made for CallSheet, and I've said this before, but I want to say it again, was opening that up to the ATP members made CallSheet super.

so much better. And I suspect that it was the same with hyperspace. And even though, you know, this was the perfect, like I'll scratch your back, you'll scratch mine kind of situation. Cause I wanted to juice membership full stop. You know, that's what I was after. And I wanted my app to get better. And then the members got access to something that was kind of not illicit, but like cool and hidden and secret. And I, I am so,

so glad that I did that. And I was not convinced I had made the right choice when I opened it up to ATP members. And I am so very glad and so very thankful for anyone who spent time looking at it and issuing bug reports and so on. So if there is another app, which is sitting here now, I have zero plans for anything else at the moment. I plan to do the same thing again, because for me, it was incredibly helpful. And I just want to thank the members that did participate one more time because it was great.

Yeah, like I said when I launched the app, the beta was great. But there's a difference between beta testers and regular users, especially for an app like mine, because an app like mine doesn't exist just on its own with its own data or with, like in your case, with the movie database behind it or whatever. It acts on people's data. And if you're a beta tester, you very quickly learn, well, I kind of have to either, like...

have it configured not to really do it or else I'll get rid of all my duplicates and I can't really be a good beta tester anymore. So most beta testers, I imagine, were rerunning it in the mode where it doesn't really do the final step of the thing and they're rerunning on the same set of data over and over again. The variable that I needed in the mix here was regular users using it on their real data for real.

And you just need real users to do that because you don't want testers to use it on their real data for real. Because once they do, they're like, well, I can no longer test your app unless I like somehow manufacture data. We haven't asked it to be a question for a future episode. People wondering how that worked. But anyway, that was the key. I it's data based. I need people to run it on their data to find out where all the problems are on their real data and their real experience. It's real customers hands. Uh, and, uh,

I did a blog post about this because recently I rolled out a version 1.3 of hyperspace and 1.3 marks the end of a journey that started with 1.0 at launch, where I was slowly adding back all the features that I didn't launch with. Well,

I say adding back because basically I launched with restrictions. Like when I was running it locally, I could just turn off all the restrictions and it could just have access to everything. So I knew that there were problems with certain classes of things. So the three big classes of things that I described in my blog post was number one, packages, which is the thing most people don't even know about, but I try to explain in the documentation. Basically, it's a folder that looks like a file. They're all over the place in macOS. Applications are packages. They are, you know, whatever mail.app. That's a folder.

It's got a bunch of stuff in it. I remember talking about it in my Mac OS X reviews back in the day. This is technology from the next world. It's a folder that looks like a file, and inside that folder are a bunch of other files, and certain packages have certain structure. There is a difference between bundle and a package, and it's too esoteric to get into. It's not that important. But the point is, lots of stuff in Mac OS, they look like files, but they're not.

They're folders with stuff in them. Photo libraries. Your Apple Photos library, it's called... It's like a .photos library file name extension. Guess what? That's a folder. All the stuff is inside there. It's a massive folder with millions of files inside it if you have a big library. .rtfd files that you save in text edit folders. So, packages. Does that read the effing document? That's a... I believe... I wrote it in the documentation. It's like a rich text format with...

Attachments? I don't know. That makes sense. With attachments. It's not the RTFM document, although that's probably taken. Anyway, I was avoiding those. But because, you know, well, messing around in people's photo libraries is dangerous. And also because lots of people don't know that they're folders. They don't know you can right-click them and do show package contents. They're not used to, as I say in the blog post, cracking them open and digging around in the guts. So if anything went wrong...

with reclaiming one of these files and my error message directs them into one of their packages, they'd be like, "What the heck is this? I'm inside this thing that I didn't even know was a folder?" Because the finder will do it. You don't need to use terminal or anything. You can do it in the finder, but it's not something people are used to. So I wanted to wait until I got most of the major problems worked out

not doing packages. So version 1.1, that was about a little less than a month after launch, I added support for packages. And that was after, I don't know, four or five or six releases in the 1.0.x series. Five releases. I did five 1.0.x releases before I was ready for 1.1. So I did package support. The next big one was cloud storage.

everybody who had all their folders and iCloud drive was like, this app is useless to me. Uh, it doesn't do a cloud storage. I mean, they haven't paid for it at that point. So it's good that they didn't have to do that, but they were still mad. They're like, I was excited to use this app. And it turns out it won't even tell me how much space I can save because it totally ignores cloud storage. I had to tackle cloud storage. Cloud storage is not like normal files. They're weird. And I had to deal with them and required using a bunch of Apple's APIs, which actually went better than I expected, but it was still tricky. So version 1.2, uh, in April, uh,

supported cloud storage that includes iCloud Drive, some versions of Dropbox, OneDrive, all the other things that use the file provider extension. And then finally, in version 1.3, which is the very end of April, I supported library directories. It

In your home directory, you'll see a directory called library, which I think is hidden by default. So maybe you'll only see it if you unhide it by switching the flag from the terminal or if you just look in terminals there. Anyway, the reason Apple hid the library directory is because for years, people would go on a Mac online

open their home directory, see a folder called library. I'm like, what the hell is this? And they would just like either start rummaging around in it and throwing stuff out, or they would try to delete the whole library folder. Don't do that. You need that folder. It's super duper important. I think it's actually protected now in addition to being hidden. I think you can't actually delete it now, but you can delete stuff inside it. Anyway,

don't, don't delete stuff in there. It's super important, but I had been avoiding that directory and the library directory. This is the top level of your boot volume because there's, there's super important stuff in there and it changes really often. Uh, any running application or any running background process is very likely messing with stuff inside your library folder. Uh,

and there's a huge amount of stuff in there. And if you screw up the stuff in there or corrupted or damage it, applications could start crashing or have problems. And you wouldn't understand why, because people don't even know what's in there. They're not supposed to know it's private stuff that belongs to individual applications, your preferences, cash files, the, you know, the container directories for the sandbox documents. It's,

you know, there's lots of sort of, you know, non-user serviceable parts inside library directories. That was the one I saved for last, even though I think it is one of the biggest sources of data that you can get back. And so in version 1.3, I supported libraries. All these things, packages, cloud stories, library storage, and libraries are,

they're all disabled by default in settings. So I'm probably going to have to add a fact item that says, hey, it didn't find a lot of data. What can I do about that? And I say, well, if you go to settings, you can turn this switch from this switch to this switch. You know, every one of these switches, even though the version 1.3 has all these features and we'll do all of them, they're off by default because I want the default experience still to be the safest for people who don't know the details. But if you do know the details, if you're the type of person who immediately hits command comma to see what the settings are,

turn on the settings that you want i even put in settings for like you can go inside packages but you can refuse to reclaim from inside them instead just use them as the source for reclaiming stuff outside of them um yeah it's it's still extremely conservative maybe i'll change my mind about that going forward but it's an important milestone uh version 1.3 doesn't seem like a big deal but this was always my plan from the beginning 1.0 1.1 1.2 1.3

I'm not going to say feature complete because I've still got more stuff on my to-do list and more features that I'm going to add, but this is what I had in mind as like,

as I said at the end of this. This is kind of what 1.0 quote-unquote should have been, but it was, I would feel, I wouldn't have felt as safe doing 1.0 with this feature set, did it? And in retrospect, I'm glad I didn't, because boy, I fixed so many bugs. At every stage, I was fixing bugs, and all of those bugs contributed to like the most dangerous thing, the library directory being as safe as it is in 1.3. So far, I haven't heard any disasters or complaints, so I figure

Fingers crossed. I think things are going well. If you tried hyperspace and it didn't do anything for you, check it out now. Also remember, I'll probably add this to the FAG as well. The file type list is also very conservative. You can probably safely at this point with version 1.3 tell it to allow any file in the file type list and so it won't filter by file type. That'll make your scans a little bit faster and also it'll probably find a lot more data. So yeah, toggle on packages, toggle on cloud storage, and if you're really brave, toggle on libraries and

and see how much more data you can find. I'm really glad that you haven't had the, you know, oh, I've lost all my family photos and it's all your fault feedback as yet. Yeah, like I said, mostly just got the one-star reviews, which I'm glad the developer replies were in there because they're like, you know, one star doesn't do photo libraries. And then I would release version 1.1. I would say, oh, I just released version 1.1 and it does support photo libraries. And you're just sitting there waiting and you're just hoping the person will notice that

They gave a one-star review for a feature that your program didn't have, and then a week later you added that feature, and you're like, come on, notice the review, revise your review. Sometimes they do, sometimes they don't. Same thing, one-star, no support for cloud storage, developer replies, it now supports cloud storage. Then you're waiting, come on, notice that I replied. You know, you win some, you lose some. That's the downside to launching conservative. But so far, no one has said you deleted all my data. Like, programs have bugs, it could happen, but, you know, so far, so good.

That's awesome. So I've been doing a little bit of vibe coding recently and speaking of bugs. Yeah. Speaking of bugs. So this all started with me wanting to post a GIF on blue sky. And I don't even remember what it was, to be honest. No, I think I was, I was trying to reply to a post that Jamel Bowie had written. Um, he, he is just so freaking smart and so cool, but that's neither here nor there. Um, anyways, I was trying to reply to a post of his, I think it was something like, um, you know,

name a movie you've seen or reply with a movie you've seen a million times. And, and, you know, everyone was replying with gifts. And so of course I was going to post a hunt for an October gift. I forget exactly what it was. It's too long ago to dig up, but anyways, I wanted to reply with a gift. And so I went to my favorite, you know, gift repository app, which is written by my friend jelly. And I went to gift wrapped and I copied the gift, the actual image. And I went to paste it in the blue sky app on my phone and nothing happened.

All right. Well, maybe blue sky doesn't want, doesn't want to take a gift either from gift wrapped or just in general. I know what I'll do is I'll go to gift wrapped and I'll copy the URL and I'll put that in. And I tried that and it still wasn't quite right. And I was very annoyed at this point. And I didn't know if maybe it was because of some like weird gift wrapped URL or something like that. So anyway,

I realized what I want is, or I felt like I wanted at the time is I want to be able to put a repository of all my gifts on some publicly accessible web web page so that I could just grab a public, you know, world accessible URL for them at any point. My friend Steve has something like this on his own website that he uses from time to time. I want to do that too. Now, none of this is conceptually challenging, right? None of it is difficult. Um,

And now that I've had, now I've added another new shiny hammer to my tool, to my tool belt, which is Cloudflare and Cloudflare pages, because that's what I used for the call sheet web presence. So I thought, well, maybe I could do that. And so how do I, what is the best, most reasonable way to do that? Well, I can, you know, put a bunch of gifts in a folder, put that on GitHub and connect Cloudflare to that GitHub account or GitHub repository. Maybe that would work. And then I thought, well, but,

you know, I need to, apparently there's no mechanism by which I can, or I can parse out or read a list of files in the file system using any of the like Cloudflare APIs. I'm sure if I like went full on node or something like that, I could do it. But you know, for just when I'm trying to just whip this together real fast, um,

There was no obvious answer to that. And honestly, it doesn't, if you know an answer, it doesn't matter. It's just bear with me. I'm just telling the story. So, uh, so I tried doing that and that didn't really work out. And so I was like, all right, well, maybe what I can do is, well, I don't know what I should do. What can I do about this? So I thought, you know, let me ask chat GPT, see what it says. And, uh,

around the same time I was thinking of it, I had this like nebulous thought in my mind and chat GPT said, Oh, here's what you can do. Just make a list of all of the gifts and store that as like Jason or something like that. I was like, you know what? That's a great idea. I'll do that. All right. So can you write me a little script or something? I forget exactly. This is like a week and a half ago and it,

And the particulars don't really matter, but, you know, I can figure out some way that you can do that. And can I run that as like a pre-build step or something like that? And it said, well, I don't know about that. Oh, wait, no, I know what I can do. What if I did it with GitHub workflow? And it said, yes, you could do that with GitHub workflow. Let's work on this together. And so, you know, I'm writing, I'm vibe coding my way through this. Well, here's what I want to do, blah, blah, blah. Okay, we'll try this. Oh, that didn't work. And here's why, blah, blah, blah. And so we went back and forth. And then finally I got to the position that, oh, I've got all of my GIFs in Wordpress.

Cloudflare, publicly accessible if I want it. But hey, since we're already vibe coding my way through this entire self-assigned project, what if I had it generate some sort of like web front end where it's just put all the GIFs on one web page. It's like 300 GIFs. It's a terrible decision. I fully recognize that's a terrible decision, but who cares? This is only for me. Can you do it? And sure enough, it did. And then said, hey, I can make it better in such and such way. Okay, sure, do it.

Okay, I can make it better in this other way. Sure. Sounds great. Within the span of a couple of hours, which granted, maybe I could have done this in less time by myself. I don't know. But in a couple of hours, I had a pretty decent looking web-based GIF repository and it was all done via vibe coding.

I am capable. I genuinely believe in my heart. I am capable of doing this by hand 100%, but I also genuinely believe that it would have taken a lot longer for me to do it because I'm so far out of the web game at this folks at this point, folks like I'm, I can do web stuff, but I'm just, it's just not my bag anymore. And I think Marco, you're right there with me on that one. And so, and so it's just not something that really revs my engine. I wanted the destination. I didn't really care that much about the journey. And it turns out in a couple hours with chat GPT, I got there and,

None of this story is that remarkable in and of itself, but this was the first time for me that I just sat there and pinged away at chat GPT and together we fixed a problem. And let me tell you,

There are only a handful of times that I really get that jolt of feeling like I've got coworkers again in the good way. Obviously, there's coworkers in the bad way. I'm talking about in the good way. That's us. Yeah. No, no, not at all. Not at all. But no, it was not as good as having an actual coworker. But let me tell you, for someone who works independently 95% of my – well, 95%.

not in this context, but in terms of like programming, I've worked independently almost entirely. And it was so great to like have a coworker for a minute and it really worked out well. And it was a resounding success. Now, and we can come back to in just a second, but in contrast to that,

I have started down the path of adding just a link to Letterboxd in, or I'm sorry, John, Letterboxd in Callsheet. So it's one of the quick actions. You can just hop over to Letterboxd and it'll open that film in Letterboxd directly. So you don't have to search for it over there or anything like that.

Well, the mechanism by which I have all these links, it's a SF symbol and it's a bit of text. And there is no SF symbol for letterboxed. Of course not. There is an envelope, but there's no three overlapping circles, which is the letterboxed logo. There's no SF symbol for that. So I dug up, I think I might've asked it to just make one and it

didn't work at all and then i dug up a copy of letterbox uh an svg of their logo just a straight up svg and i fed that into chat gpt and it tried and tried and tried and tried to try to get that into be an sf symbol and it took for freaking ever to get it to the point that the sf symbols app would even import what it created and i'm continually feeding it oh it can't find the margins it can't find the margins you can't do this can't you just do this in a straight swift ui

It's such a simple logo, you know? Probably, but I needed to, because of other uninteresting reasons, it would be incredibly preferred for it to be an SF symbol. It can be a custom SF symbol. I have other ones like that, but an SF symbol nonetheless. And for uninteresting, complicated reasons. So I tried so hard to get it to work. Eventually it did import into the SF symbols app, but what I saw was blank, blank.

So like it sucked in whatever this, you know, bespoke SF SVG was, but it doesn't appear to be anything. And I haven't gone back to try this again in the last couple of days because I've been busy with other things. In fact, I've been busy with peak of you, which we're going to come back to in a moment. Um,

But for the life of me, I haven't been able to do it. And the other thing that's a really chap in my butt is because I'm cheap and I don't want to pay the 20 bucks a month in perpetuity. You got to pay for ChatGPT, man. It's worth it. I think I do. If I continue to use it like this, I absolutely will. I started paying for it before I started Hyperspace, though.

Yeah. I'll see. There you go. So anyway, so I think that might be in my future is paying for it. Let me tell you, everybody out there, like if you're going to spend a bit of money on subscriptions in the tech world, there are two subscriptions that you need to buy before everything else. YouTube with no ads and chat GPT is unlimited plan. Okay, stop. That's not true. The zeroth element is ATP. FM slash join. Okay. Yes, you're right. Three things.

Actually, it's funny you bring that up because you know what? I've been on the precipice of engaging on, even though I'm so cheap and I really don't want to do it. But I am on the precipice of doing YouTube without ads because they have the new cheaper plan. You don't have it yet? What? You're not doing YouTube without ads? No. Oh, my God. I know. I think I'm really playing myself here. Don't buy any Ubiquiti switches until you buy that. Seriously. Seriously. Bang for your buck. Quality of life improvement for YouTube. I know.

You have to, especially your kids. If they're not already watching YouTube, they will be soon. Oh, no, they are. You've got to turn off. You've got to get rid of those ads. Oh, yeah. You don't know what's already done to them. My daughter knows what the Geico gecko is for like the one week before I turned off ads on YouTube when she was like five or whatever. Well, the thing is, I really played myself because I didn't turn it off or I didn't subscribe or whatever during election season. And that was a

entire mistake. Oh no, you got to turn this off now. I feel like making you stop the podcast and pay for you. Madness. All right. Anyway, but focus, focus, focus. So part of the problem though, is because I haven't yet paid for it is I have to wait like a day between every three attempts because it's like, dude, I ran out of compute on this one. You're gonna have to try later. Well, because I don't think I'm, I genuinely don't think the 20 bucks will get me over the edge. I don't think it's capable of solving this. Oh no, it probably won't help you with this here, but like it's, it's,

It's absolutely worth it. Yeah. When you have something that's very limited like that, you avoid using it because you don't want to spend it or you don't want to hit the limits. Right, right, right. And you need to fail a lot with these things. That is also very true. Yes. But once something is unlimited, you start finding more opportunities to use it and the value of it to your life increases by usually a pretty good amount. So this is the kind of thing where if you ever have any reason to ask ChatGPT things...

You should probably pay for the plan because then you can use it in all sorts of ways that you don't think of. Yeah, that's very fair. I'm probably sure. Or if you're me, you pay for multiple ones so you can pit them against each other. Well, actually, very quickly, I tried all this in Claude as well, and it did a smidgen better, but not really that much better. Yeah, so you can watch multiple LMs fall on their faces. Exactly. Exactly.

So as a final note on this, if you, the listener, are capable of making a approximately square aspect ratio letterboxd SF symbol, send it my way and stickers will be returned. Isn't there a WWDC session on doing custom SF symbols? I am incapable. That's literally three circles. I don't think you can just like. I am incapable. There's no chance I'll do it. Watch the WWDC session, draw three circles, fill them with the color. I genuinely don't think I'm capable of that.

Plus, I don't have Illustrator and all the other stuff that you need. I think Affinity Designer is a free trial. I don't know. I bet you can find a free vector tool to make three circles. You could use, what do you call it, that paint code thing that Marco likes? Yeah, which is not free, though. I'm sure there are free options. It's fine. But anyway. The real thesis of this whole story, though, is that

If it's not something that you really, really, really care about, like this GIF repository that was just me goofing off and this SF symbol thing, I do really care about it, but it either would have worked or didn't, right? Like it's not a do or die thing. It has been real nice for those sorts of scenarios. I've been really surprised how much I enjoyed just vibe coding my way through this little GIF project that I self-assigned. And even though it was frustrating, a lot of that obviously is my own fault, but

But that's a simple thing. Like, it definitely was...

But getting there, I just don't think I'll ever get it quite over the hump. Depending if you are a vibe coding prescriptivist or descriptivist, I've heard lots of people push back against our past discussion and other discussions that don't involve us at all about what actually is vibe coding. Is what Casey just described, is that vibe coding? The stricter definitions I've seen would say this does not qualify. What this is just is, is you using ChatGPT to help you code. And you'd be like, well, no, but I didn't write any of it. It wrote it all itself. Isn't that vibe coding?

I think like the person who coined the phrase has said something about like, you have to not care about the result, uh, whether the result works or not. And you do care about it. Like that's, that's, that's all, maybe all you care about is whether it works or not. But anyway, if you're about to email us and tell us that what Casey did, isn't really vibe coding. I would say that I think the phrase vibe coding is currently in the early stages of evolving, uh,

and like so many things in the world, what the inventor of the phrase wants it to mean is not necessarily what it will come to mean. So Casey already thinks that what he just described is vibe coding. I'm not entirely sure if you disagree, I would just say, I think this is, this is a fluid situation as they say. And,

vibe coding will mean what it means. Either we'll fade from a vocabulary and we will laugh at ourselves for using this phrase in a few years or we'll mutate into a totally different definition. Yeah, hasn't it only existed for like three weeks as a term? Yeah, seriously. It's pretty new, but already people are saying that's not vibe coding. So I just wanted to head that off. Yeah, I got news for you, everybody in the world. You don't get to control what terms mean. And once people start using it to mean a certain thing, you've lost control. And that's just what it means. Like, that's just how language works.

margo the descriptivist all right so uh we are running long so we should probably cut this off but i just want to say uh for next week on the docket already is the backblaze stuff so we aren't going to talk about it this week we've run too long we will be talking about it next week so don't worry we'll get there so please email only casey about the backblaze stuff in the meantime the opposite of what i want damn it

All right. Thank you to our sponsors this week, Squarespace and Delete Me. And thank you to our members who support us directly. You can join us at atp.fm slash join. One of the many perks of membership.

is ATP Overtime, our weekly bonus topic. In addition to all the other exclusive member content, you get a weekly bonus topic every single episode. This week's Overtime, we're going to be talking about rumors of a Vision Pro successor or successors, including one rumor for a Mac tethered product. We'll be talking about that in Overtime this week. Join now to listen at ATP.fm slash join. Thanks, everybody. And we'll talk to you next week.

Now the show is over. They didn't even mean to begin. Cause it was accidental. Oh, it was accidental. John didn't do any research. Marco and Casey wouldn't let him. Cause it was accidental. Oh, it was accidental. And you can find the show notes at ATP.FM.

And if you're into mastodon, you can follow them.

C-A-S-E-Y-L-I-S-S, so that's Casey Liss. M-A-R-C-O-A-R-M, Auntie Marco Arman. S-I-R-A-C-U-S-A-C-R-O-Q-U-S-A. It's accidental. They didn't mean to. Accidental. Tech Podcast So Long.

So Casey, I actually did my own vibe coding recently, I think. Oh. Maybe we'll see if it qualifies. So I mentioned when talking about my e-ink tablet journey between the Remarkable 2 and the SuperNote.

I mentioned that I wished that the Remarkable had links the way the SuperNote can like you can like draw something on a SuperNote note and just like lasso it and have it link to a different page or a different note or even a web URL if you want. But I wanted to use that to interlink between my major notes of personal restaurant and overcast. Like those are my major areas of notes.

and rather than having them to jump all the way back to the document picker or to swipe between different pages, I wanted just buttons at the top of each screen. Well, I learned that there's this whole community of remarkable templates that you can buy that people will sell you if they have some kind of system, like a journaling system. Okay, buy these PDFs and they interlink within themselves so you can jump to different sections and use my journaling system. But they're just PDFs.

with links inside of them. So it turns out the Remarkable supports PDF links, so two different pages within the same document. That is the only way they're supported. They're supported only by touch, not by hitting them with a pen, but they are indeed supported in that way, within a PDF. So I thought, hmm, what if I make a PDF document

template for my notes that has the major areas like as buttons across the top. Oh, that's very clever. And it would just be blank. The whole rest of the document would be blank because you don't have a bullet journal system or a theme system journal or anything like that. It's just an empty page. Right. Cause I only wanted like basically two, one or two pages per thing. Plus like a scratch pad with a few pages on that one or two pages. Wait, I thought it was the whole deal was it just has to be one page.

occasionally I need to overflow for like a subcategory. That's the thing. There's a little page inflation happening here. The stuff on the second page, you're never going to see it. I know. Well, so what I'm using it for is like... So, you know, in the context of the restaurant, we have a subproject of the stockroom, making a stockroom. And it's like, okay, well...

that that has like you know its own set of 10 things so on page one the main page i have stock room and then i have an arrow pointing off to off the right margin to say next page and then on that page is all the stock room specific stuff basically a sub project and by the way if anyone for remarkable is listening what you should do is make it so that when you underline some text it you a thing pops up that says do you want to make this a link that would be cool i mean maybe just any kind of any kind of

user-creatable link on device would be amazing pretty much in any form it takes. Anyway, so I figured, okay, what if I make a PDF? How do I do this? So I'll send you a picture of... Here, I'll put this in our Slack here. It's full of stuff, so I'm not going to share it publicly. But this is a picture of...

what i have created you can see across the top there are four like icons that i have drawn oh i love that little stick person for personal restaurant symbol overcast symbol and scratch pad symbols what you're seeing here is one of the scratch pad pages i love this i know it's supposed to be don't swing the lights but it looks a little bit like a gallows or something well i know what it means this is for me maybe maybe you have a tiff to give you a helping hand with some of these uh

The artwork or maybe have her help draw the lines a little bit straighter on your list. Well, I feel like it's actually like I like the slightly imperfect look of handwriting. That's why like slightly in my overcast icon. I could have just imported the overcast icon perfectly, but I didn't. I drew it by hand and I didn't even use the Remarkables like perfect circle methods. I just drew a freehand circle that was not a perfect circle because I just like the look better in this context.

Anyway, so I'm trying to figure out, okay, how do I create? I know PDFs can have links. How do I make one?

So first I go to chat GPT and I say what tools can do this and it's a bunch of stuff I haven't heard of and Acrobat. I'm like well I have Creative Cloud. I install Acrobat and I spent a good half hour trying and I kept going back to chat GPT. You're XY probing on this one already because you're thinking like a programmer being like what tools can do this thing that I need to do instead of thinking like a non-programmer and saying make me a PDF with links in it. Well I actually never thought to do quite that but

But I did for a while. I'm like, you know, trying to get ChatGPT to talk me through how the heck to use Acrobat. What a weird app that is. Didn't you feel good about installing that on your computer? Oh, God. I saw it on my laptop. Like, I didn't want it on my main computer. Yeah, you have to isolate. You have to isolate it in Acrobat. Exactly. Yeah.

Anyway, so eventually I'm like, I just give up on Acrobat. This is taking so long. Trying to copy and paste the links between pages sort of worked, but not quite. And it was a mess. It was just a disaster. And then I realized, wait a minute. I'm a programmer. Here, I'm going to send you a screenshot here. Do you remember what this language is? It's a name I haven't heard in a long, long time. Oh, God. Is that PostScript?

John? I don't know. It's not ringing any bells. I mean, I thought you were about to show me PostScript, but this doesn't look like PostScript. This is LaTeX.

uh yeah i've never written that by hand oh gracious so i i'm like wait a minute that's like programming language that i can use to create pdfs you know you just got done uh complaining oh i don't know web stuff or whatever but as people in the chat room have pointed out if you just made an html page with links and then printed it to pdf that might have worked too would it

I don't know. It's a theory. I don't think so. It would have been a quick thing to try because I'm pretty sure you still remember enough HTML to make a document with a link in it. I mean, maybe, but who knows with all these modern box model stuff that we have. I honestly think you should have asked ChatGPT to make me a PDF with four links at the top called this with these icons. Well, for some reason, I didn't think to try that. But I did say, here, my initial prompt was generate LaTeX markup for a page with a row of four square round rects in the corner of 0.5 inches per side.

it did it and it was big it worked like it and I didn't ask it for links yet I'm like let me just see if it can do like just the layout and then I'll figure out how to do the links and it did it and it worked and

And then, of course, I asked, like, by the way, how do I install this on a Mac? Like, what packages do I need? Did it tell you to use Homebrew? No. This is too old for Homebrew. I mean, I haven't seen LaTeX since college. Oh, it's still in use. Believe me. Oh, I'm sure. But, like, it's like, you know, if you want a programming language to lay out a page. That's where you'll see it. You'll first see it in college. That's for sure. Yeah.

yeah but there's it turns out you can go to like mac tech and there is a six gig package you can install that will do this and it's like okay great i installed it so and then you know i had to figure out like how to get the links in and what's you know when you're using chat gpt to generate code for you you can tell it like in this case i i had to tell it uh oh move into the top right corner like i they picked the wrong corner of it

And then, you know, down, down, down. And then I asked later on, like, how do I fill in the shapes and how do I make page breaks? Because I wanted this to be multiple pages. Okay. And then it told me, here's how you make the page. Here's how you can do it with a loop and everything. It was working. There was some place where the shading and it's like the example it gave me was like fill gray exclamation point 30. I asked, what's the numeric range on the gray? It gave me a whole table for all the different gray values and what they mean.

you know, go on, how to do links. Now at some point, eventually it created the, I had to create the links for me, but it wasn't aligned. It didn't, it wasn't actually working. The links were like off to the side. So I just said, the link is not aligned with the shape.

And then you see, ah, yes, it actually doesn't work. And so this is a pattern that, Casey, might be relevant to your SF symbols not working thing. A chat GPT will generate you code. If it doesn't work, you can just tell it in that thread, that doesn't work instead. It's giving this error or it's doing this bad behavior. And it will usually say something like,

oh, yeah, you're right. That doesn't work. Here's something that does. So, I mean, it's very agreeable in that way. But the sort of the death spiral pattern that you will come into if you do this a lot is it will say, oh, you're right. Here, try this approach instead. And then you'll say that approach doesn't work either. And I'll say, oh, you're right. That one doesn't work. Try this. And it will oscillate between the two non-working ones. No matter what you say, whenever you say it doesn't work, it'll say, here, try this. And you'll say, you just told me that. That one didn't work for me.

earlier do you remember that and it would be like oh yes you're right try this one back and forth a and b a and b and neither a nor b works uh yeah there's lots of there's lots of uh dead ends in these conversations but uh that's why that's why i enjoy having multiple ones of these going because like same prompt for multiple different lms and they will go off in wildly different directions often they'll all produce non-working stuff but sometimes one of them will go in a direction you find promising and then i just start over once like screen start a new thread with the new approach and then try them all at the same time it's lots of fun

Well, I tried. I eventually got what I wanted. However, let me show you the abomination that creates what I wanted. What I pasted to you now is a PHP script that I wrote to generate LaTeX code. Of course. Because I couldn't figure out how to do some of the loops and sets and counts correctly in LaTeX. And so I'm like, you know, I can just do this in PHP. You can brute force it, yeah. Yeah.

Inline everything. So I have possibly the world's only PHP script that generates LaTeX code. Oh, no, I guarantee this is not the world's only. PHP is generating LaTeX everywhere, believe me.

But hey, it was a crazy pile of hacks of ancient technology, but it works. And so now I have in my Remarkable a PDF with my little icons in the corner that I can tap to jump between types of notes. And it's not quite as good as having dynamically creatable links on the device. But in the meantime, until and unless Remarkable ever adds that, this is great. And all it took was a bunch of...

really obscure old technology combined with extremely new technology to tell me how to use it. That is wild. That's very cool. This is what I'm talking about. Obviously, AI training is its whole ball of wax that I don't at this very moment want to get into. But this sort of thing is very cool and very unlike anything that I've experienced in my career so far. It's very slick. Yeah. Again, I continue to think like,

The benefit to programmer productivity here is massive. This is like LLM-based code generation and completion and tweaking and stuff like that. It's on the level of...

Like a higher, it's like, it's like jumping two levels higher in high level languages. You know, like when we went from like, you know, assembly to C and then C to like, you know, memory managed languages. Like, you know, we had, we've had these jumps over time. This is like two of those jumps. Like it's, it's a, it's a, it's, you're still a programmer. You still have to be a programmer to really make much out of these results. But it's a leap forward in productivity in some cases and in many cases.

It isn't going to be the solution to every problem. And there's still many problems that we need to do ourselves, but this can give us a really big head start and a really big leg up and just can save us a bunch of time. The upcoming episode of Rectifs, I think that's the one. I talked a lot about this, talking about me using LMs for development. And one of the things I said there was like,

Even if it doesn't come up with a solution, which is very often the case because I'm challenging it very often to do things that are difficult or, you know, just that are hard for humans to do. And it's just failing utterly, even if it fails utterly over and over and over again, kind of like rubber duck debugging where just describing the problem to an inanimate object can help you solve a problem. I will usually by sort of exhausting the LLM's ability to do anything, uh,

Get new ideas about how to approach it, even if it's just like being satisfied that what I was trying the way I was trying to do is just never going to work or or I will suggest an approach that also won't work, but it'll give me an idea and.

It's not that you don't get these things from Googling, 'cause you do, and I would encourage people who do this, don't forget about Googling. Don't forget about looking at your reference documentation. This is not replacing all your existing tools, it's augmenting them. If you only use LLMs and stop using reference documentation and stop using Google, you'll be less productive, not more. But having the LLM in the mix, throwing out a bunch of essentially search results and various things and solutions that don't work, even when it makes up APIs that don't work,

It gives me like new things to Google, new vocabulary. Like if I don't know what word is used in this API for this type of thing and it throws out a fake API but uses the correct vocabulary for this particular verb or noun, now I have a better thing to Google or a better thing to ask the LLM about.

incredibly useful even when it just fails utterly which it does for me almost all the time because what i'm asking it to do is things that i don't have any idea how to do or that may even be impossible and it will gladly spew a bunch of crap and not lead to a solution like what you were doing marco was like a thing that somebody could do who knew the technologies involved i'm most often asking it to do things with language i already know with an api i'm somewhat familiar with uh

like subtle nuances of doing it in a particular, like I've already got something that works, but I think it could be done in a better way or whatever. And it just flails or whatever, but it gives me ideas like that. The rubber duck program is the best examples. You're like, well, I'm just stupid. How can they help you? Rubber ducks are inanimate objects. They're one of the most powerful tools program has at their disposal. Speaking of coworkers, Casey,

But back when you had coworkers, I'm sure you've used humans as rubber ducks where you go to someone's desk, describe the problem to them before they open their mouth. Because you walk through the problem describing to them, now you realize the solution and you just walk away. It's great. So if you think of LLMs as that but better, that's a good start.