694: Potential and Homework

13:01Marco Arment:really dumb but that's that's a gm thing yeah well and also like every few months somebody from rivian goes on a neil patel podcast and gives an interview and that just happened again um basically saying like we're never gonna do carplay why would we do carplay all we need to do is add a few integrations for things like spotify maybe or apple music and then it'll be fine and clearly

13:20John Siracusa:nobody would ever want to run overcast in the car. Why would you want to do that?

13:24Casey Liss:Well, it's just like this back and forth cadence of the same story, which is executive says we don't need CarPlay, and then whatever the most recent survey is, 80 % of customers said they won't buy a car without CarPlay, and it just goes back and forth and back and forth, and it's like, alright, well, I guess you're just going to wait it out and see if that number goes down.

13:39Marco Arment:Yeah, and Rivian's position has always been like, we don't want somebody taking over the whole experience. It's like, well, that's not really, like, almost no car lets carplay take over the whole experience um and there's like my car has like you know in the most recent interview whoever from from rivian was being interviewed by nila and i think it was decoder um said like that carplay requires every pixel of the screen no it doesn't most cars don't

14:04John Siracusa:use that does not i i got so mad about this i got so unnecessarily mad maybe they were talking

14:10Casey Liss:about carplay ultra which by the way the luce doesn't have that either which is interesting

14:13Marco Arment:Yeah, but look at what CarPlay means to everyone is it's a window on the screen that shows your phone stuff. And that's all it needs to be. Like my car, it takes up most of the main screen, but it doesn't take up any of the gauge cluster. And along the bottom of the main screen is my car's controls for things like climate. And that's fine. It works fine. And then they were like, well, our customers don't even want it anymore. Yeah, because anybody who wants it isn't your customer. Right.

14:42Casey Liss:Self-selecting.

14:43Marco Arment:it's the mac pro yeah yeah so we removed it from our cars and then three years later nobody who buys our cars wants carplay yeah who who could imagine that so yeah good luck i i've used your software i've owned i've owned your vehicle i've used your software without carplay and guess what i did i put my phone in a bracket next to the air vent because that's what i had to do because that's what i actually need is my phone in the car and your software is not good like the rivian map software was horrendous. The routing was bad. The traffic was bad. The directions were bad. The map data itself was bad.

15:17Marco Arment:The street names were bad. The exit names were bad. Everything about it was bad. So I never use it. I just used Waze. That's why, because theirs was worse. Guess what? Their Bluetooth implementation for audio? Bad. Limited. Their client apps for things like Apple Music. Bad. That's why I didn't use them for very long. And whenever they would announce some new update or some new service, I'd go try it again. And I would bail out after less than a day because it was so bad. Or I would try their routing because they said the routing's better. I would try it on one trip and it would be so bad. Sorry, Rivian.

15:56Marco Arment:I want Rivian to succeed. I really do because their vehicles, I think, have a lot of great aspect, a lot of promise. I like a lot about their company ethos. I like a lot about their style, their capabilities and things like off-road situations are amazing. Like there's so much and I'm so optimistic about their future lineup. Like I think the R2 is going to bring them a whole bunch of success. I really love how the R3 might be looking. The R3X would be a serious contender for my next car if it actually comes out like in the right time span. but CarPlay man that would solve so many problems and Rivian still continues to assert that we don't need CarPlay because we can just build in like you know people like podcasts right oh we'll just build an Apple podcast done we'll just have Spotify that's podcasts right makes me so mad oh you want maps here's a map it's all the same right nope it sure isn't nope

16:53John Siracusa:it makes me so mad and the thing of it is is that you can offer CarPlay and continue to do the shit you're already doing. It's not mutually exclusive.

17:02Marco Arment:I don't understand. And the worst part is they keep saying, we have to do this ourselves so that we can give people a good experience. They're not giving people a good experience. The good experience is I got to put my phone in a bracket because that's what I actually need. And I don't have to do that with my other car because my other car has CarPlay. And so I don't need to put my, so I don't put my phone in a bracket and use it while I'm driving.

17:26John Siracusa:Yep. No, it's preposterous to me. It's complete hubris. I am, my personal opinion, I think I've probably said this before, and I don't think it's an original thought, but I think eventually when Tesla sales continue to go through the crapper, I think eventually they will enable, we've heard rumors about this. I think we've even talked about it on the show. They will eventually enable CarPlay. And then spontaneously, the clouds will disappear at Rivian HQ and suddenly CarPlay will work on a Rivian because it works on Tesla's. So I guess it should work for us too.

17:55Marco Arment:I mean, we might literally hear about CarPlay being added to Tesla's next week. Like they could do something with Apple or Apple announces, you know, on some slide WDC, like that could actually be part of what we hear very soon. Like I can totally see that being a thing that they would want to like co-announce together.

18:13Casey Liss:There was something in the show notes about CarPlay on Tesla, but it just got pushed down and down and didn't even make it into overtime. There's a long queue. So thank you for bringing that back up.

18:21John Siracusa:You are welcome. All right. I told you we had to make this quick. And then I went on a tear with, well, Marco and I both. and air high five to you, Marco, because I'm right there with you. All right, let's talk about another thing that makes us so, so happy, Bamboo. With regard to last week's overtime and Bamboo, the 3D printer manufacturer and their fights with users or with their own users, Steve Riggins writes, Bamboo was not liked by the open source community well before this latest issue. They've repeatedly broken the spirit of open source software and this was, I believe, the final straw for many.

18:49John Siracusa:I own a Prusa, P-R-U-S-A, partially because of this older beef that I heard about via my friends and I did not want to encourage Bamboo to continue to act in this manner, which they clearly have. That was posted on Mastodon. Jose replied to that post saying, Facts! I stopped firmware updates right when they announced plans to lock down the comms on my X1C. I'll never buy another Bamboo.

19:12Marco Arment:I understand why people have a beef with Bamboo and it being an increasingly proprietary system. I get that. People have pointed out legitimate issues like Bamboo is a Chinese company and sending your models through their servers can possibly create some unusual or undesirable outcomes. I totally get that. I just don't care about those things from my point of view as a casual 3D printing hobbyist. I love how easy and good Bamboo stuff is. if I wanted to go kind of the more like open source route the perusa stuff looks really good I I totally get why people would do that and maybe in the future I will go that route but right now I'm perfectly happy to stay in this walled garden walled gardens can work really well as long as they're not the only gardens like that like and that's the good thing is I'm glad the world of open, free, non-lockdown 3D printing exists.

20:19Marco Arment:And I think there will always be a market for that. If I was running a big fleet of printers for a product company where I was actually printing stuff in mass and shipping it to customers, I would probably want more control over my hardware. And so I would probably do something like a bunch of peruses or something. But as a hobbyist, I love not having to deal with the increased complexities of the open source world. I love just having it be integrated because when it was harder, before I had the integrated bamboo system, I never did anything with it. I never 3D printed anything because it was too hard.

20:55Marco Arment:Now it's made it easy. What I said in the overtime for non-members about this, the basic gist of it is, please let me not care about the dispute here between bamboo and the open source people. Please let me continue to enjoy this hobby and not be an activist this one time for this one issue because that's that's currently a big source of joy for me and i want to stay that way yeah and your request was denied by the way

21:22Casey Liss:so many people wrote in and basically said to marco no sorry you can't not care so answer

21:26Marco Arment:asked and answered honestly i was not persuaded by those arguments well but you asked like you

21:31Casey Liss:shouldn't ask permission if you don't want to hear the answer you don't have their permission but you just you can do what you want because you don't need their permission thanks we are

21:38Marco Arment:sponsored this episode by Factor. Now imagine, hunger strikes, but you're exhausted. You've had a really long day. There's something healthy in your fridge that you should be making, but you just don't have it in you. Factor solves this problem for you. Factor helps you hit your nutrition goals this season and just eat healthy without the planning, the grocery runs, or the cooking. So Factor meals come to you fresh, never frozen. They have over 100 rotating weekly meals to pick from, including globally inspired flavors like Mediterranean and Asian. There's always new stuff to look forward to. And you get the Factor meals.

22:14Marco Arment:They're all crafted with functional ingredients, lean proteins, colorful veggies, whole foods, healthy fats. There's a whole bunch of ingredients, over 175 ingredients that Factor will not use. Things like artificial colors or sweeteners, high fructose corn syrup, refined seed oils. They have the whole list of stuff they don't use in any of their meals. So you can be sure that you're getting high quality stuff. And it is honestly so convenient. We were looking for something like this just two weeks ago. We started starting factor again because we wanted something that my teenage son who, you know, he gets home from school and he's starving.

22:46Marco Arment:What can he make? Well, he can go in the cabinet and eat a box of crackers and eat garbage. Or we're like, let's get him real good healthy meals. So we went right to factor for that because all he has to do is pick something out of the fridge in a factor box and put it in the microwave for two minutes, poke some holes in the top, and it's done. There's a whole meal, healthy, high protein, fresh vegetables, ready to go. It's a great situation. So head to factormeals.com slash ATP50off and use the code ATP50off to get 50 % off and free daily greens per box with new subscription only while supplies last until September 27th, 2026.

23:23Marco Arment:See website for more details. Thank you so much to Factor for sponsoring our show.

23:31John Siracusa:all right uh with regard to alarm reminders and clock alarms for more than 24 hours in advance ben smith writes you can set an alarm up to seven days in advance by using the repeat every x day of the week and selecting which day or days you want the alarm to go off given that alarm reminders are now way better at this i can't think of a reason why anyone would do this but it is possible that's

23:53Casey Liss:It's a clever workaround, but thankfully no longer needed.

23:56Marco Arment:Yeah, that is a clever workaround. It's a terrible solution. You shouldn't need to do that to set an alarm more than 24 hours ahead of time. So I understand why people do this with now, the new alarm reminders. That makes sense.

24:11John Siracusa:We talked a lot over the last few weeks about moving files and how to move files. I think the genesis of this was NASC ATP several weeks back.

24:17Casey Liss:Yeah, it was like do you copy and paste files? And we talked about cut, copy, and paste and then move modifiers. That's been the past three shows worth of discussion about this. Yeah.

24:26John Siracusa:Stephen Optubeek writes, Optubek, Optubeek, Optusomething.

24:30Casey Liss:He's given pronunciation, and I also can't remember who it is, but I know we always get it wrong. I'm sorry, Stephen.

24:36John Siracusa:Stephen Optusomething writes, you can use the move command, the MV command, to move files across volumes. It works just as you would expect. Here's a demo movie. And this was on Mastodon, and we'll put a link in the show notes. I can't believe I didn't, like, not scold you, but correct you about this. I guess you were just on a tear and I didn't want to interrupt, but I, of course you can use the MV command to move files across volume.

24:56Casey Liss:I mean, yeah, I didn't say definitively one or the other, but I just thought about it. I'm like, you know what? I've actually, you haven't tried that on, on Mac OS. And because the file moving the, the basic tools on Mac OS, like CP and MV and stuff do all sorts of Mac stuff. I'm like, I wonder, did they change that? But anyway, I didn't actually try it. Steven did. And so there's a video.

25:14John Siracusa:Then Steven continues in a second movie and a second toot. here I am interrupting the process by yanking the volume. Don't try this at home, kids. The move, or MV, only copied part of the file and kept the source intact.

25:26Casey Liss:So there you go. So it bravely tested the worst case scenario, which is if it dies in the middle, what happens? And the good news is that the source file is still there. So thumbs up to MV.

25:38John Siracusa:All right. And then someone whose identity was omitted or will be omitted here to spare them the embarrassment wrote in to say that they did not heed your advice, John, about cross-volume moves in the finder and ended up hosing an entire virtual machine that they were trying to move. Whoopsie-dipsies.

25:53Casey Liss:Yeah, they said they had my voice in their head while they were doing it, and they said, well, anyway. And then the thing failed, and they deleted the destination and realized their source was hosed. So remember, reminding again, what this is saying is, if you're doing an operation in the Finder with drag-and-drop that would normally be a copy, in other words, moving across volumes, but you override the fact that it's a copy by holding down command to make it a move, that's what i'm saying don't do because you know for the reasons stated here that like uh like i think the problem with the vm is it is actually a folder full of files and but it looks like one file so like you you drag it this is like a this is actually sort of working as designed but not as the user expects you know you hold down command turn it into a move it starts going but it fails for some reason you're like oh well i'll just do it over but what you didn't realize is that in that file that looks like a single file but it's actually a folder full of files each one of those files was like copy the destination delete the source copy the destination delete the source copy the destination delete the source oh i got interrupted so when it gets interrupted you're like oh the copy failed let me just delete the destination i'll recopy it but what you're recopying it is now a partially populated directory that looks like a file full of stuff this is again i don't think this is actually a bug this is like just working as designed but again not as you expect there are actual bugs where you're working with a single file and you can still end up in a scenario where the destination didn't get written correctly and the source is gone.

27:14Casey Liss:So don't do it. Just let it be a copy and then delete the source when you're sure the destination has been entirely written.

27:20John Siracusa:We did a member special on the movie Her and we talked during the member special about how, or at least I brought up, how cool I thought the user interfaces were on the screens. And Peter Pulio writes, fun little tidbit about the movie Her, the artist Jeff McFetridge designed all the computer UI seen in the film. His art is wonderful and worth checking out. But he's also done a lot of work for Apple, including the artist Watch Face, WWDC 2017 graphics, and Apple Pay Transit ads. Come to think of it, living so close to New York City, Marco's probably seen some of his Apple Pay Transit ads in person.

27:53Casey Liss:I have fond memories of that WWDC 2017. You remember that, like the top view of all the little people? It was really cool looking. So, yeah, I'm like, oh, that guy. Yeah, I know his style.

28:03Marco Arment:Yeah, I really like that style. we are sponsored this episode by delete me have you ever thought i should make it a little bit less easy for anybody you know scammers stalkers or whatever to find my personal info online here's what you do go to www.joindeleteme.com slash atp enter code atp you'll get 20 off delete me delete me removes your personal information that's being sold online so what these are these are data broker sites. They post your information and they advertise, say, hey, if you give us a few bucks, you can get this person's name, address, phone number, names of family members. It's a really gross business and it's kind of unsettling to see your stuff out there so easy for anybody to get.

28:46Marco Arment:But Delete Me is made to help. So Delete Me automates takedowns and opt-outs for hundreds of data broker services out there. So you tell them what information you want to be removed. You can have your stuff, you can have a family plan, and then Delete.me does all the work to wipe your personal info from all of those data broker sites that they can find. You can get an individual plan for a little over eight bucks a month, you can get even bigger discounts with a two-year plan, or when you enroll with a partner or family, it's a great service to do this data removal service for you. Delete.me was named the top pick for data removal services by Wirecutter.

29:25Marco Arment:So check it out. I personally went around looking for one of these services a few years back. I found DeleteMe before they were a sponsor, and I was very impressed by it. I used it. It was great. Take control of your data and keep your private life private by signing up for DeleteMe, now at a special discount for our listeners. Get 20 % off your DeleteMe plan when you go to www.joindeleteme.com slash ATP and use promo code ATP at checkout. The only way to get 20 % off is to go to www.joindeliteme.com slash ATP and enter code ATP at checkout. That's www.joindeliteme.com slash ATP, code ATP. Thanks to Delete Me for sponsoring our show.

30:11John Siracusa:It is the week before WWDC. WWDC is here next week. We will not be going unless there's a surprise that I was not aware of, but we will be bringing you coverage as always. And the theme for this year is all systems glow.

30:29Casey Liss:I feel like they've leaned on this one a few times. I mean, it's obvious that they're referencing the sort of HDR over bright effects that they use when you activate Siri. And like, didn't they use, maybe they didn't use the exact phrase, but I feel like the art on at least a couple of recent WWDCs has been like, hey, did you know we use HDR colors and images and graphics in iOS? Like, yeah, we know. We know. But anyway, here we are. This is WWDC 2024 take two.

31:00Marco Arment:Hopefully. I mean, we'll see what we actually get here. I mean, I think German's reporting, which we'll get to in a second, has been pretty strong in one direction. And that's, you know, there's probably a lot of truth there. But what I hope to see broadly, AI is everywhere in the industry. a bunch of it is BS and is turning out to be underwhelming but a bunch of it is turning out to be pretty useful and pretty high value Apple is so far pretty much nowhere in this game they don't make the models they have tried and they have failed to make the models they don't host the infrastructure and I think more importantly for Apple and their products they currently don't really use AI well in their products.

31:51Marco Arment:And these things are obviously related to a degree. But what I want to see from Apple is if you're not going to build the models, fine. Hire Google to do it, that's fine. Google makes good models, so that seems like a totally fine choice. If you want Google to host the infrastructure too, that's also fine. These big companies have lots of infrastructure, lots of capacity. Apple has used things like Azure and AWS for parts of their services before. So that's perfectly fine. What I want to see from Apple is the ability to integrate AI into making good products and features. We haven't seen that either.

32:32Marco Arment:So far, Apple has been utterly paralyzed and seemingly unable to ship anything good that uses AI. You look at how they have tried so far in their little baby steps. Siri, of course, is not the baby step. Siri is like the giant Apple AI-like area that they have so far just completely failed in so many giant ways over time with Siri. And its reputation is terrible. Everything bad about Siri's reputation is 100 % deserved. because it really has been massively missed opportunities, huge inflated promises that they cannot then deliver on or that they deliver inconsistently on. And then you look at how else have they integrated AI into their OSes and products.

33:24Marco Arment:So far, badly. We have crap like image playgrounds, which is horrendous. The writing tools, which are pretty rough and not super useful. You know, minor things like photo editing that are kind of like half-baked, half-there, mostly over-promised and under-delivered. We have things like the image visual intelligence. That's also pretty half-baked and pretty underwhelming compared to what Android has, which is the same feature but better and works better. We have not yet seen Apple do anything successful with AI.

33:58Casey Liss:I can't believe you forgot Genmoji. Oh, yeah, right. They've done multiple versions of Genmoji. Image Playgrounds is here, but so is Genmoji.

34:06Marco Arment:Right. So what I want to see is, It's like, okay, if you can't make the models, fine. You can't host the models, fine. But use them to build products that don't suck. Use them to build features that are compelling. Add APIs that developers can use to build features and products that are compelling.

34:24Casey Liss:They did that. They did your transcription thing. That's an example of using AI to make a useful feature.

34:28Marco Arment:They did. And the transcription, I think, is possibly the only thing they did pretty well with AI so far at an API level. The foundation models are there in the 26 releases. I'm glad they're there. They're very limited, and I'm hoping to see improvements there with 27. But ultimately, what I want to see from the 27 OSs is for Apple to join the AI revolution, to set expectations accordingly. I don't expect them to have their own models from scratch. I'm sure whatever their deal is with Google, that's fine. I don't expect them to become a model company. But Apple in general is a product and platform company.

35:13Marco Arment:So what I want to see is get on board with AI in your products and platforms for things that it can actually do well for compelling features that are well implemented, well designed, and well integrated. So far, they have shown no willingness or ability to use AI in a well-implemented, well-designed, and well-integrated way. We just have crap like image playgrounds. And I would love to see Apple deliver something with AI that is not a half-assed laughingstock.

35:44Casey Liss:Put in a slightly good word for the image cleanup, which I think is kind of like voice transcription. a useful feature that couldn't exist without ai because the previous like repair thing that they had was really falling behind the state of the art it's why i always ended up using uh edit and pixel mater pro because they had some machine learning powered uh image repair thing for years and i was like always using that because it works so much better and then apple added cleanup it's like hey we'll do that too now is cleanup as good as google's things no it's not google has better image cleanup tools as you noted um but like transcription it's like well it's better than it not being there it should be a system level feature it should be part of the photos app it should be uh something that they add and they did add it to their credit but now i'm going to take it all away because uh that feature on my wife's m1 max max studio in the photos application works about 10 of the time and you know what happens the other 90 of the time in the little sidebar in photos i see a little text message that says uh image cleanup could not be loaded try again later i don't know what that means it's persisted across major versions of mac os i believe uh just i i it's it's like oh it doesn't never work no once in a while it works but not all the time and i can't figure out why so way to screw up that one feature that i actually liked uh rumor there are we're not going to go over like every rumor for wwc we'll link to some roundups where you can see there's lots of stuff that's rumored but um we're gonna hit some highlights but one of the things that is rumored that we're not going to cover today is supposedly better machine learning powered, AI powered photo editing features.

37:18Casey Liss:Hopefully they'll actually work this time, but that is one of the things that I think is an example of take AI technology and make a useful product feature out of it.

37:27John Siracusa:I think being contrarian right now insofar as we're not going to shove AI down your throats any more than we already are, but we're not going to do any more shoving of AI down your throats. Look at us. We're with you. We agree. I think as much as I'm snarking right now, I think that's actually a potential victory to just say, look, we've tried what we've tried. We're going to make a few things better, but we're not going to just sprinkle AI dust on every freaking thing we have.

37:53Casey Liss:Do you think that's what Apple's saying, or are you saying you wish they would say that?

37:56John Siracusa:I think that they have the opportunity to say that because they've done so little now, and they can just lean right into it. But they will not say that. Oh, they won't. They absolutely won't. But it would be neat, because I don't know. Well, I just I feel like Marco, I think, opened this entire segment about AI with saying that, you know, in some ways, in certain contexts, it's very good and very impressive and very useful. And then there's eight trillion other ways where it's being shoved right down our throats. And I'm not here for any of that. And I wish that somebody would be the adult in the room and just say, you know what, we're going to do what Apple always used to do, which is we will use this underlying technology to do things that are helpful in the real world to real people.

38:38John Siracusa:Like, you know, photo editing and stuff like that, rather than just making something stupid that nobody wants, like image playgrounds. So I don't know what they're going to do. I suspect that they're going to try to just, everyone seems to think, and it would not surprise me, that they're just going to try to right all the wrongs from 2024. What is the phrase from Godfather? Settle all the family debts or something like that. So close. Whatever. I don't think I've ever seen it. I believe it's business. John, why don't you just correct me now so we can save all the news.

39:07Casey Liss:No, I'm just letting Marco correct you on movie quotes. It's the world gone mad.

39:10John Siracusa:Not a good spot for me, my friends. Not a good spot for me. I love you, Marco. But anyways, the point is that it would be very refreshing to see them treat AI as a tool or a means to an end rather than the end itself, which is what the rest of the industry is doing.

39:24Casey Liss:So I think they have, like, they've been, what Marco described, if you asked Apple, they'd be like, oh, that's what we're doing. But our rebuttal would be, yeah, but the places where you added it to your products didn't make them better. Like, I mean, I'm not saying Genmoji is bad, but like, that's, like, what is your usage of that feature? Like, it may have been implemented well, it may have even been integrated well, and it may have been integrated so well that it's not in your face because, you know, you're not being prompted for it a lot but it's not something people want to do versus photo editing which i think is a good use and transcription is a thing and developers want to use so that's a good use image playgrounds they definitely threw in your face image playgrounds is here uh as marco was talking about like when you would install the os it would throw it in your face 48 times oh that's right that's right yeah it would just constantly say hey image playgrounds exist because otherwise how would you find it would just be this app in your applications folder and it's look if there was demand for not very good image generation in a really limited environment people would find the app but they don't because it's not a thing that they want to do so yeah the features they do have i mean i said we're not going to cover everything that they're that is supposedly coming we're going to cover a couple highlights and one of them is uh the serial overhaul on ios 27 and german's got a little info on that but i think before we get

40:32Marco Arment:there just ultimately what i want to see is when apple either is not really ready to do something well yet, or if the industry is demanding it, but they don't respect it, they call it a playground. You know, we had Swift Playgrounds on the iPad, and that was like, we're not going to give you real Swift on the iPad. I mean, it has broadened since then, but like, it's like, we're not going to give you the full-blown environment. We're going to give you a toy. Here's a toy version of this. Image Playground was a perfect name for that in Apple's parlance, because they're like, this whole AI thing, we don't respect this.

41:08Marco Arment:That was obviously, Craig Federighi was very clear about that. He and Jaws and the top Apple brass did not respect AI. For a long time, even after the rest of the industry was like, this is getting pretty good, pretty useful, they crapped all over it in public. So you can imagine how bad it must have been in private.

41:29Casey Liss:What they came out against was chatbots. Federighi, according to Inside rumors sort of got religion about chat GPT or the technology itself back with like GPT version two or something. But what they all came out publicly against is that thing where you go to chat GPT.com and have a conversation with that LM were against that. That's what they were coming out with. And I feel like AI, quote unquote, AI is a much broader topic than that. But the manifestation that hundreds of millions of people see is I go to a website or I use an app and I type back and forth to a chatbot. And that's what they were against.

42:00Marco Arment:yeah but honestly that was obviously the wrong take in retrospect and even then it's like when when old dudes heard about younger people texting their friends all the time instead of calling them on the phone and they were like that's terrible why would anybody want to text their friends and you just call them and you know what everyone wants to text their friends as it turns out that's a very generational thing and i think jaws and federighi obviously missed that completely for way too long. And so as a result, Apple in general just did not see AI as something that they needed to care about when that was obviously wrong.

42:36Marco Arment:And so that's kind of what led them to be so far behind among other problems. But what they shipped was almost like trying to placate the stock market. Like, look, we can AI. Here's a half-assed app that generates some garbage that nobody will ever use because it's terrible. When Siri was launched in 2011, it started out in a pretty competitive place for the time. It just never moved forward, and then everyone else did move forward. Apple's AI stuff wasn't even competitive the day it launched. Literally, the day all this stuff came out, it was already laughed at. It was ridiculous. It was way less capable than everything that was out on day one.

43:21Marco Arment:And of course, it never got better. So they were coming from a place of deep disrespect for all of this. They seemingly saw it as something they were obligated to do by market forces or the press or whatever. And they crapped out some garbage that you could tell they didn't respect it at all. It was shoved down everyone's throats through promotion in the settings app everywhere. Image playgrounds are here. And you had to click it and dismiss it and everything to get it to turn off that stupid red badge for that stupid notification for their stupid promo and their ad for their settings. app not that i'm upset 48 times they obviously didn't respect it and look at how much that has cost them they are they are still so far behind what i want to see broadly is signs that they respect it now and have some better idea on how to actually use ai in their products in a again in a well-designed well-integrated way not just some like bolted on crap thing like you look at the writing tools uh api they did too writing tools are pretty rough in large part because they just punted and didn't make a ui like at all there's no ui you just like right click your menu and do some stuff and it's like well what if you want to like adjust stuff or see the diffs or anything nope there's no ui because apple doesn't know how to make uis anymore either apparently but that's a whole separate topic but like i just again like i want to see them take this world seriously and show us that you can make good products.

44:49Marco Arment:Because that's what Apple's strengths have usually been, is take technology that's out there, that's floating around, use their skill, their product sense, their design sense, and their UI chops to make good products out of compelling tech. We have the compelling tech. We've had it for a while, longer than they think so, but it's been here for a while. Maybe they finally agree that it's compelling tech. Like, now let's see, can they use this compelling tech for any good product features, products, or integrations? So far, they haven't. And so if they continue not to, I'm kind of worried, honestly.

45:29Marco Arment:If Apple can't figure out how to make good products with AI or how to integrate AI in a good way into their products for new features and stuff, what are they doing? Where are they going to go the next 10 years? Those are major doubts that I have about them that hopefully in a week we will be able to say, oh, it seems like they've turned over a new page or a leaf or whatever the expression is. It seems like they're now in a better direction. And look, there are some cool integrations here. And maybe the betas have some of them working already, hopefully, fingers crossed. And maybe we can see signs that they are able to harness this new world and make great stuff with it.

46:10Marco Arment:We have not yet seen those signs.

46:12Casey Liss:it's fair for you to not mention this because they didn't ship it but what they announced in 2024 and didn't ship is exactly what you're talking about like all the stuff that they said it was going to do but it never actually did is let's integrate this intelligence into like i think the apple intelligence strategy as articulated as articulated in wc 2024 was and is the best strategy for apple which is we have all this data about you let's let the model know it all and let's let you talk to the model and have it manipulate all this data in a privacy preserving way on your device because like we're the only one who we're going to give that kind of access to we don't want to you know allow chat gpt to have essentially like the equivalent of like the computer control thing where it just controls your mac or whatever we can do that because we're apple but these other things can't so we're going to roll out apple intelligence and you'll be able to ask it when your mom's flight is arriving and it can see your email and see your messages and see all this end-to-end encrypted stuff and do all like that was a thing that would take this technology and make apple's products better they just didn't ship it like that's the was the problem and what they did ship is everything you listed which is like well what the heck is this you don't have a chatbot and you hate chatbots you've got image playgrounds which no one's using you got genmoji which no one's using you got writing tools which is half-hearted at best and terrible you've got the image cleanup it doesn't run on my wife's computer so i'm mad about that um and everything else like the bulk of what was in wwc24 what i thought of as apple the apple intelligence strategy after that keynote just didn't arrive for two years so this is a chance for them to get back to that but on the topic of chatbots and again getting back to ermin Gurman, I have some good news, bad news.

47:42Casey Liss:I think it's, well, we'll get to it after Casey reads what it is. But I have mixed feelings about what is rumored for Siri and iOS 27.

47:51John Siracusa:So from the 28th of May, Mark Gurman writes, illustrations created by Bloomberg show the revamped Siri interface, a new chatbot-style app, and other major iOS 27 changes that the company plans to announce at WWDC. The new Siri will include the delayed features announced in 2024, such as the ability to understand personal data and analyze on-screen content. But those capabilities are just one part of a broader wave of updates, a rebuilt model that uses Google Gemini technology, AI-powered web search, and a completely redesigned interface. There's also a dedicated Siri app designed to compete more directly with chat GPT and other AI assistants.

48:27John Siracusa:So in summary, also Siri and the camera app alongside photo and video, plus customization of the camera app's toolbar icons. And there are a whole bunch of screenshots in this article. We'll put a link in the show notes. We also have a link in the show notes to a MacRumors article, which regurgitates a couple of those screenshots.

48:44Casey Liss:Yeah, these aren't screenshots. These are mock-ups. I'm sorry. Yes, I'm sorry. They commissioned an artist to draw what they had a description of. Here's what the new Siri interface is going to look like. And so I think the key parts here is basically like, hey, all that stuff in 2024. Now they're going to do it more or less. And also, did you want a chatbot? Guess what? The thing that Craig Federici and Jaws were coming out against two years ago saying, we don't, the chatbots, we don't think that's it. apple is going to have one built into the os you'll talk back and forth with siri which will really be a gemini powered lm the interface they show is basically like if you pull down from the top like the dynamic island expands to be a search box like always but what the search box has in it now is like search or ask where you can ask and presumably that goes to apple's you know model that's powered by gemini um you can also go to chat gpt in this thing and supposedly it's third party pluggable so there may be some competition there but then they have screenshots of a chat app an lm powered chat app where you're talking to to to siri uh and it's all like dark mode and you ask questions and give responses and it looks the mock-up looks like chat gpt basically in dark mode only you're not talking to chat gpt and uh you know obviously tons of people want to use you know ai chatbots because they do they use them i'm still not entirely sure that apple needs to or should make one because i am i'm i feel like what they're going to make cannot possibly be as good as the the real chatbots like using chat gpt directly using claude directly those are i i just don't see how apple can keep up those things get updated all the time and keep improving and keep getting better and better and apple probably can't keep up and they're going to give you this out to go to the third party ones anyways so it's like this weird waypoint on your way to going to a real product that people actually like and use you can stop off midway through and talk to apple siri back and forth and yeah it'll be smarter than siri has ever been before because that's not hard but really if you want to actually have a productive serious conversation you should go to one of the third party chatbots because they're going to be better um and when you know federigi and jaws were saying we think chatbots is not it that's not the way to do it they were trying to contrast that to hey it's just going to be they're just going to be features of your phone that'll be powered by this stuff under the covers but you won't know it your phone will just be more capable and they felt like just punting and saying well we're not going to do any of that instead there's a window where you can type text back and forth like they already have that it's on the iphone it's called the chat gpt app it's called the cloud like you just that those are third-party apps and apple seems like they are if it's rumors to be believed they're saying we have to have one of those two and i just i have no confidence that the one they have will be competitive i just even though they're even though they're using google gemini under the covers like So you would think like, well, at least it'll be roughly competitive with Google.

51:35Casey Liss:I'm I don't believe that. I think it will be worse. I think it will have fewer features. I think it will advance more slowly and kind of like how in current iOS, when you go to chat GPT through Siri with the integration that they've had with OpenAI since 2024 or whatever, that it's worse than going to chat GPT directly. I think that will be true of talking to Siri. So I'm not enthusiastic about that. I'm enthusiastic about them rolling out the promise features from 2024, where just I can ask my phone something and it can look at all the data that only it can see because it's private on my device and do useful stuff.

52:06Casey Liss:That's something that can't be matched by the third party clients because Apple won't let them because Apple doesn't give them access to your stuff. So that I'm looking forward to. But I cannot see myself ever saying, you know what? I want to talk to an LLM chatbot and you know what I'm going to do? I'm going to launch the Siri app and do that. I just don't see that happening. They're so far behind that I don't feel like they are going to be competitive in this realm.

52:27Marco Arment:Yeah, I don't think they've ever shown the ability to be competitive in things like this. You know, in many ways, what Apple's software has seemed like in, oh, geez, maybe a decade, is it just seems like it's moving at a fairly leisurely pace in almost every way except designs, which they're happy to rush and shove through. but like what with the exception of redesigns which they love uh apparently um with with that exception the rest of it seems to just move glacially now i understand this is a significant challenge at their scale they have a lot of platforms they update them all pretty much in lockstep um that's not a small job that's also a self-created problem and optional but hey that's Not a small job.

53:15Marco Arment:I get that. But what it has seemed like for a long time is that they just aren't stepping on the gas in this new era. It seems like they're coasting in software. They're not really pushing things people want to push. They're not really participating in this whole new boom much. and even among things that they even if we set aside AI for a minute Apple software just doesn't move that quickly and they don't you know like what you were just saying like they will put something out there and then it just dies like it just gets ignored or you know they'll do the first 80 % of something and show off a good demo at WBC and you'll never hear about it again it'll never get that last 20 % of effort put into it it'll just kind of sit there in mediocrity forever until it's eventually discontinued or replaced.

54:08Marco Arment:That happens with so much of Apple software. And, you know, it's from from things like APIs all the way up to apps, like entire apps like that. That happens a lot. And it just seems like they're just coasting and not really stepping on the gas. It seems like they don't really maybe they don't think they need to. I don't know. But the entire industry is exploding and booming right now with all sorts of new potential. So not only we're not seeing Apple take advantage of the new potential or use the time and resources to polish up all the stuff they already do. Neither one of those things is happening.

54:47Marco Arment:And so again, like show signs of some change of direction or ability or even just caring. Show signs that they can care. And follow through with something. Show signs they're stepping on the gas. Show signs that they're actually active and trying to be competitive in some way instead of just coasting and resting on their laurels and collecting fees. Because they're really good at those things. But let's see them really push on the products a little more.

55:22Casey Liss:another one of the rumors is third-party integration integration of third-party ai that it'll be pluggable that it won't just be this one deal with open ai but you'll be able to pick from the leading vendors and you know this is the other strategy that lots of people talk about is like apples need to do this they just need to integrate with the third parties just you know they don't they don't need to do everything they're already the the platform that people use their mobile the mobile platform people use these apps from the most so why don't they just integrate why don't they just give integration i mean one of the reasons they don't want to do that is obviously they're willing to give their own uh lm stuff access to all your contacts and all your messages and all your mail and blah blah because they're apple and they trust themselves right and they don't trust you know facebook open ai anthropic to get access to that same data in theory they could do it in a privacy preserving way by having really limited apis but in the end those companies are probably just like well we'll just continue to get people to to download our app and onto their phone and use it that way so there is a place for system integration and the rumor is they're supposed to you know have more third-party lm integration does that mean every single place you could ask siri there will be a choice to ask some third-party thing do you will you only get to pick one will it only be surfaces in some places that remains to be seen but that's one of the rumors another one in the screen swatch where you see um the camera app having a you know instead of like photo video portrait blah blah that siri will be an option because they're basically taking visual intelligence which was previously hidden under like a whatever was long press on the action button or whatever the heck the default shortcut is they're going to put that right into the camera app which i think is a good idea but that brings up the third-party integration thing okay i think the camera app should have a thing that says hey what is this you know that's a question you've been able to ask with like google lens for like a decade i guess and lms are way better at it than those old things used to be will the only option be to use the quote-unquote siri camera to ask siri which then asks the google gemini model or will you be able to say in settings somewhere, what do you want to use for the thing where you point your camera at something and ask it what it is?

57:17Casey Liss:And you could say, oh, I've got a ChatGPT subscription. So I'm going to set that to ChatGPT and not Siri. It seems like that's not going to happen, but that's the type of third-party integration that it's like, you have to choose Apple. Are you going to do everything and actually keep up with it? I think you're not going to keep up with it. Or are you going to provide the integration needed? In this case, like in the camera app, I think they could do that in a privacy preserving way allow a third-party model you know like do some integration maybe not in this release maybe there'll be a future release or whatever but when i look at this feature i'm like okay but if i really wanted to know what it was would i ever use siri camera or would i launch the chat gpt app and use that directly just because i know from experience that it's going to do a better job or what i'll use google lens because i know from experience that's going to do but it's not called that anymore whatever the heck it's called now i in practice i frequently use the google app i think I described when I had a halogen bulb or whatever.

58:08Casey Liss:I didn't go to an LLM. I went to Google image search or whatever, because it does a good job. Now that's probably powered by Gemini behind the scenes. Anyway, I don't know. I don't care. But the point is I went to their app and I think that will continue to be the case. So I expect WWC to be this weird mix of we're doing a bunch of stuff and we don't suck as bad as we used to. And we're like, okay, well, you know, good. Not sucking as bad as Siri is a low bar, but you need to do that. But also will there be some push for like, and also we have third party integrations in select places. And that seems like it's going to be more half-hearted, which is kind of a shame because I feel like that's what they should be doing.

58:43Casey Liss:Like if they, if they are not going to be able to keep up with the big third parties in terms of the stuff, they should be really good about integrating them. And in the areas where only quote unquote, only Apple can do it for privacy reasons, that's where they should put all their effort. Do those features that nobody else can do. And, you know, maybe someday plug a third party thing into that, but, but we'll see. I don't want to prejudge it because, again, maybe they're using Google's model, so maybe it won't suck as bad as we think it will. But I put it this way. I look forward to being able to talk to my phone, my$1 ,000 whatever iPhone, whatever pro, the same way I can talk to my$25 Amazon Echo puck that I won in a raffle 10 years ago.

59:26Casey Liss:Because I can talk to that one. You know what? That one's backed on the server by some decent model. I talk to it. I don't care how I phrase sentences. I just say any old crap and it gives me reasonable responses. I cannot do that with my phone. And that hopefully will change next week.

59:41Marco Arment:It's hard for us to have any faith in Apple's abilities in this area because Siri hasn't even done like the 10 years ago feature set reliably and well over this time. And when you look at modern LLM tech, it seems like it would be a godsend to something like siri because you know where siri fell down in so many so many ways you know some of it was just like basic reliability and responsiveness and stuff like that server failures response failures and everything that's that's a different problem they also need to solve but where siri would often would often fall down is like miss parsing what you meant by your command or you ask or like it asks something you you give a follow-up and it it like kind of figures the wrong direction on what that follow-up meant or what you were asking or what you were saying or it drops the thread or whatever.

1:00:32Marco Arment:LLMs are really good at the natural language interface. They're really good at taking sloppy human input, figuring out what they meant and what a good answer to that would be and delivering it. And so you would think that this would be exactly what Siri needs, not to become ChatGPT, but to just become a good Siri. That's ultimately what we need from Siri, what we want from you. Because, you know, John, you're right. Apple's never going to be as aggressive about updates, as aggressive about features. Like, they're never, because again, they can't step on the gas. They are so slow at so many of these areas.

1:01:14Marco Arment:So we know they're not going to replace the desire for most people to have the chat GPT or cloud or Gemini apps on their phones and use them all the time. Siri is not going to compete with that for most people. And Apple shouldn't try and I don't even know if they are trying. The app redesign that was rumored for Siri could just be a better way to access Siri instead of it just being in these ephemeral floating blobs that just are gone the second that you've stopped looking at them. Which is probably a good strategy for it. But ultimately what we want from Siri is for it to just deliver on what it has always supposed to deliver on better and reliably and they should be able to do that with modern llm tech it should help them do that that's what i want to see i don't need them to become gemini or claude i need them to become siri to actually do siri well that's what i want to see and what and if they just put a new ui on the same old crap siri again uh then that's not really going to do it for me um but i i think i mean look we we keep giving them chances over and over again every year there for the last few years there's been rumors that siri's about to get better it hasn't uh maybe this will be the year that that it gets something other than a superficial interface tweak or a superficial you know you know tweak to its basic behavior but I want to see something beyond superficial.

1:02:43Marco Arment:I want to see the results of them having stepped on the gas ever on Siri and actually being able to deliver something better.

1:02:51Casey Liss:You keep using the car analogy, but I think the Siri situation is that the last car they were in exploded. And then they had to build a new car and clean up the wreckage from the old ones, so thus the delay. Although, again, we talked about this on a past episode that, like, supposedly, like, the 26.4 update was supposed to have a bunch of Siri advances, and they kept delaying that so if they uh announce you know all these siri features that we just described at uh wwc at the keynote and they say but actually some of these features are not shipping until the fall well then we're going to be like oh which features you know the ones that are actually good and that makes siri not suck we'll see we'll see you'll see if you see what they've learned because i feel like they've had they have had some hard lessons about announcing things at wwc but everything that we just went through with the gorman river seemed pretty solid a chatbot interface, a Gemini powered Siri, the photos thing or whatever.

1:03:42Casey Liss:So I think they'll have some good features and I think they will, you know, because again, the Gemini deal, the Gemini deal is the thing that gives me the most faith. I know Gemini exists and works and I used it. And so if that is powering their backend, I know they'll probably make it a little bit worse, but hopefully they won't be able to make it that much worse.

1:03:58John Siracusa:I don't know. We'll see what happens. I just, I hope there's more to this than just AI talk in, I mean, both WWDC and the rest of this episode, because we've caught on a tear.

1:04:07Casey Liss:Well, you should go to the macOS topic next.

1:04:09John Siracusa:Oh, God, because none of us have opinions about that. All right, so... It's not AI. It's not AI, at least. All right, the 10th of May. It's also not intelligent. Oh, sick burn. The 10th of May, Mark Ehrman writes, though the Mac software introduced the same liquid glass interface seen in iOS 26, the design language hasn't translated as smoothly to the larger displays and different input methods of desktops and laptops. Part of the reason is that liquid glass was created with more modern hardware in mind. I'm sorry. What? I believe it was -

1:04:38Casey Liss:Keep reading so we get to the end of this ridiculousness.

1:04:41John Siracusa:Okay, so yeah, fair enough. So let me repeat what I just said. Part of the reason is that Liquid Glass was created with more modern hardware in mind. The crisp OLED displays are used on iPhones, some iPads, and Apple Watches. Macs also continue to use LCD displays, which don't render translucency, shadows, and glass effects as effectively as OLED screens. I'm sorry, what the f*** is he on? What is he talking about?

1:05:02Marco Arment:This is like we can fix the butterfly keyboard, but you need to put a gasket under all the keys. That'll fix it, right?

1:05:07Casey Liss:Well, no. So here's the thing. Because of Mark Gurman's infuriating writing style, let's call it, it's impossible for me to tell if the BS that you just read is Mark Gurman offering opinion or conjecture or if it's sourced. I think it's Mark Gurman offering opinion or conjecture, and that opinion or conjecture is stupid. Yes. Because it can't render translucency on an LCD? Why? like the contrast is higher and you get better blacks on an oled but rendering translucency has to do with compositing multiple layers to come up with the final pixel color like aqua on 10.0 on lcds had plenty of translucency and was rendered fine so the oled thing just makes no sense to me the best i can come up with is if you have a bad quality display and you have a ui that has like barely any difference between two adjoining shades of white that it looks bad but that's not the lcd's fault that's the fault of the person who made a ui where you have two colors of white that are almost the same right next to each other that's the problem and it doesn't look any better on oled oled correctly renders the tiny difference in whites between the two things and so do lcds so this this is old news i know but like it's infuriating because it's like of all the nonsense reasons to try to excuse liquid glass on the mac blaming on the lack of oled screens for mac is just the most ridiculous thing i've heard and i don't think this has anything to do with apple i don't think an apple source is saying this i don't think anyone at apple would agree with this so i'm just really dumping on mark german but in case you heard this they're like oh liquid glass looks bad because max don't have oled screens and that's going to change soon hey it's not going to change that soon so don't worry about that maybe that's not that's not why liquid glass and the mac is bad yeah because it turns out

1:06:44Marco Arment:But blurry text is not going to be solved by OLED screens. You know, if you look at most of Liquid Glass's failures, like, because again, like, I want to be clear, I don't hate everything about Liquid Glass. I just think it has some bad designs. Number one for me is text in a computer UI should never be blurry, period. I will hear no other arguments on that topic. Blurry text is bad design. So where do you see blurry text in liquid glass? Everywhere. Okay. Why? Because they made a barless design, basically. You have navigation bars, toolbars, and sidebars, and they decided all of those things should either be gone but still have their controls float or be very, very, very translucent to the point where they get these blur overlays and these blur effects at the edges.

1:07:39Marco Arment:So Liquid Glass, the design, is filled with basically kludges, hacks, bad hacks that they're like, okay, we started from a place of we want everything to be transparent and floating over the content to respect the content. Okay. But we're going to float everything over the content and be translucent. So then, okay, how do scrolling bars work? oh well I guess we'll we have to blur things you know progressively blur them as they get closer to the edges so they scroll better or whatever like that was a bad solution to a precondition they created with the design that should probably be revisited because what's what's better than blurring text is not creating the conditions that you need to blur the text to use the controls.

1:08:27Marco Arment:Okay, let's look at designs to do that. Maybe bars are the solution. I don't know. Just throwing it out there because I think we might have had those in the past and they might have been fine. Maybe some kind of bars to separate content from control areas. Bear with me. maybe the controls shouldn't be on top of and blended in with the content maybe to respect the content having the controls be separate is not only better for the design but also better for the controls and better for the content i know that's very radical today but just consider the possibility so anyway what liquid glass needs and and what i am hoping to see although i don't expect much but what i am hoping to see is tweaks to the design that still keep some of the aesthetic of it that like the parts that are good because there are parts of it that are good um but that remove the need for bad hacks like blurring text everywhere because that is always a bad hack the other thing i would love to see removed um is they have this amazing system where when you have an app with a sidebar the sidebar is supposed to be translucent now and content is supposed to appear under it now what if there isn't content under it because what if like the thing that is on the on the right side uh like the content area what if it doesn't have additional content that is under the bar on the left, logically speaking, or that content would be inaccessible.

1:10:10Marco Arment:Well, they have a whole system they recommend that you do where you mirror a fake mirror of the content on the right under the bar on the left just for the purposes of blurring under that bar. So you're blurring mirrored fake content under it. So you get color blobs going through there. So liquid glass looks coherent. That is the most ridiculous thing ever. That's the kind of hack that like, if your design requires that to look coherent, it's a bad design system. That needs to be revisited. That needs to be iterated. So what I want to see is Liquid Glass being tweaked to need fewer crappy hacks that make for crappy UI.

1:10:55Marco Arment:And that's blurry text and fake blurred content.

1:10:59John Siracusa:Yeah, yeah, I mean, you're not wrong. And the thing of it is, is that I am actually mostly okay with liquid glass, which I know is blasphemous, and I'm probably going to get fired for saying that. But I'm mostly okay with it, but it's certainly least well-considered on macOS. Again, this is not an original thought. But I would really love to see – I think some people in our little commentariat or whatever the word is I'm looking for are like, oh, they should just crap, put it in the can and walk away from liquid glass. And A, they're never going to do that. Never going to happen. Yeah, I don't think it's necessary.

1:11:31John Siracusa:Like, there are some grumbly people on Mastodon that are like, oh, I should just walk it back and never touch it again. I don't think that's necessary. I think it looks pretty good, even on macOS for the most part. But I think it would be nice to do what they did, as we all said last summer, do what they did with iOS 7 and walk back some of the more aggressive stuff to make it a little more, I almost said well-rounded. That's probably not the phrase I want here, but make it a little more palatable. and I think there'll be some of that. I doubt they'll call attention to it, but I think there'll be some of it and I'm looking forward to it.

1:12:05John Siracusa:I'm looking forward to it on all the platforms, even though I think as we all agree and again, not an original thought, I think macOS needs it the most, but I would like to see it pretty much everywhere. I think everywhere could stand to be improved a little bit.

1:12:17Marco Arment:Also, I would say like, again, even setting aside design changes, the implementation is still very sloppy and half-baked too. Like one of the worst things about on iOS, like a floating glass button, like a toolbar button, if you tap the edge of that button, it plays the animation of you tapping the button, but the button tap does not fire. So if you don't hit the middle of a button, you can think you're hitting a button and it doesn't actually trigger the action. And there is no way for the underlying app to change that a lot of the time. so like there's little implementation bug or like you know bad animations like there's there's little bugs like this all over the system that make it difficult for us as developers to make good uis with liquid glass and so if let's assume that you like liquid glass and support it and want it to stay well then at least give us the ability to make good apps with it like get make it work better.

1:13:22Marco Arment:Obviously, I'm going to have my opinion about changes to the components, things like bars and stuff, fine. But even if you don't change anything like that, make the current system work better. Because right now, it still is full of paper cuts.

1:13:36John Siracusa:So continuing with what Mark Gurman said, I'm told the company is preparing what people internally consider to be a, quote, slight redesign quote for macOS 27. With the next update, Apple aims to address the shadows and transparency quirks. Last year's operating systems didn't necessarily suffer from design problems, I'm told, but rather a not completely baked implementation from Apple's software engineering team.

1:13:56Casey Liss:So this gets to what Marco was saying is like, okay, well, the stuff didn't work well. And so I totally agree that like what they shipped was filled with bugs. I've had to deal with them for an entire year on macOS. They're really bad on other platforms that also have problems. And this is source. This is not just German. So he's saying what he's told and what people internally consider, you know, to be a slight redesign. whoever the internal source uh was that said the system doesn't necessarily suffer from design problems but just did not uh complete implementation why not both as they say uh yeah the implementation was bad and not completely baked i 100 agree you know what else was bad the design as marco just got through and explaining like i i don't know why this i mean i guess i think this doesn't bother people just because they don't see how this design leads to all the problems they're seeing and they think like oh we can fix this by just tweaking things and changing transparency and you can improve things but you know i think i made a list of these three things many months ago when i was trying to boil it down to what's wrong with liquid glass but it's basically the fundamental thing of like um i don't know picture an image editor and you open an image and the image is a picture of somebody and the entire picture fills the window it's like okay there's the picture all right but liquid glass says okay but in this application you don't just look at the picture we want you to have tools so you can like edit the picture or crop the picture or scribble on the picture or whatever and the liquid glass solution to this is take that picture that's filling the window and take a bunch of capsule shaped things and drop them on top of the edges of the picture so they're floating on top of the picture and now you can't see part of the picture because the controls are floating on top of it and then that's where we get into things that marco was talking about okay but what if i want to see the thing that's like the tool palette is blocking the picture.

1:15:44Casey Liss:What if I want to see that? Can't I just scroll the picture so it's not underneath the toolbar? No, you can't, because then what would be underneath the toolbar? So now you've got to do the mirror image thing. And it's like, this is just plain a bad idea. You don't start with content edge to edge and then drop a bunch of capsules on top of it. That's a bad idea. That is fundamentally a bad design for a user interface for a million reasons that we are all experiencing, maybe developers more than other people, but just there's no way around it. You either can't have content underneath the thing, in which case it's not floating on top of the content anymore.

1:16:16Casey Liss:Or you can. And once you do have something floating on top of the content, then you want to show it through. And that's also a bad idea. And then you don't want to be able to see the whole image. Now you got to do the mirror blur thing or just it's just a bad idea. And as Marco said, like this was this wasn't a problem that needed to be solved. This is a new problem that they have introduced with a bad design, which, by the way, also had a bad implementation. Now, in Mac OS 27, it's still going to be the ideal is a bunch of content edge to edge with a bunch of things dropped on top of it floating.

1:16:44Casey Liss:That's not going to change. All they're talking about is maybe we'll tweak some of the blurred edge things and the transparency and the blah, blah, blah. But the fundamental design of liquid glass we are stuck with for probably a very long time. And that fundamental design is bad. It's not the worst thing ever. It doesn't break the computer. You can still use it. It's fine. But it's a bunch of bad ideas. And it just kills me that they're going to spend brain power and years and years of effort to try to take what is a fundamentally bad idea for a user interface and try to make it as least bad as possible when they could have gone in a million other directions that would have been like, let's identify problems that people are having with the UI and solve them versus let's identify something that has heretofore not been a problem and make it a problem and spend the next half decade fixing that problem.

1:17:28Casey Liss:And that is frustrating. But I still, I welcome any improvements to Liquid Glass, but it's, you know, we have to wait for the next big redesign for them to fix this because it's just so fundamental to everything they've done. we got to wait for the stupid 20th anniversary phone. They'd be like, see, this is why they did it. It's like, no, you could have it borrow us in that phone too. It would have been fine. But anyway, yeah, it is something to look forward to is to see how they tweak it. And also honestly, as from a developer's perspective, yeah, please do fix the bugs because setting aside the bad ideas in the design, when things just don't work, that makes everybody sad.

1:18:02Casey Liss:So I hope they do make a lot of progress there.

1:18:05John Siracusa:All right. Beyond, this is still German beyond adjusting the look of liquid glass. Apple will focus on bug fixes, battery life upgrades, and performance improvements. Another new iOS 27 and macOS 27 feature is a potentially helpful addition to Safari that mimics something Google Chrome has had for years.

1:18:21Casey Liss:Oh no, I smell AI coming.

1:18:23John Siracusa:Apple's testing a feature that can automatically organize tabs into groups. When enabled, the system says that tabs will group into topics you browse. The feature isn't labeled as an Apple intelligence enhancement in the system, but it's clearly using some form of AI to work.

1:18:37Casey Liss:I mean, I think that is a useful feature. No one wants a manual. Well, most people don't want to manually organize their tabs. Most people have no organization whatsoever. And if there was a feature of Safari that said, it looks like these are the tabs where you're trying to figure out what refrigerator you're going to buy. And these are the tabs where you're ordering takeout. And you know what I mean? Like that it could put them into groups because Safari has had tab groups and pin tabs and stuff like that for a long time, but only the most meticulous people actually use them. So some kind of machine learning powered sort of, oh, it looks like you're, it looks like you're trying to write a letter, a clippy style thing where it makes a tab groups for you.

1:19:08Casey Liss:Could be cool. this is also a good test of apple's ability not to shove this crap in your face right is it going to be like image playgrounds again and where you're not going to be able to launch siri and without it shoving this you know we can smartly organize your it's like okay okay chill out if it's a feature people want like i know you got to make it discoverable or whatever but like don't i don't need it to be thrown in my face a million times just do a good job on the feature support it and see how well it's adopted. But yeah, I don't, I hope hopefully it doesn't mess up the interface too much, but yeah, there's no escaping, uh, Apple intelligence machine learning features, even in Safari.

1:19:46John Siracusa:So there's been a lot of hardware that we keep hearing has been delayed by Siri. There's the alleged home pod with a screen. There's the God forsaken Apple TV. I am desperate to get a new Apple TV so I can trickle down. I want to trickle down because I have like a three generation back one in the tailgate tub that is also my travel Apple TV. And that's the problem. The current one.

1:20:08Casey Liss:You just bought a new Apple TV for your tub.

1:20:09John Siracusa:I should have. I didn't realize I was going to be waiting 18 frigging years. No, the one in the living room, which is the most modern one, I can't keep the generation straight. The one in the living room is great. Occasionally it feels a touch slow, but 99 % of the time it's great. The one we have in the bedroom, which is one generation back, is 80 % of the time pretty good, if not great. Occasionally it's a little meh, but generally speaking, it's more than enough. The tailgate tub one is, what is that, two generations back, and it's getting real rough down there. And again, the tailgate tub one is our travel one.

1:20:40John Siracusa:And yes, you can laugh at me and say it's unnecessary to bring an Apple TV with me. But you know what? Not only is it unnecessary, but it's something I enjoy. So get off my back. But anyways, the point is I really need that one to get upgraded, which really means I want to do a trickle down. And so please, Apple, for the love of God, I don't think I'll be able to get to MacRumors Buyer's Guide quick enough. but suffice to say it's been like 300 frigging years, uh, since the Apple TV has been updated. It is really preposterous. Oh, I was able to make it four years, 1 ,324 days, October of 2022.

1:21:13John Siracusa:So we are coming upon four years, four years. This is basically half of my youngest child's life. She has only known one Apple TV, which is a phrase I kind of wish I could have back now, but I've already said it. So here we are. anyway uh so maybe an apple tv with an a17 pro or perhaps a revised remote i again i've always been an apple tv remote apologist especially the current one i actually really really like uh so that doesn't bother me any but i know i'm standing mostly alone on that and for those of you who enjoy home pods uh particularly the home pod mini maybe there'll be a new one of those too

1:21:48Casey Liss:you never know yeah so the apple tv and the home pod mini are rumored to have been done for just ages they're just sitting there the hardware's been done for months and months and months but of course they can't ship them like so many of these things like the home pop with the screen which is also rumored to have had the hardware been done for months and months and months because they rely on siri not sucking and having features that they said we're gonna be ready but aren't and so the hardware has just been sitting there gone i mean i guess they could have shipped things like the apple tv without it and just you know but they've just like we're gonna it's a flagship feature of the you know tv os 27 it's got you know it's all of a piece so a lot of hardware has been delayed and if you're gonna delay some hardware this is probably the hardware to delay it's fine especially like home pop with the screen which i don't know if people are clamoring for but the rumor is that uh now that the 27 os's are going to be announced this hardware is free to be released the question is will apple mention any of this hardware at wwc some of the rumors say yeah they're going to introduce all of this hardware at wwc because why wouldn't they it's been sitting around it's totally ready to go they probably already have all the promo it's like why wouldn't they launch them but then i look at this hardware i'm like when apple announces hardware at WWDC, they usually like to have some kind of like flag bearer, really impressive, exciting thing like MacBook Pros or something, which are, you know, again, those are the OLED ones are not coming out until much later.

1:23:01Casey Liss:So they're not going to be announced. But, and I don't think any one of these devices or even all three of these devices are significant enough for them to give them a slot at WWDC, but who knows, maybe they don't have a lot of other things to talk about. Um, so I'm not actually expecting these to come out at WWDC, but I am expecting them to come out shortly after because or at least be announced shortly after i think they can't ship obviously until the 27 os has actually come out so maybe fall ish um but yeah the one i'm the one i actually am most excited about is the apple tv not because of the trickle down reason but just because it is old and the vague rumors of a possibly revised remote even if it's just like we made one of the buttons microscopically concave when before it was microscopically convex or whatever they already do that on the buttons but it's so slight anyway any change that remote would be good for me it is much better than the remote it replaced but i still feel like there's tons of room for improvement and i have frequently found cursing that remote because when the skip intro button comes up and you know it's only there for a brief period of time i have a short window of time with which i have to grab the apple tv remote off of whatever end table that it's on and i have to press the one and only largest button on this remote to activate the skip intro button and my success rate at that is terrible and i'm i refuse to blame myself i'm a good video game player i can press buttons why is this so hard because i haven't fully disabled the touch interface and why have i not fully disabled touch interface and settings because it is actually the fastest most efficient way to navigate the ui but it also totally destroys my ability to press buttons without accidentally swiping or something so i really don't like the remote even though it is better than the one that replaced it and i'm looking forward to any mild revisions and hey a17 Pro.

1:24:43Casey Liss:And, you know, I think they need that one because it's, it'll run quote unquote, Apple intelligence, whatever the heck that means. But maybe that will mean, uh, when I talk to my Siri remote and ask it to do TV things, it'll be slightly better. It's already pretty good because it's, it's problem domain is so narrow. Normally when I say, you know, press the remote and say, watch, blah, blah, blah, it does a pretty good job of finding that maybe if they make it LM powered, I can have more sophisticated conversations like show me the, you know, watch blah, blah, blah, but show me the version I bought on the iTunes store and not any of the other ones.

1:25:10Casey Liss:And it will understand that when I say it just like that. My Amazon puck can do it. So I'm looking forward to that. And I'm not particularly excited about the HomePod with the screen, but who knows? Dazzle me. But again, I don't think this stuff will be announced. What do you think? Will there be any hardware at WWC? And if so, will this be among the hardware?

1:25:29John Siracusa:The only reason I can imagine there being hardware, and I think Upgrade was saying this this week, is that that would give a convenient excuse for Ternus to be emceeing for at least a portion of it and maybe we can look at the tea leaves about uh the way this is going to be handled we should talk about that before we end the segment but i could see that as like a convenient excuse but leaving that aside no i don't think they're going to talk about hardware yeah and

1:25:56Casey Liss:again i didn't gather up all the rumors there may be some other rumor however as far as i know like you know nothing nothing big it's just these these ones are people rumoring these ones because they're ready and just because they're ready doesn't mean they're going to be announced or you know, a WDC, because again, they can't ship until the 27 OS is around and that's not going to happen until fall at the earliest or not fall. When does iOS usually come out? Is it fall? Yeah. It's usually, you know, early September. Yeah. It's early fall versus sometimes later fall. Mac OS will come out later fall, but unless they just ship it when at the state it's in an early fall, like they did with 26.0.

1:26:26Casey Liss:And that was terrible. I mean, we know they're going to ship it. They're going to ship it with the, with the new iPhones. That's it. Like you said delay. Yeah. I'm saying for Mac OS, they used to delay Mac OS until like October sometimes, but this most recent year, I feel like they didn't do that.

1:26:38Marco Arment:Nope. They, I mean, this is, you know, modern Apple. They, they ship whatever they've got at that time, whether it's finished or not, they shove it out there.

1:26:46Casey Liss:I have a, yeah, I have a, I was gonna say constants, but I have, I have let vars in my code for Mac OS 26.0, Mac OS 26.1, Mac OS 26.2, Mac OS. These are all, I have conditions on all of these things in my code to say, okay, in 26.0, you need to do this, but then 26.1, you need to do this in 26. Because of like tweaks to metrics and controls that didn't work. Yeah, the 26th OS release has just been a nightmare in terms of how the bugs have shifted and moved around in macOS using SwiftUI or whatever. So I'm looking forward to 27 not repeating that. And maybe I can just have one thing that says 27 or later.

1:27:23Casey Liss:Imagine that. All right.

1:27:26John Siracusa:Let's talk names. Mac Rumors had noticed way back when that a bunch of shell companies had registered, I think trademarks, it doesn't really matter, but had done something to just basically pee on and claim as their own a bunch of different names. And of those that have been used already or taking away the ones that have been used already, MacRumors has come up with the following list as potential names. California, Condor, Diablo, Farallon, Grizzly, Mammoth, Miramar, Pacific, Redtail, Redwood, Rincon, Shasta, or maybe Shasta, I don't even know, Skyline, and Tiburon. of these options are there any that jump at jump out at you as absolutely yes absolutely no or

1:28:09Casey Liss:hmm maybe well so to begin with uh i want to say that i hope that they continue to give mac os names they don't have to be california names but it's one of the last places of like fun for the sake of fun in apple's products everything else is so regimented especially now they're all in the same number and everything the big cat names were fun california place names are less fun maybe be the big cat names but still fun and this is one of the you know the one of these sort of like uh inconsequential things that's going to be announced at wc is what is the new name of mac os it was more fun when mac os was in better shape and we could just enjoy this as a frizzy thing on top of an os that we loved and now we're just kind of all grumpy about the glass or whatever but still this is a fun part of of the announcement and again i love the fact that it has no consequences now this list as you noted there's been a bunch of names that were in this trademark list of these shell companies getting trademarks for apple that they did use like sequoia was on there i think sonoma i forget which ones have been but they did use some of these but i think also they have had mac os releases that didn't come from this list so i'm not entirely sure that the new name will be on this list it could be something that's not on this list at all in particular the first name that you read is interesting where i feel like uh if they ever use mac os california that's the end of the california names that's the last one like once you use that there i feel like they would announce uh mac os california and this will mark the end i guess they would announce that they would just find out next year but surely that one has to be

1:29:36John Siracusa:the last one if they ever use it i don't know i could see them i know they don't typically look back that much but i could see them saying you know this is the 50th anniversary of the company or 50th birthday whatever you want to call it of the company so we wanted to do something big

1:29:49Casey Liss:uh let's call it mac os california yeah and i have as much as i just like liquid glass mac os 26 should have been skyline right the code name was solarium or solaria or whatever it was the code name was like a sun sky based word it should have been skyline i'm glad they didn't use it because because i don't like mac os 26 and so i'm glad they didn't use skyline for it but it was like a a glass os sky glass i know skyline's a place or whatever but it's a chili chain i like skyline i that's one of my favorite names i don't predict that it's going to be used but on this list i like that one and also the other one on this list that uh is funny to me is uh farallon i'm pretty sure like so many things from my youth like that i didn't know i'm pretty sure that there was a classic mac os networking company with that name uh that i read in like mac magazines when i was a kid only when my child's brain saw that word and just scanned it it pronounced it internally as let's see falron which is not how you pronounce it it's f-a-r-a-l-o-n but i had no idea there was a place with that name i just knew it as a networking company that i internally mispronounced because i had never heard anyone speak that name i think until you just said it right and so my whole childhood i'm like it's one of those falron networking things or whatever nope it was clearly a place that was probably from that part of california and name themselves after it.

1:31:12Casey Liss:And, uh, you know, I don't know the names of places in California. I think we've all proven that we don't know these names. Um, although I did know Mavericks. Uh, but yeah, Farallon is, uh, probably my bottom choice because I can't pronounce it. Uh,

1:31:26John Siracusa:for me, I could, like I said, I could see California being a thing. I don't know where skyline California is. Uh, I don't know anything about California geography.

1:31:33Casey Liss:There was that movie with the aliens invaded. It was called skyline. I thought it was because like all the aliens are coming from the sky. No, it's a place. Well,

1:31:39John Siracusa:it's also skyline drive in Virginia. and I think it's actually Skyline is only in Virginia.

1:31:43Casey Liss:Oh yeah, I'm sure these names are used elsewhere, but apparently there is, let me see, this is Skyline, Skyline, California.

1:31:49John Siracusa:Either way, I view Skyline as a Virginia thing, so I both love and hate that as an idea because you know that all the obnoxious Californians would be like, oh, it's really about California because California is the only place that matters. But I do like Pacific. I like Mammoth and Grizzly. I do think Redwood is pretty good, but wasn't that used for like one of the Windows versions or something like that? What am I thinking of that that was a code name? Yeah, yeah.

1:32:12Casey Liss:Not an official name. I mean, they just did Sequoia. I feel like Redwood is.

1:32:16John Siracusa:No, that's fair. That is true. I don't know. I think there's several.

1:32:19Casey Liss:Apparently Skyline. I'm looking up Wikipedia. Skyline is a hilly urban neighborhood in San Diego, California. There you go.

1:32:26John Siracusa:Today I learned not as good as Skyline Drive. Also, I think Marco had mentioned this quietly under his breath, which may or may not make the edit, but a very good chili, as it turns out. Yes. Nevertheless.

1:32:35Casey Liss:Very good chili. Isn't that the one? It's like chili on spaghetti.

1:32:39John Siracusa:Yeah. Yeah, or maybe vice versa, but yeah.

1:32:41Marco Arment:I've actually never been to a Skyline Chili, but I grew up in Columbus, Ohio. I think they're based in Cincinnati, or at least they're popular in that area. So they were kind of all over the place, but I never actually went into one.

1:32:51John Siracusa:Yeah, I've had it. Some very dear family friends, the wife in the family is from Ohio and occasionally will make Skyline Chili, and it is delicious. Anyways, I don't know. I think there's a lot of decent options on here. I do agree with you, John, that the California names are fun. they're not as fun as the big cats, but there's only but so many big cats you can choose from. And I do like that they have some sort of thing other than just a numeric version number, even though as we've proven, I think it was like a year ago now, maybe two years ago, that Marco and I have no freaking clue what came when or anything like that.

1:33:28John Siracusa:But still, I do like them. Let me start with some negative. I mean, I think we've maybe already covered this, But is there anything else that, Marco, let's start with you that you're not looking forward to? What are you dreading about next week?

1:33:41Marco Arment:So in general, Apple WBCs are a combination of exciting potential and homework. And the balance of that shifts year to year. What I didn't like last year was that Liquid Glass redesigns brought a ton of homework, and I wasn't super fond of much else besides the transcript API, which admittedly, that was a really big deal for me. But the rest of iOS 26 mostly just brought work that developers had to do, because redesigns do that. This year, it's more likely to be a better balance of that, because it's not a redesign here. And so what I'm hoping for this year is hopefully some design calmness in the sense that they're not going to hopefully redesign things to such a degree that we would need to redesign our apps again.

1:34:44Marco Arment:And also maturation and expansion of a lot of the on-device model stuff. I was very happy to get those APIs last year, but they aren't really that useful yet because of just inherent limits in the on-device models and everything. And also, we were all so busy dealing with liquid glass for the first half of the year that we couldn't really build against the on-device models that much because we were just underwater doing liquid glass stuff. So what I hope to see this year is, again, a better balance of new capabilities and new stuff we can do versus keeping up and treading water that we have to do to keep up with the system.

1:35:28Marco Arment:Also, adopting and maintaining a liquid glass UI is especially cumbersome because things are so different from 18 and whatever came before. it's hard to maintain the code basis for both. It's hard to make your app adopt Liquid Glass, especially adopt it well and not just to a surface level, but also maintain compatibility with iOS 18. What I want this year is for things to move forward enough that we can drop compatibility for iOS 18 in our apps for the most part. And that requires a lot of just stability, reputation improvements for the OSs, tweaking them in such a way that a lot of the haters who are holding on to 18 and sequoia can comfortably move to them and will happily install them pushing that base forward i think is the biggest thing they can do for us so what i am dreading to answer your question is if they don't do that if if what they ship is just another like you know butterfly keyboard gasket on top of liquid glass like oh we fixed it and it's not really fixed and it's just like some token tweaks and it's not anything actually meaningfully better.

1:36:40Marco Arment:That would be, I think, the failure mode here. I don't know how likely I am to get what I want design-wise, but I think I probably will get what I want technically. I do expect the foundation models frameworks to be a little bit better, a little bit tweaked, a little bit improved. And then the question is, this entire world of app intents that we have been promised since the 2024 vaporware demo, The idea of your apps that can vend things using a new app intent system to the OS and have Siri tie them all together, that is a giant question mark. Are we going to actually see that? Is it going to actually work?

1:37:22Marco Arment:Is it going to be good? And what will our apps be able to vend? And how will that system work? Is it going to be like the system spotlight database, which it's built upon, which does not work very well? um i i hope not or i hope they have made spotlight work better somehow even though they never have achieved that see also siri like there's a lot of these things are kind of like built on on shaky foundations or or likely to be built in shaky foundations that i am a little bit worried about but ultimately i hope that they give us a good amount of new capabilities to use without it being full of a bunch of grunt homework on shaky crap that doesn't work.

1:38:07Marco Arment:Because that's what we had last summer and it sucked. And it was really hard for us to develop good features and move our software forward. So this year, what I'm hoping for is the opposite. Give us decent stuff that works well that we can use to make better apps. John?

1:38:21Casey Liss:Yeah, I think the thing I'm dreading most is kind of like Marcos had a repeat of last year. In my short five or six year career as a mac app developer uh this year dealing with mac os 26 has been by far the the worst stuff that i've had to deal with i just spent so much time uh fighting with like small display issues with basic interface elements that i didn't control um like i for example i use uh on mac os there's like store controls where you can just say like oh so you have a bunch of subscriptions in your app or a bunch of in-app purchases we have a view for you that shows them the apple way and i was like i'm definitely using that because i'll get past app review i'm like look i don't even control this view it's apple's view i just say put store view here and it just reads all my metadata about my things and blah blah and then i have to worry about oh your button has the wrong text or you didn't put the price in the right place i'd be like look i'm not even making this ui it's the apple ui and i thought that was so smart when i rolled out my app with an app purchase in mac os 15.

1:39:23Casey Liss:And then 26 came and broke that window over and over and over again by making it clip things or show embarrassing scroll bars or make the buttons really tall or make it so you can read all the text. And, you know, this is another example of not my fault, but definitely my problem. I just didn't want to ship an app that looked embarrassing. Even it was like, well, it's usable, but like, why is there a scroll bar with like two pixels worth of travel in this window that you can't resize? I'm like, well, I don't know why this scroll bar is there. It shouldn't be there. it's a bug in apple's thing and so i had to do just the most herculean stuff to that's why i had all those things of like 26.0 this is what i need to do to make it work 26.1 oh they changed it but now i need to do something else to make it work 26.2 now i need to do this i even as i mentioned on a past episode have branches in there for os version but also on intel versus arm why should swift ui view from apple be different on intel versus arm i just experimentally i have determined on mac os 26 point whatever on intel looks like this and on mac os 26 point ever on arm looks like this luckily i have multiple machines to test with but it's just that was such a waste of my time because was that adding features that helped the user at all no could i i could have ignored it but then when people would go to the purchase screen on my app they would see something looks like someone doesn't care about their apps like why does this look janky why is this button the text and this button truncated why is there a scroll bar that scrolls two pixels i feel like that's an impression i didn't want to make in my apps so i spent so much time fighting with the new look of liquid glass and mac os 26 for basic controls that really have nothing to do with the functionality of my app i also fought with it elsewhere my setting screen which does have to do with the functionality of my app had similar problems and i basically did six months worth of bugs and bug reports on that by the end they pretty much got them all fixed i think i have you know the the case where you're using whatever we're on now 26.5 i think it has maybe one or two little hacks in it but kind of like marco was saying all that code to make it look right on 26.0.1.2.3.4 that's all still in the code and has to be because what if someone is using my app on 26.0 i don't want my app to look bad i can't clean that code up until i drop support for 26 which is not going to happen for a long time so yeah that's what i'm dreading a repeat of that and i don't honestly i don't think that will happen again it shouldn't happen 27 should basically be the same as 26 but with some minor tweaks that don't affect anything and plus a bunch of bug fixes so i'm not expecting it but dreading in terms of like i have ptsd from from liquid glass on the mac i don't want to have to go through that again um and then well you should go to your thing casey What are you dreading, if anything?

1:42:04John Siracusa:You know, I don't feel like I'm dreading that much. I feel like, as I said earlier, I am dreading that if they really get, you know, they get obsessed with AI stuff. Like, you know, oh, it's AI this, AI that, AI this, AI that. Like, I do not want to see that. I'm over it. I don't want to.

1:42:22Casey Liss:Are you dreading that in the presentation or in the products or both?

1:42:25John Siracusa:Yes. I think either way it could be the case. um what i'm i'm also dreading that and the apple has gotten better in the last few years but still is a long way to go i'm dreading looking at the documentation for all of the new stuff because you know it is likely to be missing or trash they have gotten better they still got a long way to go so i'm dreading looking to that looking looking at that um what i'm looking forward to though is, first of all, we haven't really had a chance to think about what the presentations are going to be. But when I was listening to Upgrade a few weeks back now, something that Jason or Mike said occurred to made me think that the likelihood that they do some completely corny and cheesy and cringe changing of the guards pre-video for this, you know, so like, you know, handing off from Tim to John.

1:43:21John Siracusa:I think it's going to happen. And I'm actually kind of excited for it, if I'm honest, because I think it could be really fun.

1:43:26Casey Liss:I thought you were going to say it wasn't likely to happen and you were glad it wasn't going to happen. But no, you went the other way. You think it's going to happen and you're excited for it.

1:43:33John Siracusa:I think it could be kind of cute. I really do. I might be the only one.

1:43:36Casey Liss:I don't think they're going to do it and I'm not looking forward to it, but I don't, you know, it's fine. I don't care about that stuff.

1:43:42John Siracusa:But no, I mean, I think in terms of things we're excited for, which I think we've all kind of bounced off that as well, you know let's let's go from not so happy to happy i would love to see some advancements in swift ui there's no like particular bugbears off that come to the front of my mind is specific things i really want them to change a better search support would always be great because i've had to do custom search related stuff in call sheet for its entire life for various and sundry reasons to have some better first-party support for doing things more flexibly would be great i think a lot of swift ui you could just say, I would like to do blank in a more flexible way or allow, you know, be allowed to do whatever, but have more flexibility with it.

1:44:22John Siracusa:Um, I think that's true of most of Swift UI. I actually, and again, I know a lot of the older, like the, the spiritually older folks online are very anti Swift UI right now. I'm like masked on whatever. I actually really like Swift UI. I think that if call sheet were to be written in UI kit, that it would look one half as good, or I'd still be working on version one. And that's perhaps an indictment of my own skill. And maybe I'm okay with that. But I really, really like SwiftUI, but it also really, really has problems. And so I am not calling for a whole cloth, like let's go a different direction with SwiftUI.

1:45:00John Siracusa:And I've seen some calls for that on Mastodon. I would really like to see, and I'm excited for seeing what improvements come to SwiftUI on basically any of the platforms, even if it's one that I don't really use, myself like if i i would love to see john get some swift ui improvements for mac os because that will eventually help me when i'm using john's apps or other apps that are mac os apps that are based on swift ui so i'm excited to see what they do with swift ui this year that's likely to be more of a state of the state of the union thing um than a keynote thing but i'm still looking forward to it

1:45:32Casey Liss:well let me let me talk about what i'm excited for because it connects to what you were saying about swift ui this is phrases like what are you excited for i'm not actually excited for this because i don't think it's going to come but if it did come boy would i be excited about it and exactly what you got out there's functionality that's still missing from swift ui on the mac there's been a couple of uh complaining posts about it recently but it just really just again ptsd reading that person's description of all the different drag and drop systems and how none of them are fully functional i lived it friggin lived it with switch glass from from like day one of swift ui when i should not have been using it on the mac having to roll this all on my own and here we are years later it you still cannot implement basic drag and drop with all the features you would want on the mac and it's it boggles the mind they're on like the third or fourth try at doing drag and drop just today i was messing with dot file importer which i don't know if there's an equivalent of that in ios maybe not but basically open save dialogues how do you do an open save dialogue box in a mac os swift ui app they have features for it i don't think they have modifiers and stuff for it but they're fairly limited there's a bunch of stuff that you can't do in them that you can do with the app kit ones and also buggy because if you have like oh i'm gonna do a file importer on this view but there's a file importer also on this other view it's like oh now they fight with each other for reasons you don't understand no i don't understand i don't like you end up making like an invisible clear one pixel background view to attach your file importer for because that's the that's what makes it work like stuff like that should not happen this far into swift ui so i would be super excited if they came out with swift ui advances that say hey missing functionality it's here now things that didn't work and were buggy they're fixed but i'm not expecting that yeah but what i am excited for which i know is coming and i know this every year is because uh it's developed in the open swift the advances in swift that are happening in swift evolution you can look there's no surprises it's been happening all year it happens in the open we know every single thing that's going to come or pretty much every single thing that's going to come in swift and i can tell you having watched a year of advances there's a bunch of really good stuff in particular uh the you know whole like um uh could not type check this expression in reasonable time sometimes that's not what it means but for the times when it really does mean that it couldn't type check that expression in reasonable time that's going to get way faster in the new version of swift and that is just a quality of life improvement for everybody who's using swift and all the other nice features they're added like i'm always excited about that and it's kind of anticlimactic because if you follow swift evolution you know all the features but it's also exciting because like i've been seeing people discuss these features over the course of the year and argue about them and seeing whether they get adopted or whether they have to go back to be modified and come back in and then wwc is like they're they're coming out it's like here are the ones that graduated you know here are the ones that graduated to actually be features of swift 6.4 or whatever version we're on now um and i can tell you the features are good and i i'm also excited for an xcode built on top of this version of swift that performs better because of these improvements and i suppose also a new version of xcode that has better lm integration and stuff like that but i am excited about that and it's easy to get excited about that because you know it's coming like it's it's one of the the few things that apple does in the open where you don't have to guess uh what's going to arrive um the other thing i'm excited about i think i said this when we talked about this at the beginning of the year i am actually excited to see the features that didn't ship in 2024 actually arrive in a usable form because i think they're good ideas and I think they will be useful to me as a user of my phone if they work like like Marco said the way we know they can we know it's possible we see the technology working in these very limited scopes I'm not asking for the world I'm just like Marco said I'm just asking for Siri to act like Apple has been advertising it acts like for the past decade that I'm actually really excited about I know it's boring and it's not like well does that mean Apple is okay and they're competitive and setting all that aside i'm a person with a phone and i want to be able to do those things with my phone more reliably and every other device in my life that has been hooked up to any kind of llm powered engine has gotten better for me in that regard and that it understands what the heck i'm saying and responds in a reasonable way i'm so excited for my phone to get there i don't think i'm going to run a beta of ios because i'm not that excited about it but i'm excited to just get over that hump and it seems like this is going to be the year marco anything that you

1:49:47Marco Arment:are particularly excited for i've kind of woven it through my my commentary the rest of the episode but yeah just in in general what i want to see is you know again like what what makes for a good wbdc is they show off features that a will ship b work well or at least work and see that I can use to make a better app. That's like, you know, because there is the consumer part of the WVDC announcements of like, here's what all of our new us's are going to do. And that's great. But it is like technically still developer conference. And I am technically still developer. And what I would like most of all is, yeah, Apple, you do good stuff, but we know your stuff is going to like be a little bit half-assed, let's be honest.

1:50:38Marco Arment:because that's what they do. As we were saying earlier, they put out some stuff that kind of works and walk away whistling and call it a day. But where their platforms really shine, despite them not acting this way, is with third-party apps and what we do for them. And the better tools they can give us, the better we can make our apps. So that's what I'm always most hoping for every year is give us tools so that we can make better apps. that's the most empowering that's what has the most impact and that's what ultimately pushes their products forward yeah yeah i agree uh john you put in here what will we miss the most i'm

1:51:19John Siracusa:not sure where you were going with that uh but i have a couple of things you're not sure well i have a couple of theories uh number one uh i will miss seeing the two of you uh obviously we like i said earlier we didn't really get the nod this year didn't at all get the nod this year we never really understand why we do or don't get the nod but we didn't uh so we're not going to be seeing each other, unfortunately. And that makes me sad. I will miss that the most. I'm guessing, John, you will miss being able to use your piece of crap computer with the newest operating system the most?

1:51:45Casey Liss:No, what I'm going to miss the most at WWDC is the formerly planned announcement of the new Mac studios, the computer of supposedly going to buy because before the whole AI is using all the RAM and everything in the world thing, Mac studios with presumably M5 family of chips were going to be announced at wwc and now they totally aren't for reasons that are understandable and technically are not really apple's fault but really suck for me personally so what will i miss the most the mac studio sad me and obviously you know i you know missed the mac pro a little bit but it was kind of a mercy killing but like that was that was my prize was like okay well the whole smack pro saga is over got my shirt with the the dates and the little tombstone it's like it's sad blah blah but guess what this is the year that i'm gonna buy a new computer and they're gonna roll it out of wwdc and no they're not i mean they can surprise me that would be super amazing if they come up with like an m6 max studio and they're like somehow we got ram chips i'll be ecstatic but i just don't see that happening so yeah that's what i'll miss the most my max studio and also you too but yeah i won't i won't miss a six hour plane flight that is true i will not

1:52:53John Siracusa:miss that uh but i love that we rank marco and i rank below the max studio the computer that you didn't even really want in the first place really really feeling great over here i talk to you two

1:53:03Marco Arment:every week but you know where's my mac studio he really wanted better podcast co-host but he stuck

1:53:08Casey Liss:with us right exactly where's the mac pro podcast so so you're gonna be riddled with holes oh my

1:53:14John Siracusa:gosh i don't think we should take this any further marco anything you'd like to add or are you good i am good all right but i think uh that's pretty much it is there any uh stagecraft or anything

1:53:23Casey Liss:you want to talk about or are we good stagecraft i mean like uh the things you were talking about like what kind of weird skits are they gonna have and stuff um i guess yeah i guess yeah that is a one other topic um the question of the balance between tim and turnus i think we're going to see

1:53:38Marco Arment:a lot of both of them i think tim is going to open and i think he's going to hand it off to

1:53:42Casey Liss:turnus all the time yeah i think and you think turnus is going to be talking about things that

1:53:47Marco Arment:are not hardware related um that's a good question yeah because like normally federighi would take a pretty big role in this i obviously i think i think federighi will still be in this too you know know prominently in this but maybe like the kind of mc role that used to be just tim i think maybe maybe he will share that with turnus you know yeah we'll say that it'll be interesting i mean

1:54:10Casey Liss:it's not like it's a mystery like we know the transition is happening we know the time we know everything about it and honestly how they divide up responsibilities is not even that important because who cares it'll be fine either way um but it will be i mean it may be our first opportunity to see Ternus talking about something that's not hardware. I'm not sure that will happen, but if it did, it would be our first look at CEO Ternus. You're no longer confined to your world where you just talk about whatever the hardware product is. Now you can talk about whatever you want. You can talk about the new Apple TV show, Ternus.

1:54:40Casey Liss:You can do anything. You're the CEO. Or soon, anyway.

1:54:43John Siracusa:Yeah, I'm really actually surprisingly excited to see how they spin this. Well, the spin implies that it's bad, but you know what I mean, like how they handle this transition. And I suspect it's going to be mostly business as usual and with perhaps a bit more Ternus than we had seen in the past. But not I don't think it's going to be like Tim says hi and then turns the entire thing over to Ternus and then says goodbye. I mean, it could be. It's been similar to that in the past. I don't think that's what's going to happen, but we'll see.

1:55:13Casey Liss:Do you think there's going to be a Tim goodbye? Because this is like his last basically Apple event before he's not a CEO anymore. Do you think he'll have any kind of like little nod to that or he'll just like do it by the book and be out?

1:55:23John Siracusa:I think there may be a nod to it, but I think it won't be much more than that. I think at the least, it'll be a kind of a wink and a nod sort of thing. Not literally, but you know what I mean. And at most, it'll be something like, thank you for the time that I've spent with you and blah, blah, blah. especially actually and i could very much see him couching it as and to the you know staff at apple to the employees at apple thank you so much for our you know 15 years or 10 years whatever it's been together um it's meant the world to me blah blah i could absolutely see him doing that yeah

1:55:54Marco Arment:well so that they'll be watching for that um i don't know i i would be surprised if he confirmed this was his last event that doesn't that's not really his style it's not really apple's style i i would be very surprised by that well we'll see what happens i mean he could just

1:56:09Casey Liss:say something nice and not mention anything about it being his last event i'm just like huh why did he say that nice thing about employees and thank you um the one way where we could see a ton of turnus quote-unquote legitimately or normally is if all that hardware we described the home pop of the screen the apple tv and the new home pod mini if that all comes out yeah have turnus do all that because he would have done all that anyway like he's the hardware guy you know he would have been the the top level mc for that handing off to individual ones but if that stuff doesn't arrive i'm i don't know where you would work in turnus from a hardware perspective so i I guess we'll be looking at that, but yeah.

1:56:41Casey Liss:And the other thing to look for, I guess is, is there any change whatsoever in Ternus's demeanor? I don't think there'll never be a change in Tim Cook's demeanor. He's going to be the same forever, but will Ternus like, will we see anything different from Ternus? I don't think we will because he's Tim's cook is still the CEO and he's still kind of in the mode of like, I don't quite have this job yet. Just keep, keep on keeping on. But, uh, I'm always watching for that. I want to see him spread his wings and come out from the shadow of Tim. And maybe that won't happen this June, but I hope it'll happen eventually.

1:57:10Marco Arment:All right. Thank you so much to our sponsors this week. Factor and Delete Me. And thanks 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. This week in Overtime, we're going to be talking about the newly announced NVIDIA RTX Spark, which could be a challenge to Apple Silicon. We'll talk about that in Overtime this week. You can join us at atp.fm slash join. Thank you, everybody, for listening, and we'll talk to you next week. Now the show is over. They didn't even mean to begin because it was accidental.

1:57:48Marco Arment:Accidental. Oh, it was accidental. Accidental. John didn't do any research. Marco and Casey wouldn't let him because it was accidental. Accidental. Oh, it was accidental. Accidental. And you can find the show notes at ATP.fm And if you're into Mastodon you can follow them at C-A-S-E-Y-L-I-S-S So that's K-C-L-I-S-S M-A-R-C-O-A-R-M N-T-M-A-R-C-O-R-M-N S-I-R-A-C-U-S-A-C-R-A-C-U-S-A It's accidental They didn't mean to Accidental Tech podcast so long

1:58:40John Siracusa:I've nerd sniped myself with call sheet. I feel like, generally speaking, my work efforts are fairly static. You know, I work roughly the same hours every day. I feel like I get roughly the same amount of output out of me every day. But then I'll get this burr up my bottom, like, oh, man, it would be really cool if I could do blank. And when that happens, I cannot stop until it's done. And I feel like, Marco, you definitely have talked in the past about some of these. I don't know if manic is really the word I'm looking for, but these, like, hyper intense moments where you're just absolutely fixated on something.

1:59:17John Siracusa:And maybe, John, you go through that, too. I don't know. But I feel like I have a kindred spirit in Marco, if nothing else, on that.

1:59:22Marco Arment:Oh, yeah. This is much of my life.

1:59:24John Siracusa:Yeah. But I think the difference between Marco and I is the way you describe yourself anyway is that it's very big fits and very big spurts. I always forget which one's good, which one's bad. But either way, it's very up and down, whereas mine is mostly level. And then I'll go through these moments that it's like, oh, my God, I cannot peel myself away from the computer. That's one of these times. The most recent time that this happened was – and this is going to be something that we could let ourselves get sidetracked about. I'm really going to try not to. Maybe we should talk about this topic another time.

1:59:52John Siracusa:But I am mostly divorcing myself from Plex, and I have started to lean into it. Whoa. At least for my own use, I've started to lean into Jellyfin, which is a purely open source equivalent. I do think it is worth discussing this another time, but that's not terribly pertinent to the conversation, except to say that I had added support in Call Sheet for getting information from Jellyfin as to what's currently playing. And so when CallSheet, not long after CallSheet launched, I forget exactly when I shipped it, you could, it would sniff out if there was a channels server running in the local network, or really actually not even a server, excuse me, but a client running the local network, or if a Plex client was running the local network.

2:00:38John Siracusa:And it would do its best to interrogate those clients and figure out what are they playing and do the best it can to put that information by default on the very top of the very first screen you see when you open the app. And it was okay at first. It got a little better over time. With channels, it uses Bonjour, which is the zero config stuff that's baked into all of Apple's devices. It's an Apple technology, if I'm not mistaken. And it works really well. It's pretty consistent, and I really like the way that John over at Channels has handled that. Plex, however, what they have is kind of spiritually similar to Bonjour, but it was written about 950 years ago, and it shows.

2:01:23John Siracusa:And so, apparently, I just recently found out that the internal name for this, or allegedly the internal name for this, is G'day Mate, or GDM, which I find hilarious. But what you do, and I haven't looked at this piece of code in a little while, so I might get the detail slightly wrong. But what you do is you have, or CallSheet will broadcast a multicast message that says, hey, are there any Plex clients out there? And if a Plex client hears that message, it will reply and say, yes, I am a Plex client. Here I am. And here's my connection information. and then CallSheet will then go and connect to that Plex client directly and say, hey, what are you playing right now?

2:02:03John Siracusa:When I first launched this feature, I don't want to say like two-ish years ago, it worked okay. This is very old, very rickety tech. If you're the kind of dork that has like a VLAN in your home network, and let me tell you the kind of people, the kind of dorks who run Plex are the kind of dorks who would have a VLAN. And if you don't know what that is, congratulations, you're normal. But if you have like a VLAN or something like that.

2:02:28Marco Arment:How did that avoid, you know, like we pronounce it vlog. Why don't we say VLAN?

2:02:33John Siracusa:Because it's too close to Vlad. I don't know. Anyway, if you have like a VLAN or something like that, it doesn't work. Like in so many ways, it doesn't work. And then as Plex is in poopified over the last year or two, again, we'll talk about this another time. As Plex is in poopified and had dramatically rewritten their client apps, a lot of these features got even more rickety than they already were. which is saying something because they were pretty freaking rickety. The problem I had though is that for whatever reason, for right or wrong, I was staunchly, devoutly against doing any sort of thing where you have to log in in order to get information from these playback clients.

2:03:13John Siracusa:I didn't want to have to do like an OAuth flow. I've done it in the past and I hate it. And I'm sure there's packages for it now, but I just didn't want to bother with it. I didn't think it was fun for the user. I didn't think it was nice. It didn't feel good. And so I didn't want to do anything that even vaguely relied on login. So then a few months ago, I started to really try trial Jellyfin and really try to figure out, okay, can I make this work in my life? And I have decided that at least as I sit here today, I am pretty much all in on Jellyfin. I still run my Plex server, and the family is still on Plex.

2:03:49John Siracusa:But for me personally, for me, myself, and I, I'm pretty much exclusively on Jellyfin now. I will say that the Jellyfin client apps are all pretty much terrible at this point. All the Infuse people are yelling at me and screaming at me. I have tried Infuse. It was not for me. I don't remember why. It doesn't matter. But there is a client that I enjoy called Send Player. I'll link it in the show notes. It is written or based out of China, so take that for what you will. But it is very affordable, and it's a one-time purchase. But anyway, so I'm all in on Jellyfin, and I wanted my playback information from Jellyfin to be showing up in my app in CallSheet so I can just tap and see what I'm working on or what I'm watching.

2:04:33John Siracusa:And it turns out that there is no sort of peer-to-peer way to do this with Jellyfin. So now what I needed to do is I needed to do some sort of like login flow. And I asked Claude about this, former and I think future sponsor. I asked Claude about this and Claude said, well, hey, here's what you can do. You can do like a whole OAuth thing if you so desire. But you can do, and I forget what Jellyfin calls it, but you can basically do one of those things where you ask the server for a PIN number. and then the user goes to the server and enters that PIN number to complete the handshake and say, yes, this thing is blessed.

2:05:09John Siracusa:And so that's what I did. And that's already shipped in CallSheet. It shipped, I don't know, like a month or two ago. And it's pretty good. I like to think I made it as clean and easy as I possibly could. And so now when you start up CallSheet, it is almost freaking instant that you can see what you're playing in Jellyfin. What's also great about this is that because you're not doing some like multicast, peer-to-peer sort of scenario, it will work depending on your setup, of course. It could even work remotely. So the peer-to-peer thing for both channels and for Plex, that only worked if you were on the same network because it was all multicast stuff.

2:05:47John Siracusa:Well, I have my Jellyfin server exposed using Tailscale and specifically a Tailscale service. I'll put a link in the show notes. But basically what that means is in the same way I was talking about funnel at the top of the show that lets you expose something to the entire internet that's on your tail net serve is kind of like doing the same thing but only within your tail net so you have to be attached to my tail net in order to access my jellyfin server but it has a full https certificate you know it's it's it's got a easy to remember domain name etc etc well what's great about all this is when you enter in, you know, jelly.you know, whatever your, if tailnets name is.ts.net, well then, you can see that information even when you're watching something remotely, which is really cool because it'll reach through tail scale and ask your Jellyfin server, hey, what are you playing right now?

2:06:38John Siracusa:This works super well. And it works so freaking fast. It's incredible. By the time call sheet is spun up entirely, typically what you're watching Jellyfin is right there, front and center, right at the top of the screen. I love it. I genuinely do. But all the Plex people, of which I sort of kind of, but I'm mostly not one anymore, all of them are saying, hey, with love, your Plex implementation is a pile of crap. Because it is. Because the way that Plex does things is so rickety and so old. And so today, or maybe it was yesterday, actually, yesterday I got thinking, if Jellyfin has this sort of code thing that I can do, I could probably do something like that with Plex, right?

2:07:19John Siracusa:and now I'm in a little bit of a pickle because what I figured out it's a good pickle but a pickle nevertheless what I figured out and I've mostly implemented like the plain Jane easy approach the no frills version is already implemented it's not on test flight or anything yet but it's close to test flight and I can ask have you try to log into Plex and what it does is it goes and asks the Plex central servers, the central Plex servers, hey, this user would like to log in. And it gives you a very long code. It's like a 20-character long code. But the nice thing is the user never needs to see that.

2:08:01John Siracusa:You just send the user to a particular URL that includes that code in it. This is very much like what John has done with our discount codes for ATP Merch. And so you go to a special URL and the user just needs to enter, you know, their login credentials with plaques, you know, in the spirit of OAuth. I never see any of these credentials or anything like that. And they don't even need to enter a code. They just say, you know, is this okay? Yes, it is. And that's that. The catch though, is that this only works with asterisks and daggers and double daggers. This only really works for playback on your local server.

2:08:39John Siracusa:It doesn't really work with remote servers. And the reason that is, is because Plex, while it has a central authorization and authentication mechanism, it doesn't really have any sort of central, what is this user playing anywhere sort of server or situation. And one of the great things about Plex is that I can share my Plex server with other people without needing to punch more than a single hole in my firewall, without needing to do the tail scale dance, without needing to do anything else, I can share my Plex server with other people so they can stream my media and vice versa. So if John, for example, has shared his Plex server with me, which last I saw he had, and I'm streaming something from John's house to my house, this call sheet integration wouldn't work because I'm interrogating my local server to see, is Casey playing anything locally?

2:09:32John Siracusa:And I'm not going and interrogating John's server because in order to do that, I would have to interrogate all of the servers I have access to. And not to humble brag, but that's kind of a long list for me. So the question that I kind of have, and I'm curious what you two think is, this seems to work a billion times better for pretty much every use case, except when you're watching something from a remote server, which is not an entirely uncommon use case for Plex. Or you can use the old school, like the old school one, it's still functionally there, but it almost never works in my experience.

2:10:11John Siracusa:You can use the old school multicast thing and you can hypothetically get whatever you're watching across any server on the entire planet. So I'm thinking as I sit here now, I'm thinking I'm going to ship this new Plex integration and leave you the option to do either or. Although the old version is so bad, I've been considering pulling it because it's really very rickety. But I'm curious if either of you have thoughts on how I should handle this. Because again, the summary is, I do this new thing where the user does have to log in, but it's very quick and very easy. In fact, it's actually even easier than Jellyfin, which I didn't think was possible.

2:10:44John Siracusa:But you can only see what you're watching locally, like what you're watching from the local server, I should say. Or I can leave it as is and just tell Plex people, well, sorry, Plex has a really rickety system. There's nothing I can do. Or I hope it works for you, even though it never works for anyone. So I don't know. Do you guys have any thoughts on this? Any opinions about this?

2:11:03Casey Liss:You got to think in 2027 mindset, Casey. Here's what you do. WWC Apple announces a new version of Shazam kit that works with TV shows. You just record audio, figure out what they're watching based on the audio that's in the air. And then you look that up using TMDB. Done and done. That would be amazing. No API interface. No figuring out what server you're streaming from. No figuring out what are you watching it on. Is it on Apple TV? what network do i have to scan with this log in with that and figure out no no it's just it's in the air uh and if you could take over the camera and have a look at the screen like you got to do it the the machine learning ai way just sound and video is happening in the room what is it what show is that what episode what actor is that on the screen done and done i mean that that would

2:11:46Marco Arment:be like that would be the coolest and most modern version of that and you can the most privacy invasive yep yeah but and you could even do something do something like you know take like take a picture every three seconds for like 10 seconds with your glasses with your glasses just for like in like one burst like identify this boom boom boom send those three pictures and then just

2:12:07Casey Liss:and send the pictures you take to to some kind of moderation center in another country right

2:12:11Marco Arment:exactly no no i was gonna say like you know send them to like you know the the open ai avis or whatever and say what movie is this like like that could be done but no i think setting aside those kind of solutions for now because that doesn't really sound like you or necessarily a great idea. You have a choice between something that has a more limited feature set but works a lot better and something that is more broad. I would go for the former every time because if people are promised functionality that is slow, rickety, incomplete in practice, that's going to make it a more frequent negative experience for them with your product.

2:13:00Marco Arment:Whereas if you say, this version works a lot better, it just doesn't support this one use case. Right, right. They know that up front, and the more common cases where you're watching things locally work a lot better. That is usually the right approach. Make a product that covers less content area, but covers it better.

2:13:21John Siracusa:Yeah, I think you're right. And that's what I'm leaning towards. And like I said, I'm leaving for now anyway, as I sit here today, I plan to leave the old and rickety version there if you prefer it and basically have a toggle between the two. But I really think that you're exactly right because so many people have written me over the last couple of years and said in very, very nice ways. Did this ever work, dude? Because it surely seems like it doesn't. And again, the people who have said this have pretty much universally been extremely nice about this. And so I feel like giving a better experience to your point, Marco, even with the caveat that it's only for local stuff, I feel like that will generally be better for almost everyone.

2:14:06John Siracusa:And I think that that's probably the right answer. But like I said -

2:14:10Casey Liss:I mean, the reason I half jokingly gave my, you know, Shazam kit that can understand TV shows answer is because I watch most of my TV shows in a thing that you're not supporting at all. And I want the experience of being able to open up Call Sheet and not have to search for the show I'm watching. I just wanted to know I would personally be willing to give Call Sheet access to the mic to, you know, TV show Shazam kit what I'm watching because that would save me having to do the search. i would personally be willing to point the camera at the screen although that would probably be a lot less reliable than the audio but like the problem the reason i'm suggesting this again only half jokingly is because it sidesteps all these other issues because like even if you just want to support apple tv apple is not going to roll out an api where like hey third parties do you want to know what someone's watching on the apple tv on any app just use this api that apple will never do that tv makers do do that tv makers watch what you're watching and report back to their servers but apple will never nor should they apple shouldn't do this but i the bottom line is i'm on my couch i'm watching a thing what is that person from you could save me a bunch of steps if i could launch call sheet and just see the show up at the top and that's not gonna work because i watch my shows on like the hulu app the apple tv app like i don't watch them on plex i don't watch them on jellyfin right i would do the login dance of whatever i had to do to make that work but we watch on all sorts of things like we still have that hallmark plus streaming subscription like there's just too many apps and so that's why i'm saying there's not going to be any app integration can we sidestep it can we get around it can we use the power of machine learning and advanced technology to narrow it down even to say these are my top three guesses for what show you're watching tap on the one you're watching and i know integration will be better and i applaud you for pursuing this but like you're just really nicking away at like the corner of a giant iceberg because it's great for the five people who watch all their stuff through jellyfin or watch all their stuff through Plex.

2:16:01Casey Liss:And maybe those nerds are kind of like a core audience, but your app has much broader appeal. And so I do think most people who use your app will never know or see anything about this feature. So in that end, I would say, maybe try to time box the time you spend on this, speaking of being nerd sniped, because as cool as it is, you're never going to get anything close to enough coverage that the majority of users of your app are ever going to see this feature.

2:16:27John Siracusa:Yeah, I know you're right. And on the long-term list, which it's been on the long-term list since before LLMs were really a thing or before I was aware and exposed to them anyway. If you look at Home Assistant, and, you know, it wouldn't be me if I didn't bring that up at some point. If you look at Home Assistant, there is a mechanism by which it will show what Apple TVs in your house are currently playing. Now, it doesn't really show a whole lot about it. it's like a title in very little else. Whereas the nice thing about using Jellyfin or Plex or channels is that they'll often give me an exact movie database ID number for the thing that the person is watching.

2:17:04John Siracusa:But I'd like to, at some point, set an LLM at the Home Assistant code base and say, hey, there is a mechanism by which, I think there might even be a Python library to do it with that Home Assistant leverages. I have a bookmark somewhere that I could dig up. But suffice to say, look through this code and write me a swift version of it so i can do this too and what it ends up doing is the the four digit like pairing code that you would do i forget when you do this sometimes with airplay sometimes with other things you basically have to do that dance once or maybe twice and then at least home assistant will show you the title and like a image for what the person is watching not like literally a screen cap but like a poster image or something like that and so I'd be curious to see if I could make heads or tails of that with the help of an LLM on my side I did try to dig through this code two three years ago whatever it was and it was nigh impossible for me to figure out I think in no small part because I am I am able to speak Python but I am far far from fluent in it and I think that was a large part of the problem but over time, maybe I will, especially if I start digging into the more agent-y stuff where I say, hey, go figure this out.

2:18:19John Siracusa:Let me know when you're done. Maybe I should do that at some point and set Claude off and say, hey, build me a Swift library that does this thing. Here's an example of it that works over here. See what you can do. And maybe that'll work and maybe not. But the downside of that, though, is that even if I can figure out what you're watching on Paramount Plus or what have you the best i could do is when you tap on it it pre-fills out a search for star trek strange new worlds or whatever it's called you know what i mean it's just it types those words out for you rather than jumping you directly to the the entry in call sheet for star trek star trek strange new

2:18:53Casey Liss:worlds you do the i'm feeling lucky and show the first search result in line that's true

2:18:58John Siracusa:is this not it type of thing but you get my point but i don't know we'll see what happens but i i I have very much enjoyed trying to figure out how to do this. And I've been working with Claude to do it. These are things that I think I am capable of doing without the help of an LLM. But again, it's much quicker to ask it questions and have it give me direct answers than having to go spelunking through Plex documentation or third-party copies of Plex documentation or what have you in order to figure it out. But I mean, pretty much all the code I've written for this particular feature is, I think, 100 % me.

2:19:29John Siracusa:I don't think it's 99 % me. I think it's 100 % me. it's just the figuring out what the magical incantation I need to do against the Plex API that was largely a conversation almost entirely a conversation with Claude and so again that's the sort of thing that I think is really powerful let me have Claude help me with the thing that I'm not uniquely skilled at and then I will do the things that I think I am uniquely skilled at and that's been that's been fun and rewarding as well