Hi, I'm Leah Wommelsdorf, and I'm a part of the worldwide developer relations team at Apple. Today, I am joined by members of the watchOS engineering team, and we are so excited to answer some of your questions about the platform. So before we get started, I'd love for you all to introduce yourselves. Dan, I'm giving your name away, but Dan, do you want to introduce yourself, talk about what you work on for watchOS? I was going to call myself David, but now you stole it. You can't go undercover anymore. It's all over. Hi, I'm Dan Keen. I manage the workout team on watchOS and I'm super thrilled to be here. I'm Anne Hitchcock and I am a watchOS frameworks engineer and I work on all kinds of stuff. What are we talking? Throw some frameworks out there. Throw some frameworks out there. So system spaces stuff. So system UI. I've worked on some connectivity stuff this year. I work with a lot of other teams. So I've worked with the Foundation Models team this year. I worked with the App Intense team this year in their work with watchOS. So a lot of all over the place. That's amazing. Cool, I guess I'm next. Yeah, I'm Devin. I work on widgets and live activities and then how they integrate with system spaces like SmartStack and WatchFaces. Yeah, and hey, I'm Matthew Combs. I manage the watchOS system spaces team. So that includes a lot of what you kind of think of when you think of watch. So how we launch apps and how we get notifications and all the underlining look and feel of the platform. So what I'm really excited about is we have a lot of frameworks and a lot of system experiences covered here today from HealthKit to UI frameworks to connectivity. And what that means is we're really well posed to answer a lot of your questions about all things watchOS. And this year has a lot of big updates for watchOS. There's new frameworks coming to the Watch with foundation models, vision framework, core AI framework. They're all on watchOS 27. And there are also some big updates for liquid glass and new health kit APIs like the menopause API and the workout zones. So there's a lot of new stuff that we're well-suited to talk about. But also what's cool about watchOS is there's such a long history of the platform. And this is a great avenue to discuss all of your questions, whether it's about the old stuff you've been thinking about for a long time. qualities that make great watch apps, or the new stuff that you're starting to think about. So we're so excited for your questions today, and I'm really excited to dive into it. So I'm going to check to see some of your questions. Okay, so for our first question, and this is from, forgive me if I'm pronouncing the username improperly. This is from mcrow. Okay, the question is, a session video mentions foundation models are now available on watchOS. Does that run directly on the connected iPhone or directly via PCC, private cloud compute, in the event of an independent watch app? So, Anne, I'm going to kick this off for you. Let's talk about foundation models on watch. Okay. The foundation models on watch do not require you to be connected to your iPhone. But they will require that network connection. You can use PCC or you can use anything that conforms to the language model. So the language model protocol will have official support for Claude and Gemini coming soon from those vendors. But you can conform to language model yourself. But it will require a network call. So you're not going to have models running on the device. But I've had a lot of fun this year working with that team and playing with that framework. And there's so many fun things that you can do with that that are great use cases on watch, like summarizing text to fit on the watch display and, you know, making more accessible text out of really dense text that you might have. So it's a lot of fun. Yeah, this is by far the thing I'm most excited for because I think it's going to be where we'll see the developer audience take it so much further than even we thought it would be possible. Like, this is such an amazing and emerging technology, and we're finally opening it up on a platform that has a lot of great health data and insights. And now you can go see how it applies and you can use it to build some really amazing experiences. So, yeah, by far Foundation Models is my most excited thing. But I think it's things like that, the tech summarization or take all this data and go feed it to an LLM and see what kind of insights you can provide to your users. It's going to be such an exciting opportunity. So I think there's a lot of really cool stuff people can do in terms of learning resources. I know there's some dub dub videos, some sample code. I'd love to hear that part purely, like, hear more about what someone needs to do to onboard to thinking about foundation models framework. And then also the other piece of that story, network connectivity. You know, you're going to have developers are going to have to think about what type of experience makes sense in my app. Some graceful fallbacks if there's, you know, maybe it's not a cellular supported watch. What types of things should people, developers, be thinking about as they prepare to use foundation models framework in their watch apps? Yeah, they do need to keep that in mind. And check there are some APIs in Foundation Model to check, is Foundation Models, you know, is this available right now? So make sure that you're checking that before you're making these calls. And definitely if you're using PCC, be aware that needs an entitlement. That is mentioned in that reference session. And so plan for that fallback that if you're using PCC or any language model provider, have you reached the quota? What's your token usage on that? So there's a lot of fallbacks that people need to think about when they're designing their apps using foundation models and large language models. Yeah, and it sounds like there's a lot of watch, like whether you're using a foundation models enabled app or you're just making a regular watch app. I feel like there's a lot of design foundations that also come into play when making a great experience on watch. Like some qualities that come to mind are glanceability, predictability. Like what other things come to mind that we should really be keeping in mind as we think about the next wave of watch apps? I think it's to be a lot about trying to find the right experience for the wrist. This is a different form factor. It's something that we talk about year over year with Apple Watch and WatchOS. And I think, again, the excitement I have about this new technology for us is that we will see what you do with it to go build out these great new experiences. But it really comes back to the fundamentals. And the fundamental experiences on Watch are that they are literally smaller. They are on your wrist. Yeah. They need to be glanceable. They need to be quick. They need to be concise and easy to digest for a user because you are going to look at your wrist and then put it down and go do something else. And so it's about bringing that information to your users in a timely and immediate interaction. And I think they're not just smaller and glanceable, but they're more intimate. They're more personal. And I think a large thing that the foundation models can give you to the insights and stuff that we've been talking about is how to make that more personal, how to make it feel that the content that you're going to display really quickly for the user here is really personal for what they want it to be. Yeah, it's something that we think about a lot with how we've been using new large language-runal type features for interactions on our platform and where you guys can take it as well. So, yeah, I mean, go back to those fundamentals and then see how you layer these experiences on. Because at the end of the day, it's still a watch. It still has all the power and performance constraints of this platform. But it's now opened up in a whole new way than we've had in the past. Yeah, I'm really excited not just to think of new ideas myself and to see what you, the developer community, comes up with. And as you're trying new things and you have questions or you have discussions you want to have, I really encourage you to one-file feedback through our feedback channels. And also check out the Apple developer forums to ask questions as you hit roadblocks or maybe want to deepen your understanding. But I think this is a really exciting year for watchOS. Okay, I'm going to dive into our next question. And, okay, we have, oh, you guys are really interested in the foundation models. Okay, I'm going to do one more on the foundation models before switching gears a little bit. So this is from Dirkio. And this question is, when using the Foundation Models Framework on Apple Watch, does it only use PCC through either cellular or Bluetooth to iPhone, or can it also utilize the on-device model in the iPhone if it's on the same network? So, it does not use the on-device model on the iPhone. It uses PCC over the network. So you're going to make a network call over your cellular or Wi-Fi connection. And so you can use PCC or you can use another language model conforming protocol, but you're going to go over the network. And so that's something to keep in mind. You're always going to have that network latency for those calls. So really think about what you're making those calls for and the timing and when you're going to do them and how many you're going to do them. because you're always going to make that network hop. That is good insight. And again, on the topic of foundation models, I'm really excited that there's such great documentation about all the opportunities for every Apple platform. So I really encourage you all to check that out as you continue exploring. Okay, we have more of a platform question from username LazyVar. Fun name, very cool. Okay, the question is, As an iOS dev moving to watchOS, yay, I'm adapting to the watch's unique background constraints. What core architectural paradigms or hidden pitfalls should we keep in mind when handling heavy off-main thread work like CloudKit syncing without relying on iOS habits? There's a lot to get in here. Thank you, LazyVar. Okay. Who is up? Matthew, do you want to kick things off? Yeah. I mean, I think I was mentioning the constraints of the platform. And for a lot of what we want you as iOS developers coming to watch is, thankfully, a lot of the APIs that you are already using work on the platform. You have NSURL session. You have things like CloudKit and off main thread, like Swift Async. There's a lot that you already have in the skills that you can bring to the platform as well as just a whole wealth of SwiftUI. And code that you have will be able to be reused. The key thing to know is that we, as the platform, do take strident rules. So I'm trying to make sure that you and everyone, because there's so much that's happening on the watch at any one time, that we are giving the runtime and preserving the battery life for our users. Because we know how much all-day battery life matters to our users and to your users. So those are the key things to know about. We do have limits put in place by the system. You'll see sometimes things like watchdog timeouts that are there put in place to make sure that both you and everyone else that's running on the platform are good citizens. But for the most part, you should be able to just reason through similar kinds of background runtime considerations you have on phone. But just know that the timeouts are very strict and they're strict for good reasons. You brought up Swift Async and I do want to say one of the things that I really enjoy you doing on the watch is how constrained the platform is Yeah, so if you're coming from iOS to watch OS, it's not just the smaller screen You also have a lot lower core count and stuff like that. So being cognizant of the work that you're kicking off is very helpful Yeah And you know that because sometimes like constraints can sound scary. Oh, no constraint But it's an intellectual opportunity and really thinking about I think like something we talk a lot about the watch platform is that some things are great on an iOS app or macOS app and on a watch app it's a totally different design problem and there as you alluded to earlier there's this great guidance from the watch os 10 WWDC video it might be called design I actually am not remembering the exact name but it's a really great video about designing great watch experiences instead of just taking everything from an iOS app and putting it on the watch think about those core moments and experiences that really shine on the platform with the intimate qualities of the watch and how people leverage them. Yeah. Design and build apps with watchOS. There we go. And you'll recognize one of the speakers. I was going to say, we've got a lot of famous people today talking about their videos, a little self-promotion. Yeah. But no, it is really cool. And I loved what you said about different frameworks, like many frameworks being available on watch with similar like instead of just being available it's not just oh it's available it's tailored to the platform like devon i think you could really speak about widgets and live activities like there's so many system experiences that are tailored to the watch and what comes to mind for you is the opportunity for developers yeah definitely i mean i think um coming from ios actually there's a lot of similarities with widgets and live activities the runtime you get for widgets on watch os i mean other than a slower uh smaller core count it's actually the same as what you get on iOS. So that typically transfers over pretty cleanly. And even if you already have a lock screen widget, that should just work out of the box on watchOS, maybe with some small layout tweaks. I think we do have some unique API services, though, especially around SmartStack. You have opportunity to get your widget suggested at the right moment for the user using the widget relevance API, which is heavily exercised on watchOS. So I think there's a special opportunity there where you can really think about just the right way to be proactive with your app and push information to the user instead of needing to dig in and look for it. We have an interface where you can just bring it when the time is right. And Dan, what comes to mind for you thinking about health experiences on watch? I know there's a different level of data availability, a different time span. Does anything come to mind as a unique opportunity on watch that's different from on a phone? Many things. Oh, good. I set you up well. We did. If you think about it, so the watch has a lot more information about you that it's recording at that time. So your calorie information, your heart rate, stuff like that, it's coming from the watch. It's sort of that as the primary case for it. I was talking to some developers on Monday just about interesting ways to use heart rate in their apps in terms of surfacing, wow, your heart rate was really high during this time. That's kind of interesting. totally not in a fitness related context. So I think that there's an opportunity to, again, going back to that personal intimate thing, we're measuring things about the user. We're recording them into HealthKit. It's all there. There's times when that could actually really be applied to things that you don't necessarily even expect it to be. Yeah, I totally agree. And again, an opportunity and design constraint problem. That's really cool. And I have one more, I have one more treat for LazyVar and follow-up question. For those, I think what's so cool about Apple Watch and WatchOS is that you can actually have a presence on Apple Watch without building a watch app right away. Through system experiences, like live activities, through notifications. So for people who already have apps on other platforms and they're WatchOS curious, whether they're thinking about building their first app or they're like, how do I have a presence on Watch? What types of things do you recommend the developers think about? How should they approach that transition or what's step one? I mean, I think that there's a lot of really good opportunities. You mentioned one of the biggest ones, which is live activities. We're like, we really love how we have this experience with the phone. Live activities show up on your lock screen. They're like these great glanceable interactions. Mind you of another platform? So we've decided to bring them to the watch as well. And they come automatically. But there's an opportunity for developers that if you want to customize that experience and make a really great looking live activity that shows up on the smart stack, it's really easy to do. You don't need a watch app at all. So we keep on looking for more and more opportunities like that. One that we also brought in WatchOS 26 is controls. So you can go into the WatchOS Control Center and add your controls from the phone app to the watch. So I have one to unlock my car. The developer did not have to do anything, actually, for that one at all. There was no customization needed. We look at what kind of controls that you provide, find ones that make sense to bring to the watch and do it automatically. So it's a really great thing to think about spaces on the phone already, like the lock screen or control center, and we decide to bring those over to the watch so that your users are likely using them already. And I think that's just a really good entry point into the platform. Of course we want you to also bring an amazing app, but apps don't necessarily make sense for every use case. And so for those experiences, the live activities, for example, are just like a really great way to still bring your experience front and center to people's wrist, have that personal interaction. And it's really not a whole lot of work. Yeah, yeah. Like it's validation. Like you need to test your code, try it out, and like be aware of that. But I mean it's a really lightweight way to get on the platform. Exactly right. Yeah. Very cool. Okay. I've exhausted that. I'm going to keep going and check out some more. No, lazy. That was a great question. And I really enjoyed it. Oh, we could plug the live activities fundamentals session. Oh, that was a great session. So if you want to dive into live activities, that is a great session. There you go. On the topic of fundamentals, another set. So live activities, fundamentals and widget kit foundations. I know two different naming conventions, but there's two great videos that for people curious about watchOS as a platform, I really recommend that you check out and you can build a strong foundation. with these system experiences. Yeah, so we did have one more thing for LazyVar. One more thing. That's a good thing. Oh, man, there's even more. That was complex. Okay, I have actually another really good question from LazyVar, and then I promise. Okay, this one is about a different topic, though. Any tips to speed up physical watchOS debugging loops? What are the best practices for minimizing Xcode installation delays and handling the initial system symbol sync system symbol sync. That's like, I'm impressive. I delivered that well. More efficiently. Can we force a local wifi tunnel or pre-cache these system symbols on the Mac? Let's talk about, yeah, let's talk about tooling. Ann, do you want to kick it off? Oh, oh, goodness. This is something that Ann and I spent a lot of time thinking and talking about. So like, this is a conversation that we have frequently. Yes. Yes. Goodness. Yeah. I would say, yeah, ensure your connectivity to the Mac is good. Be nearby. You know, I don't think there's a lot of magic to this. You keep your OS up to date. I know that the Xcode team has done a lot of work over the past couple releases to make this quite a bit better. So please give it a try. Update, get that, you know, get the beta installed. Give it a try. No, this year I've heard a lot of, there's been a lot of attention. Yeah, they have put a lot of work into this and not only have they put a lot of work into this, but they put in a lot of tools to help diagnose any problems that you might have. So running into a problem, even after all the work they've done, please file feedback because they've put in a lot of stuff to help them diagnose any problems that you're seeing now. So please file that feedback because that will help find any of these lingering issues that you're seeing now. So yeah, install Xcode 27, give it a go now because they have heard you. They have heard your frustration and they have tried to make it better and they want to hear, is it working? If it's not working, file that feedback and send them the logs because they're ready. Yes. I'd say just like the top tips and tricks that we think about, we now, starting with Xcode 26, but also in Xcode 27 with Device Hub, we have a direct connection from Xcode on your Mac directly to the watch. In the past, we would proxy through the phone, and we have seen substantial improvements in both reliability and throughput. by having the connection go directly from the watch to the phone. Sorry, watch to your Mac. So that, you mentioned, like, infra Wi-Fi. That's actually something that we are doing. So we do have that connection. Do make sure that you are on a network that actually has the ability for devices to talk to each other. We've seen some issues with people, especially like on corporate networks, where a peer-to-peer is not enabled. So do you keep an eye on that kind of, like, network setup? More recent Apple watch hardware also has a 5 gigahertz Wi-Fi chip That helps also again for throughput So there's just a couple of these like things that we can do to help the actual like network latency That that's why you're saying update And and using more recent hardware helps a ton there. We of course though Support and need you to be testing on the whole range of Apple watches that are out there right now because your users are likely not on the latest Apple Watch. And so if you need that connection, like testing on physical hardware is great. Using DeviceHub to test all the various different hardware configurations, though, is also a great way to make sure you get that validation across all of your potential users. Absolutely well said. Were you going to add something, Devin? No. Amazing, I will. There's a lot to, we're really eager to hear your feedback on Xcode 27. So as you are debugging, let us know what you think. Okay, I'm going to move on to a new question. Let's see. Okay, this is one I'm excited about because I have a really good answer. But PearlJam66 says, what are your favorite new features of watchOS 27? You can go first. I know. It was like, who's going to speak first? Who's going to speak first? Well, I mean, I've got to plug my team, right? Yeah, let's hear it. We've got really great new insights in Workout Buddy that I have personally found wonderful in my workouts. And I'm really enjoying them. And that's not to say less of any of the rest of the amazing features in WatchOS 27, but I really enjoy it and think they've done a great job with it. Very cool. Could you give us an example? Like what kinds of insights? Sure. Measuring speed over time and how it compares to previous weeks, months, year, how that's going. And as my cycling. As we're full and personalized. Exactly. And as my cycling has been getting better, it's actually been uplifting to hear. That's awesome. That's exciting. Oh, can I have more than one favorite feature? Yes, you can double down, triple down. I didn't know that was an option. Oh, so I've built a lot of apps and I play with a lot of APIs. I mean, they're not really playing. But anyway, so I have had so much fun with Foundation Models framework on watchOS. So that is a favorite for me. But also being able to update widgets with watch connectivity. So, yes, that is great. So if you've been waiting for that, it works now. I feel like say it again into the mic. Update widgets with watch connectivity. Yay. Capital letters. Yeah, it's great to see that one get across the line definitely. I think it opens up a lot more opportunity for apps to stay in sync between the iOS app and their Watch companion and just push it right to your wrist. Exactly. Well, that's actually one of my favorite features too. So you saw that one. I guess also plugging my own team. I think there's a lot of enhancements to general performance and reliability around widgets this year and making sure they load in a timely manner, making sure you have more of that runtime to update information. Also, we have new suggestions in the SmartStack. I found a really handy one the other day. It gave me a reminder of a friend's birthday. But yeah, always making new suggestions. Yeah, that's one of my favorites too, actually. Very windy. It's also just really fun. And I think people that update to WatchOS 27 will find some delightful new surprises in the SmartStack on holidays and other significant events. So look out for those. Now I have something to look forward to for holiday. Check out the SmartStack. I think my favorite feature is what we're doing with some really great work with Siri AI. It's something that's not out in beta one, but something that will be coming. And we're really, really excited to see what we do in that space. I think it's going to be really, really awesome. And just in general, I think the new intelligence features on the platform are fantastic. I am so excited. Very cool. I was going to say, I'm really excited about all the health updates, the health kit updates coming to watchOS. And not just the experiences within Apple's apps, but to see how developers use them with heart rate zones, cycling power zones, and the perimenopause and menopause API. I'm so excited for the full breadth of features. I was going to say, like, if I got two answers. Okay, just this once. Number two was the zones, but now you've stolen it. I know, but Devin stole Anne's answer, so really we're all just stealing. It was a co-answer. It was a co-answer. Mutual appreciation for multiple frameworks. So that was really good. Thank you, Pearl Jam 66. Okay. I have a question about SceneKit from Zev Eisenberg NYT. Exciting. Maybe that stands for New York Times. Okay, the question is, I have a watchOS app that uses SceneKit. SceneKit was deprecated in watchOS 26, and its replacement, RealityKit, is not available on watchOS. is there a recommendation, is there a recommended modern way to render 3D graphics on watchOS? So Matthew, I'm curious about what comes to mind for you for this. Yeah, I mean, I think SceneKit deprecated but not gone, so you're still welcome to use it in your app. But we really do think that the path forward for watchOS in particular is what we provide in SwiftUI. And so SwiftUI Canvas has been around for a couple of years now. It's a really great drawing system. It's not SceneKit, but it is still a really well-featured and, I think, appropriate for the watch platform graphics library. So the note there is likely time to start walking away from SceneKit for watch and going towards SwiftUI and SwiftUI Canvas. I will say, Canvas is very fast, too. It runs on the GPU. Take advantage of that on the watch. Yeah. And if there are things that you can't do, can't figure out how to do, if you can't figure it out, talk to DTS, file feedback. Post in the forums. Some other people might know. Yeah, because please figure out. If you haven't figured it out or if it can't be done, we would like to know. Absolutely. Name of the game. We're all trying to make the watchOS platform as optimal as possible for you, the developers, and for all the people who love using their Apple Watches. So, yeah, if you can't figure it out, we want to help you get there. Very cool. Okay, I have another question. And this is also kind of in the theme of thinking about having a great, strong presence on Apple Watch. And this is from Prechaya Tries. Tries. I think it's Tries. Like, yay. Okay, cool. The question is, what do you think are the things that most app developers might not know that they can do on watchOS that will make them want to extend their app to the device? For example, a long time ago, they were interested in making a VisionOS tennis game using the sensor on Apple Watch for the racket, but it didn't seem possible at the time. So, yeah, what do you think people like – what do you think are some of the really awesome opportunities on Apple Watch? It's interesting because I wonder what was missing for them then. I think that there's actually a possibility for doing that as a peripheral for Vision OS. Going back to the earlier thing about the watch being on your wrist and having all that health data and everything, there's also CM sensor data and stuff like that. CM being like CoreMotion. Yeah, sorry, CoreMotion sensor data. For the other folks in the room. CoreMotion, little street lingo. Yeah, so you could track the motion of the device while you're doing things, which could be cool for anything from a tennis thing to, I don't know, maybe you're into pottery and you wanted to see how steady your hand is as you're forming the vase. I could see that being interesting. So I think there's a lot of possibilities of novel experiences. I think some of the greatest WatchOS experiences are the ones that are kind of like proactively brought to you. And so I'll keep on talking about live activities because I think that's a great one. But there's a lot of these really awesome kind of outside of your app experiences that we try to bring proactively to users when the time is right. And so that's really the promise of the smart stack. And relevant widgets are a really great way to do that as well. So for people, I realize we're also throwing out some vernacular that might be unfamiliar to people. Like what makes a great live activity or what makes something meaningful in the smart stack and most relevant? Like what are some of the qualities people should think about when exploring these experiences? I'll start with widget suggestions. Yeah. So I think with widget suggestions, we really just care about what is the user or what is the customer doing currently? Like what is their real world context around them? And when you register your widget to become relevant, you associate it with like a location or date or sometimes like semantic locations, like when the person's at work or at home. So just making sure that, you know, what you're showing is really relevant for their real world situation around them. I mean, your watch is something on your wrist. It's in the moment with you. And the things we're showing you should, you know, help give you information about that or, you know, help you unlock your front door when you get home. Like actions that you can take around you and being accelerant to that. Yeah, I think another advice there is try to avoid being in that space too long, actually. You might find that users begin dismissing your app if you're taking up that top spot too often. So, yeah, just being judicious about when is the right time to surface this thing, what might the user want to do with pretty good confidence. When I think of live activities on watch, I think of the Mango Cart app that you made in a video. Was that WWDC 25? Yes. Yes, it was 25. All here. Bring your live activity to Apple Watch. Yes. And so like that live activity was a mango delivery? It was produce shopping. Produce shopping. Yes. Produce delivery. And so that was, yes, you could go shop for your produce box and get it delivered to your house. And it would show the status of your delivery. So you've ordered and how long is it going to take to get to you the progress of your delivery? And meaningfully, it's not just there all day, like five hours after it's delivered. It's like, okay, it's been delivered. Now it's gone, right? Yes. It's like kind of the kind of thing we should be thinking about a lot. Right. I think with live activities too, you have the opportunity to have like notifying updates and non-notifying updates. So I think there can be a difference between things you really want to alert about. Like your delivery has arrived right now. I need to go outside and get it. Definitely alert for that. Small status updates, you know, can be something you passively track. So just be mindful of, like, what's the important piece of information here? What's the signal worth alerting for? That's something that we spent actually quite a lot of time thinking about and worked with the sports team to develop. They were working on bringing live activities to the sports app, was, you know, we want to have alerting live activities. They don't alert for every score, though, because you look at every sport, and they don't necessarily mean the same thing. And so basketball, they alert other quarters. soccer, they alert every score. And so these kinds of thinking about when and how is the right time to alert for live activity is super important to the watchers experience in particular because we actually bring the smart stack back to the forefront and give you an opportunity to draw something really cool or show a great update. And then the user can dismiss it. You can get more updates in the background as the game's progressing. And then a big score, you do another alerting activity-- And and kind of go into instruction a little bit more I think we've also thought a lot about notifications versus suggestions in the smart stack and sometimes it makes sense to present information to a user But not necessarily alert and we've done a noise suggestion this year So I think a little bit before you normally get the notification about high noise. You'll get a suggestion that smart stack So it's more like a soft reminder. We're not quite ready to like alert you about it yet, but You know, we know something's going on here. It's worth showing worth elevating. Yeah, this is all so The best place to learn how to make these decisions, like if I'm making a sports app and saying, oh, this is how often we should notify people with the live activity, like what is the best resource to learn about how to make these decisions? I'd say probably the developer website more than anything. I think, you know, the tying these together end to end is something that we kind of leave to the developer audience to figure out how this should all fit together. but every one of these tools serves a unique purpose in and of itself. And we found great ways to pull them together to build some of our experiences, what we provide from the apps already on your watch, or when we work with Apple developers that build the sports app, and when and how do they provide notifications versus live activities versus a smart stack widget. It's something to look at and get inspired by. And hopefully that you too can take that and run with it and find really awesome ways to use that data yourself. That was like great API. So we don't want to be too prescriptive, I think is the answer. Like they have all these tools and they're for you to figure out how to pull them together. Yeah, I think it's very demand specific. Yeah, I think, yeah, it is. Ultimately, you as a developer know your app and your domain and you will know. You figure out what's right for you. Yeah. Absolutely. And I will say, I think times of year like WWDC, when we get the chance to speak to the developer community and hear about the problems that they're facing or the things they're trying to solve or the questions they have, I think that also informs the direction of our frameworks in which wrinkles we iron out. It's a feedback loop that we have. And I think it's a challenge that we have as Apple platform developers is to find the right entry points and the tweaks that we want you as developers to make for your specific use cases. That's what makes a really great API. It's flexible enough that you can go do some really amazing things. And we try to build it with that in mind. I mean, it's a very noble goal. And I think I'm happy to be a part of it. It makes me smile. Okay. I have another question. There's so many good ones. Okay. You all have excellent questions. I love that one about new liquid glass. That's something that I was really excited about. You know, yeah, let's talk liquid glass. We haven't given it any love because, yeah, it's time. Okay. this is a question from simon from helix what's new for liquid glass specifically on watch os let's talk glass yeah let's talk glass um i'll dive into it yeah like so uh we were really excited with watch os 26 to bring liquid glass and the new design language to apple watch and there has been a lot of refinement we've done uh even since that initial uh release uh so something that we've brought in watchOS 27 is the updated look with more of the dark edges, dark speculars, what we call them. Some nicer separation of content as stuff scrolls so that you are able to more easily preserve legibility in your app, as well as just a lot of really great performance improvements, just something that we've noticed in our own use. And so we worked really, really hard on making the liquid glass experience really shine on the platform. I think my favorite thing is just using the really great reactive glass experiences where you drag the buttons and stuff. With the interactive glass. Yeah, with the interactive glass. That is really fun. It's just fun to play with. And we're excited to have that in Watch and throughout the whole system. So it is something that comes to a lot of apps automatically on WatchOS especially. We found that our design system that we brought in WatchOS 10 really naturally leaned into where we were going with a new design language. And so for a lot of your apps, the Liquid Glass updates even last year and this year are already there for you. And so, you know, we're excited to continue to work with you and the developers and provide feedback again. Like I think if anything that this year and the kind of how we were talking about all the operating systems with these really great performance and improvements that we've made, it's all about that feedback loop from developers and users coming back to us. And I think Liquid Glass and the updates there are really great refinement on that system. Yeah, I think like the feedback channel for Liquid Glass in all the platform 26s, The developer relations team met with developers around, really around the world, getting the feedback, seeing which areas were either confusing or challenging or weren't meeting the needs of all of the users. And that really informed the direction for iOS and watchOS 27. Yeah, I think the, like, what's new in SwiftUI session this year, too, gives you some really great new tools to allow them come to watch as well for you to customize and make sure, like, what glass is working for you and your app. And I have found it delightful when I see the apps I know and love update with Liquid Glass. And it feels really cohesive to the system. And it makes me, especially as I go between different platforms, I can see that consistency. It's a more delightful experience. I have two follow-up questions. One, is there anything different that developers need to do for watchOS 27 to use Liquid Glass effectively in their apps? For the most part, no, which is actually, I think, a great answer. Yeah, I mean, that sounds good to me. Yeah. For a lot of our developers out there, especially ones that have been developing with the watchOS 10 design language that we brought a couple of years ago, the new Glass designs and the updates we've made come to you automatically. That is really cool. And then my second question, the settings, the personal settings that someone makes on iOS for how clear or tinted the glass is, does that have an equal presence on watch? Is there also a watch-level preference, or is it synced across devices? It's actually, we don't provide one for watch. And if you really feel like we need one, please give us that feedback. I think it was a decision that we made for the platform that we felt that because our user interfaces are primarily dark that the layout and the smaller screen size meant that we felt like we could really hit that balance of Preserving legibility by and still give the the feel of liquid glass Without needing a slider to let users pick where they felt like this is the right balance for them But you know, we're here if you have more suggestions Of course, we also have all the accessibility features, reduced transparency, increased contrast. They are for you on Watch as well if you need those. But we think that the platform is, like, as designed is in a good place. No, that's great to know. Like frameworks, Liquid Glass is tailored to WatchOS. So I think that's a really nice sentiment. Thank you. I'm seeing a lot of love for watchOS 27. I'll just say a quick shout out. Ashafri said, we love watchOS. Thank you all for your hard work. I'll say that to you. Thank you. And asked, what is something new in watchOS 27 that you think is a hidden gem? And I think we've covered that with the new features. But I just wanted to thank you for the love of the platform. I like that one. Oh, you can drive number three. I can just expand on it a little bit. Well, I mean, we talked about this already with the zones API. But I think unlike foundation models, which is like this huge, beautiful feature that everyone is really excited about, zones is something that has actually been requested for years. And it took us a while to get to the point where we shipped an API that we think is actually really flexible and really easy to use and provides a lot of value that I think we learned by using it internally for a little bit first. So I think in terms of a hidden gem, I think that one fits. hidden gem. Yeah, I mean, share the hidden gem. Good question. I'm sorry I was going to move past this. Don't move past this. We have hidden gems. My favorite hidden gem is the rewardable API. Yes. What? Say more. So new in SwiftUI this year is a new API for rewardable, and it was mentioned in the State of the Union, but it is also available on Watch, and so we allow developers to really easily like have collections that they drag around on the platform we actually had used it ourself to implement control center so we are really excited that this is something that we can see in your apps as well now so yeah rewardable that's my first time developers can reorder containers on watch OS ever we are so excited Wow okay I'm really glad we dove deeper here into this Devin do have another hidden gem? I don't think so, no. I want to talk more. Only public gems. Only public gems. I want to shout out more to heart rate zones because Dan, you've opened the can of worms and I'm really excited about heart rate zones. And power zones, power zones. Don't forget the cyclists. And cycling. I know. Cyclists in the room are saying, come on. There were like intricacies around time end zone, which is actually something that's really interesting. So a lot of people were just asking, just give me the boundaries. And we've been able to provide something that's actually much richer for them and still incredibly flexible where you can choose different zones for your specific workout app as well, which is, I think, really cool. It is super cool. And what I also like is you don't have to be just a workout app, but just a workout app to leverage the new workout zones API, because there's the aspect of what zone is someone in during a workout, how hard are they working, but there's also the after workout experience of talking about, you know, how long was someone in a high heart rate zone during the week's workouts? There's many opportunities to think and especially leveraging more intelligent experiences there's a lot of data that you could do cool stuff with so i'm really excited to see how people use it great question okay um we have a question about widgets that i'm excited for from medium fidelity it's very cool i'm thinking of the different family family configurations for widgets okay the supported families. Okay. Does watchOS 27 contain any changes that alter guidance for keeping a widget up to date? Are we looking at about 50 widget, parentheses, complications, updates per day? That's regardless of if they're the background app refresh or APNS based, right? And I think there's a few things to talk about here, the notion of widget budgets, but I think maybe the secret question buried underneath is how do I keep my widget up to date? What things should I keep in mind is a widget the right thing for me should i use something else so devon i'll kick things off with you to get into medium fidelity's question yeah um so i mean off the bat i don't think we have much new guidance for watch us 27 um but i can you know reinforce just good practices um so yeah typically i mean i think you can expect your widget to update every 15 to 20 minutes if it's being actively used um and i think there's a big difference there between whether your widget resides on the watch face or the smart stack things on the watch face are getting viewed all day so they're at the top tier of the budget, and they will be viewed favorably by the scheduler and be more guaranteed to get updates. I think updates can come in a few different ways, too. You can invalidate your widget from your app, and you have some novel information, and you need to tell the system that it's time to get reloaded. But there's also your timeline, and it has a reload policy attached to it that gives it an expiration. And that's also a very powerful and slightly more deterministic way to make sure you get refreshed. We might hold a stale timeline for a little bit, but there's stronger guarantees that will give you a refresh after that. For the SmartStack, I think that if the user's added the widget to the SmartStack, again, kind of in that similar bucket of getting background refresh through the day. But that background refresh is not going to be quiet in that 15 and 20 minute bucket, because it depends on how often the user actually goes to look at your widget. This is similar to on the iOS. If you have a widget on a home screen page a few pages away, you'll get about the same behavior. If the user looks at it once a day, you might get a refresh once a day. Yeah. - I think you dropped a lot of important information. - There's a lot in there. - There's a lot there. - There's a lot of nuance. - There's a lot of nuance. The things that stand out to me are that the system, the system, not just on watchOS, on iOS, on all platforms that have widgets supported, the widget budget of updating is optimized for both the platform and how often people are engaging with your widget. There's a lot of really great documentation about this that Medium Fidelity, the content of your question actually includes. It's an article on the developer website called Keeping Your Widget Up to Date. So I think that is a really good resource to continue referencing. There's also, we talked about it briefly, there's a new Widget Kit Foundations video from WWDC26. And that's going to cover all the different techniques to keep a widget up to date. And the foundations, if you're like, what are they talking about? This is a really good resource to just learn the basics to not only build your first widget, but think about how to keep it up to date and the different tools at your disposal. And similar to everything we've talked about, we love to hear what's working and what is an area of friction when using things like that. I think something that we've seen, especially when we've worked with developers on adopting widgets, is sometimes you naturally reach for the first thing. It's not necessarily the right tool for the job. And so we provide a lot of different types of APIs for widgets and live activities and controls available through the WidgetKit framework that we hope kind of hits at the various different needs of developers out there. Of course, if there's something missing, provide the feedback. We want to hear it, especially with your use case. I think that's something that helps us a lot to understand where and how and why you want to have this thing that does not exist. But I think there's a couple different types of widgets out there. Our typical timeline-based widgets are really great for data that updates throughout the day. I think a canonical example is weather. It changes, obviously, frequently, but you have a forecast, and so that's a really natural fit for a timeline. Calendar is another great example. It's a natural fit for a timeline. We found mismatches though, or some people want something that's a little more immediate and maybe a little more timely. And that's where the relevant configuration comes from. You say this widget is relevant at this place, time, location. And when those things are met, then we give you the runtime to go update the widget and provide the information in that moment. If you have something that's more like a session, a sports game, a ride share that has a more obvious beginning and end, live activities are a really great fit for that. You get substantially more runtime in order to provide substantially more updates for that fixed amount of time, knowing that when it's over, that runtime goes away. But in the moment, we want you to have a really live experience. And if you want to provide, like, I want to perform an action or, you know, how the user just do something quickly, that's what controls are for. So if you try to frame the uses of, like, what you're looking for within each of these types of APIs, I think you'll find more success trying to tailor the experience to the right tool. Yeah, I think on your point about there's just a bunch of different data models for different apps. I'm constantly surprised and challenged by the different ways we find that apps have different requirements and different ways to refresh. And we think we've covered a lot of the bases with the different options we have. We have a lot of different ways to get onto the watch face and the control center and the smart stack and all these things look the same. So, yeah, I mean, I think you got to try the different models. I think even internally we pivot a lot between different ones and you just got to find what works for you. I think that's something that we have. We also find iteratively with our own widgets internally. We sometimes start off with one technology. And as we are working on the design and how it feels, we decided, you know, actually, this is a better fit for relevant configuration. And we started off with a timeline widget. And that's okay. That's, like, the fun part of the discovery. You mentioned the, like, cool thing about the constraints of the platform, right? It's a journey. It's a journey. And so you have to kind of be open to that. But all that being said, I mean, if you feel like there's some case that you have that isn't served well, I mean, we're always open to feedback and trying to learn more about what those cases are. You know, we're you know new suggestions or new conditions. We should make suggestions and new reasons for for refreshes Yeah, well said I love learning the use case of developers because that really helps us get to the heart of the problem One more thing on this. Oh, this isn't necessarily new to this year, but last year in watch was 26. We shipped a PNS space Which are refreshes? Yeah, I think that's part of the original question So there's a whole new path to being able to you know from a server Push new information to your widget and get a refresh and now in this year we have watch connectivity based on that's right Yeah, so we are continuing to try to find more and more opportunities to provide The right signals into the system to provide that like update we want your widgets to also feel fresh, too And so it's just a matter of we mentioned the constraints of the system. It's a matter of finding that right balance So whether it's being pushed from the phone is being pushed from a server it's driven by background app refresh, or you have a timeline. These are all the various different tools we provide. And again, going back to what we were saying earlier, we want to provide all these different tools and opportunities for you to go find the right way to combine them together. And if you just can't quite find that right fit, that's the right time to come to us. That was beautifully said. I love it. Okay, we're nearing the end of our time together, so I'm going to look at one more question to round things out. and there are so many great questions that I really – So many great questions. I know. I'm like, wow. Thank you for being engaged and we will – I would love to engage more on Apple Developer Forms after this. Wow. Thank you all for this. I'm really excited. There's a lot of we love watchOS. This is really nice. We have to love watchOS. We all love watchOS. Who else loves it? I love it. There is a – okay. Okay, LazyVar. We have to end with LazyVar. We have to end with LazyVar. I wish I had like a special badge on the screen to give. Excellent questions today, all of you. And thank you for upvoting your favorite questions. This is the last one for our time together this time around. Okay, LazyVar asks, What are the best practices for optimizing heavy asset fetching on a standalone watchOS apps first launch without relying on watch connectivity? I'd want to keep the initial sync fully independent of iPhone, but struggle with the watch's radio scheduler limitations. Oh, goodness. You know, I think you really think about what you need to get. What can you bundle? Because you're going to have to, you know, you can use background URL to get things. But you're going to be, you know, in the background, that first launch, because you don't have any time before that first launch to do anything. So it's like really think about, you know, what you can do to provide the things that you instantly need on first launch beforehand to provide a good experience while you're getting those other assets that you need. So, you know, provide something for people so that they can start doing something, see something and not see a spinner until they get the other things that they need. So, you know, really think about what are those assets that you need to download versus what can you what can you get and what other experience can you give them that's useful to them in that moment while you're getting those other things. I think we're all smiling because this is like a canonical Apple Watch challenge. But that's what we need to talk about. No, I think it really is. It's what we think about a lot for the experiences that we build, right? And it's kind of the magic trick of working with those constraints. Like, yeah, it is a constrained environment. The connectivity is a challenge. Getting the runtime you need is an obstacle. But like Anne was saying, it's about trying to find the right way to craft a user experience that masks the uncomfortable things that we all have to do as developers to get the assets downloaded on disk and still provide delightful experience for your users. Yeah. And you may not have connectivity. The first launch of that app may be someplace out in the backcountry because we all love our Ultras and our series watches. And we go off the grid with them. And they may launch your app and have no connectivity. And what are they going to see? And so we as watch app developers think about what happens. What happens when I don't have anything and I'm completely offline? What should I do now? I love it's a it's a constraint and it's also an opportunity for something that's truly delightful and forward thinking and making the best watch app experience. And I think those are the moments that really stand out on the platform when it's so thoughtful. And, you know, I said that was the last question, but I have a personal question for me. I don't have a cool username yet, but I'll work on that. I would like to see what watch bands you guys are all showing off because I I'm going with the black one. It's very cool. But I don't know. I think like let's see what and you're wearing, too. I feel like you got to show off for a second. Okay, so I've got the Matchy Matchy Loom and Unity Bloom because it also matches. I love the colors, and this is one of my favorite watch bands. Nice. I like the double. You look like a superhero. It's cool. It's how you really know testing code. That's awesome. Cool. Well, seriously, thank you all for your questions. This has been so much. I've had a lot of fun. I think I can speak for all of you. I know you've also had fun. Have you had fun? We've also had a lot. This has been very rewarding for me. It's really fun to talk about watchOS, talk about the qualities that make your apps so excellent, and keep people coming back from the platform. I love trying your apps, and not just hearing about your triumphs, but also hearing about the things you're still trying to figure out and learn, and pushing each framework to be better on Apple Watch. So I love, we love hearing your feedback and learning about the cool stuff that you're building through all of the channels that we get to connect with. So thank you so much for being a part of the Watch Developer community. It is such a special platform that makes people's everyday lives better through convenient interactions, staying healthy and connected. I want to say thank you to our lovely panelists. Thank you. This was really awesome. Thank you. Thank you. Oh, thank you. Thank you for being such a great host. Yeah, thank you. This was really fun. Thank you to our support staff, helping us go through all the questions. If we didn't get the chance to answer your question today, that's okay. We're still here in different formats. Please visit the Apple Developer Forums at developer.apple.com slash forums, where we'll continue the discussion. We've had great WatchOS and Health and Fitness Q&As throughout the week at DubDub, and we also have people throughout the year who are checking in on your questions. It's a really great way not just to ask questions but to learn from other people's questions. It's a really great learning resource. Sometimes you're hitting a wall and someone else has figured it out. So I really encourage you to use that. And if you have feedback, whether you're running into bugs or have enhancement requests for new stuff you want to do, check out feedback assistant at apple.com. So we really appreciate your feedback. And we want to bring, the big theme is we want to bring the best developer experience to you. So thanks again for joining us. And we hope you have a great WWDC.