I just wish that Apple Music Classical would get a web app. That's the only reason why I'm paying $12 per month. Although I suppose most of the classical pieces are available on regular Apple Music as well...
I use it too...but it's a really poor app. Search is incredibly slow and clunky, discovery is nonexistant, they've still got the same shuffle logic bugs they had in iTunes, etc.
The most annoying part is they CDN from California for UK/EU users. Start playing something not downloaded, hit skip a couple of times and enjoy 10 seconds of buffering.
The Apple Music web app is so terrible it made me doubt the engineering ability of the whole company. Can't sort collections, can't edit playlists, songs randomly getting skipped, etc. etc. It was like they actively hated the users and wanted to punish them for using the web app.
Just a theory. But I think they are preparing for the notion that there's a growing number of PWAs in their app stores. And most of those would be using things like maplibre or Google Maps. So, to address that (given that they can't really stop PWAs), it makes sense to make Apple Maps usable outside of the Apple platforms. This way, people can develop PWA apps and have some level of integration with Apple maps on IOS. Just a theory.
The browser restrictions are probably because developing hardware accelerated map rendering engines for the web is a bit of a project and the support for things like WASM and Web GPU in Safari is probably requiring dealing with some Apple specific quirks. Maybe they'll get around to that eventually. I think for most web developers, no Firefox support would be a show stopper. There's no point to this strategy unless they address that.
They already had MapKit.js as a mapping SDK available across various browsers, off their own hardware. It's been available on a variety of browsers, even Firefox, this is just a beta that doesn't support it.
Do you regularly use one of them? I've been aware of them for years but I've been never motivated to activity use them due to overall poor UX. Sign-in is already a hassle there.
Yet I appreciate them maintaining them because I once had recovered my access to my devices when I was almost locked out of Apple ID (don't quite remember the detail though).
As others have said, they already have a few (Mail, Music, iWork, FaceTime, etc). Like FaceTime coming to browser, and apple music going to android, this is probably an attempt to cast a wider net for their ecosystem. Also the EU DMA law is could be causing some strange behavior here.
I wouldn't be surprised to see apple try and release some expanded subscription that includes mapping features. Not sure what TBH and there have been no leaks, but they're searching for revenue streams, and the App Store is getting eyed by regulators.
Oh and the browser restriction is probably temporary - MapKit.js works on all major platforms, even Firefox, so its safe to assume this will get there too.
So weird! I was just looking for this 2 days ago, and was like, "Huh, I thought they had a web app??" Turns out it was always for devices native only. I had no idea this was coming but 2 days ago was the first time I was looking for a web version! Hahaha! :)
edit: Just tested it. Nice! Faster than Google Maps in my estimation. (panning and zooming the map builds and focuses faster). Google, please don't delete my account for criticism! hahaha! :)
Satellite (actually aerial) images are very expensive to collect and not all that useful.
Oddly, until recently Brisbane's satellite maps on Google were over a decade old even though it showed a current year copyright. More up to date on Apple Maps, except it doesn't have Brisbane in 3D, even though it does have Gold Coast.
This -- Apple highlights rail lines in their appropriate colors, which is an amazing way to visualize how lines are routed. Google's is kind of half-baked in comparison, IMO.
Seems strange Apple offering this for free for other platform users.
This also makes me wonder how much does it cost to run a Map services. I assume the actual server and bandwidth cost are negligible. But the updating and Data would be the most expensive part. But what incentive does Apple have to open this up?
I’m on both iOS, macOS and Linux. One thing that’s keeping me using Google Maps is not having Apple Maps in the browser (on Linux). This definitely could lower the switching threshold.
I was also surprised to see that there is no cost for using Apple Maps (maybe because it's a beta?).
How will this affect services like Google Maps, Mapbox, and similar providers?
Protomaps makes hosting maps pretty cheap for open streetmap vector tiles. Most of the cost is actually the CDN bandwidth. It's not going to be nothing depending on the number of users but it shouldn't break the bank for Apple and probably is relatively low to other content they distribute (e.g. Apple TV) or OS updates.
The way protomaps does this is by serving a single large file with all the map data via bucket storage and then using lambda functions + CDNs to extract tiles from there on demand. So, they don't pre-calculate the tile files and this simplifies the update process to replacing a single file. The CDN caches the extracted tiles so this is relatively cheap and doable even for small startups. So, this minimizes compute and storage.
Generating the map tiles requires a bit of compute obviously but it's a constant overhead; and they have to do this anyway for their native apps.
Probably the hardest part for them was building a hardware accelerated render engine for the web. Similar to Maplibre, Google Maps, etc. That would explain why it doesn't work on Firefox as well. And obviously Safari is a bit lagging with things like web GPU and WASM that I imagine would be useful for this.
I fear running all these services is expensive too - not just the data & updates.
You need quite a lot of infrastructure:
1. map tiles
2. satellite view
3. geocoding. Where you have several services like forward, reverse, IP2coord. Likely also different services for different countries.
4. A-B routing. Again with several services like car, bike, walking and transit. Especially transit is a completely different thing. Also traffic data requires a different data pipeline.
Does this in some way imply that the web is a significant part of their strategy?
The button in the bottom left explains why this exists: It's a gateway to get more information about businesses and other attractions from entities out there in the world that don't live in the Apple ecosystem. Apple Maps ultimately needs a direct line to the real world to be maximally useful, especially against its competitors, and this is their attempt to build that bridge.
Pretty much. Even as a disillusioned mac user I wouldn't rely on their maps app but no compatibility with my browser of choice means that I'll never bother using their product under any circumstances.
Ha ha ha Apple still thinks 'the web' is Apple or Microsoft
(https://support.apple.com/en-us/120585) and Firefox is not supported at all.
It's time to shake the last rotten apple from the tree.
...which only underscores how pointless this is: if it works in Chrome on MacOS and Windows (https://support.apple.com/en-us/120585), it will also work on Linux, so why exclude Linux?!
I'm impressed with how well they've enforced that as well. I tried spoofing my UA to be safari (which I fully expected to not work), but it also didn't accept when I set my UA to Chrome.
What's especially odd is that Apple acknowledges Firefox's existence in their WWDC videos about web features, when they mention browser compatibility or who they're working with.
Like others have pointed out, it seems to work fine in other browsers once you trick it into letting you in. General compatibility doesn't seem to be an issue. So, what is it that Firefox and Chrome on Linux (and only on Linux) don't support?
H.265 is what they don't support. I'm not an avid enough user to know where Apple Maps makes use of media, but the source code contains media player controls, so it must somewhere. Retrofitting compatibility by launch may be as simple as re-encoding the H.265 content. Not at all worth the effort for beta 1, but with an obvious path forward.
> So, what is it that Firefox and Chrome on Linux (and only on Linux) don't support?
H.265 is what they don't support.
Do codecs need to be supported by the browser itself? I thought this was unloaded to some media decoding framework. Linux does have h.265 support at least in mpv.
> Do codecs need to be supported by the browser itself?
Not necessarily. The browser could defer to licensing established by the operating system vendor, but Firefox places the expectation upon itself to have parity across platforms and to not support encumbered technologies.
> Linux does have h.265 support at least in mpv.
And if you've negotiated the licensing fees you can even use it, but chances are... Microsoft and Apple have dealt with the licensing for you on their platforms, so the ballgame is different there.
> To start, Maps on the web is available only in English. Maps on the web will be available for additional browsers, platforms, and languages soon.
Published Date: July 24, 2024
> browser compatibility could be something for the launch
This is indeed how many bad/junior engineers approach this issue but it's backward - anyone with any experience doing launch QA knows well that browser compat needs to be built in from Day 1 - retrofitting it is disastrously expensive from a launch-delays perspective.
You cannot fix bugs if you don’t collect them. Neither Mozilla. If you have not enough resources, just collect and track. Fix them when more people are available.
Same for native application ports, ship them as early as possible. Just mark them beta or alpha. At least you collect bugs. Bonus, you filter which are generic issues and which are platform dependent issues.
If it is in such immature condition it should be kept internal.
If it doesn’t work at all in a web-browser which handles HTML5 and modern CS it is probably not a website - but a proprietary protocol which needs a special client-application.
I thought the same, until I realised I still had `/unsupported` in the URL. Spoofing a Chrome UA and dropping that path from the URL let me load (and use) Apple Maps fine under Firefox.
I normally find that stuff I build for the web just works in Chrome and Firefox and it’s Safari that requires hacks and workarounds, even when I’m using standard APIs that are widely supported. I’d have to go out of my way to have something work in Chrome but not Firefox.
Apple says that MapKit.JS works on Firefox, so this beta web page is probably just working out bugs before they release for FF. Perhaps a rendering issue?
And even in their supported browsers (Chrome at least) I got the "unsupported browser" on Fedora Linux.. Wonder what makes a online map need such a specific (even if its widely used) setup.
Everything works if you use User-Agent switcher extension. So they went through the trouble of making an "unsupported" page and redirecting you to that page instead of doing nothing
Of course I can. Add cleaning water, check oil levels, replace a light bulb. No much else I can do, but others may, and other won't even do any of this.
Point is, this is not a binary choice. Between user and developer there are many people with varying skills that will use a user-agent switcher if needed.
In my experience (systems engineer/devops for both Windows and Linux for more than 25 years), very few users are actually savvy. Even those working in tech.
there is a good reason that most of the people prefer apple for its simplicity, its because apple only shows you what is required. i agree with you there.
Extremely frustrating. If a user is smart enough to use Firefox, they're probably also smart enough to open another browser if a site does not happen to work on Firefox. (Which I haven't experienced for a while, except when using ESPHome which requires WebSerial)
* Epiphany with WebKit2-Renderengine. The literally block their own engine.
* Firefox with Gecko.
What year is it? 2001?
No web developer should be allowed to “block” webbrowser. Test for features and say “this thing doesn’t work because of and I don’t care about another solution”. Same shitty experience with Microsoft Teams which blocked - at least some months ago - the call buttons for Firefox, despite everything works fine. And Confluence which claims they don’t block but started, Epipany is now hiding as Safari and…surprise…everything works.
Interestingly it works on Opera although the colours are weird (lots of dark greens and blues). On both versions (Edge and Opera), my local bakery is mis-located (by hundreds of yards) compared with its (correctly) reported location on an iPhone.
It's not financially worth supporting, Firefox has 6.53% of desktop and 0.53% of mobile marketshare (Statcounter), with a switching cost of zero if users encounter a breaking issue.
Not surprising it got to this point, Mozilla has been stagnant on features most users care about and catered exclusively to the privacy crowd for years - which isn't a large group and competes with Chromium offshoots (giving it a smaller niche, privacy but demanding an alternative rendering engine).
Sure, but the QA cost to support Firefox is significantly higher than the small fraction of people that will refuse to use a site that doesn't support it when they encounter an issue.
> catered exclusively to the privacy crowd for years
Not even that. Firefox on iOS doesn’t have an integrated adblocker. It’s been requested for years at this point, and browsers like Brave do have one. Pure unwillingness. It’s why I got all my non-techie family and friends to switch to Brave.
Seems like a baseless restriction. I can't find anything wrong with Firefox support itself as I changed my user agent under Firefox and Apple Maps works fine.
It sucks when companies restrict normal access to a website when it's uncalled for. It's not the first time I've gotten "Use Google Chrome" for no reason.
"Not supported" on its own does not mean "actively blocked/disabled", which is what this is.
"Supported" means that they provide a certain effort for make the configuration operational for their users, by designing said support if needed and providing assistance as required.
Yeah they are just being lazy and not testing it... so rather than verify it works and fix bugs, they just check your browser agent and redirect you if you aren't using Chrome or Safari. So ghetto! Reeks of the late-90s/early-2000s "Use IE6" messages that companies used to put out when they built a site using Microsoft web components or proprietary APIs.
"Hey look, I can save my PowerPoint as a web page! And it even has the animations!" Except it's 2024, and we have standards, and for them to say, "Oh we don't adhere to the standards" is shockingly bad.
> Not only Firefox is not supported, but even Chrome on Linux doesn't work.
Which strongly suggests that it makes use of H.265 content somewhere (the source code corroborates such functionality), likely as a carry over from content created for the iOS/macOS versions of Apple Maps where support is granted.
> It's embarrassing for a company such as Apple.
To be fair, it is still in beta. There is still plenty of time for them to recreate the content in a format with wider support before release.
Much more embarrassing is that we enable this state of affairs. The situation that keeps Firefox and Linux from jumping all over H.265 is not some natural property of the universe, it's just a social construct that we uphold by willing choice.
> The situation that keeps Firefox and Linux from jumping all over H.265 is not some natural property of the universe, it's just a social construct that we uphold by willing choice.
Can you elaborate and/or link me to anything related to this?
Patents. To distribute the codec itself or content, you might have to pay patent fees.
For codecs, they are not flat fee[1], but per piece shipped. Which obviously, presents a problem for linux distributions. Even if they had money, they cannot count how many instances there are.
[1] Well, there is a ceiling, if you ship a insanely huge number of them. Linux isn't it. Cisco is, which is why we have openh264 binaries by them.
Basically, H.265 is based on some patents and you would have to license them to be allowed to implement the Codec. Mozilla categorically doesn’t want to do that until they can implement it without any patents.
I understand they "cannot test every possible browser" and that "users may get subpar experience".
I don't understand why there isn't "continue at your own risk" button. Maybe with a scary warning. It's kind of stupid that I have to spoof UA for a website to let me in. And in most cases, everything just works fine.
Maybe one day I'll create a website to inform about the issue.
Firefox, while still significant, represents a smaller portion of users. But I think that the absence of initial support for Firefox doesn’t necessarily mean it won’t be supported in the future. Yet I hope so.
I never understood the value proposition of Apple Maps. Can you imagine being the executive, deciding to create Apple Maps? "Ok, how much does it cost to build and maintain? $BIG NUMBER. What? No way. We'll never make it back by selling adverts on the map." And, still, they built it. We have heard many times on HN that Google Maps (virtually) throws money out the window to keep it running so smoothly. Just keeping all the transit info correct for suggesting routes must be a nightmare.
Yeah, it's a feature enabled by default outside of the EU (in the EU it asks you if you want to enable it). Makes for some fun stats/recaps, and is useful for tracing back steps (wait, where was that awesome store/restaurant/park/whatever we went to while on a trip to XYZ?) at the expense of Google knowing a lot about you.
On iPhone I only see Signifcant Locations; on my phone I only see a list of 3 places (despite 400 records). Compared to Google Timeline it’s much more curtailed function.
Whilst I agree with what you say I'm so grateful for Apple Maps simply on the grounds that I try and use Google products as little as possible. Things like Apple Maps keep me in the Apple ecosystem as they add value to my life. I wouldn't use Apple CarPlay either if I had to use Google Maps (granted, I know Waze and others also exist).
It's defensive, (and it was built at a time when money was free).
The iPhone launched with Google Maps. Then Google decided to push feature updates skewed towards android phones, leaving iPhone users behind. Apple saw that a vendor could screw their users over (and potentially cause defectors), and decided to invest to ensure they don't have a dependancy.
The best part is that they can now offer it to App Store developers as a free iOS SDK (and paid API on web). Meanwhile the same developers would have to pay an exorbitant cost to use Google Maps. It's part of the moat that makes iOS the more profitable platform to develop for. You can also see this playbook with the release of free Weather APIs.
Yea Apple/Google maps has to be expensive to build and maintain, but at least for apple, they were able to buy their way to bootstrapping the map. What's impressed me is all the fly-over and custom 3D modeling they've done. It does really feel like they just wanted to make a good map at some point, even beyond what people needed or expected. That said, mapping products probably has good caching and fault tolerance you can design in to reduce cost - maps don't go out of sync that fast (for caching) and you'd never know if their "suggested routes" data was out of date occasionally, because you can never drive both routes at once.
> The best part is that they can now offer it to App Store developers as a free SDK. Meanwhile the same developers would have to pay an exorbitant cost to use Google Maps.
Apple Mapkit is free up to 25K api call a day, after that you have to contact Apple for more (and pay I guess?).
At the time, Google Maps on iOS was written by Apple, not Google, and Google was holding back API access for Street View until Apple sent back more location/tracking/demographic data on users that Google wanted.
Rather than sell out their users, Apple dropped Google Maps as the backend and launched their own maps, and then let Google write their own Maps app where they could do anything they wanted.
Maps are core technology, which Apple prefers to own. Imagine wanting to release CarPlay (or a full blown car) and Google having you by the balls over maps and navigation. That wouldn't be a good situation. As to $BIG_NUMBER, they seem to be managing fine - Maps sucked pretty bad when it came out, but it doesn't suck now, I prefer it to Google Maps where I live.
Maps is table stakes for a smartphone, and having such a key feature provided by your main competitor is a huge risk. So purely on that basis, it could be worth it.
Then, on top of that, there is value in the data you're able to collect. Traffic data is really valuable. Tracking the movement of vehicles and pedestrians lets you create very accurate maps based on "real world" data, you could use it to figure out really specific things like traffic light timings, diversions, pedestrian crossings, parking space, layout of private roads...
At one point, Apple was working on a car, if you were making a self driving car, all that data would be useful for you, and beacuse of the value of it, competitors may not even sell it to you. So your only option is to generate it yourself.
As for transit data, that is fairly simple, most transit agencies will publish their timetables in GTFS format, there are tools to automatically export this in scheduling software. That will probably get your 90% of the way there, so you might have a few on the groud people in major cities to tweak and make it more accurate, which is nothing for a company on the scale of Apple.
> Then, on top of that, there is value in the data you're able to collect. Traffic data is really valuable. Tracking the movement of vehicles and pedestrians [...]
...but then they decided to market themselves as "privacy-focused", so they can't really do that, right? Or are they actually doing it?
> At one point, Apple was working on a car
...but then they killed the car project, so that goes out of the window too.
Allegedly, Apple have built in privacy features so they can't associate individual users with routes, or know what the entire route is[1]. Apple does show traffic data in the app, so they obviously do collect the data somehow.
When Apple built maps, the car project was still alive, so it would have been a factor in deciding on the investment. They could still partner with a car manufacturer and use the data.
I do suspect that my first point was key in green lighting Apple Maps. Google could have asked for more and more money to provide maps for Apple, or they could pull out completely, and force users to use the App Store app, which would have left the product direction of Maps completely out of Apple's hands.
I haven't been an employee since 2015, but by then Google had already been doing the route trimming and splicing for live traffic data. (If you had location history enabled, some of that same data at lower granularity was stored in another service, of course)
Collecting dots/vectors on a map doesn't necessarily invade my privacy. The problem comes with linking that dot with a person. As long as that link is lost and unrecoverable, I have no problem with Apple (or anyone) collecting it. The second problem is actually ensuring that.
The main problem with this is that the data is naturally linked to your phone, and you have to trust the provider to anonymize it. I suspect that's at least part of the reason for Apple painting itself as privacy-friendly: building trust with its users that they won't misuse their data.
I mean most of those vectors will converge on my home dot; with time data any vector intersecting with my home can tell a lot about my life. Additionally, is it anonymized per user (ie all my vectors are still a set just not identified as me) or each vector is an individual product unliked from all other vectors and user data.
> Additionally, when you use Maps to make a navigation or directions request, details about your route are sent to Apple, including:
> [...]
> A random identifier, which is created when you ask for directions and exists for the duration of your navigation session
That’s just because your phone sends it, but your phone also has a list of “significant locations,” around which it could avoid sending data for a mile or less.
> but then they decided to market themselves as "privacy-focused", so they can't really do that, right? Or are they actually doing it?
Here's the genius behind Apple's marketing: when they say "privacy" they (mostly) don't mean from them! They are mainly talking about third parties. Apple collects a ton of first-party data, and nobody seems to be concerned about that. I also the pond Apple swims in (big tech) is so disgusting and polluted that even their minor effort at cleanliness seems pretty good.
Apple has a lot of technical solutions that mean data is collected, but is never associated with a particular user.
As an example, location data is shared with Apple, but it’s associated with a random unique identifier rather than your account. When your trip ends, your device switches to a new identifier. Traffic information is only shared if a certain threshold of users travel on a route [1].
Other examples include the entirely on-device photo scanning, the same rotating identifier system for transcripts of Siri interactions, etc. and, of course, being the only major cloud provider to offer E2EE on everything.
Not perfect, but a huge difference from their competitors.
I do appreciate their sharing that, but I hate that it requires entirely just trusting them. They've so locked the user out of the device that it's difficult or impossible to verify anything for yourself, and even if you did, they could trivially push a change at any time because they have ultimate control over the device.
On the flip side, I tend to think a company so large would have at least one whistleblower or something on the inside, and/or would be so concerned about legal fallout that they wouldn't risk it.
On the flip side of the flip side, Apple is notoriously secretive (even among insiders) and very tight-fisted around employees sharing/leaking information. They also have some of the best lawyers in the world and a near infinite ability to fund any legal action, so may feel (and in fact, be) untouchable. And should Apple go evil, there aren't really great alternatives anyway for the average person, and they're generally so invested in the walled garden that walking away would entail a major disruption to their life.
I agree though, while not perfect, they are certainly much better than their competitors (not counting small players, e.g. GrapheneOS), and I'm grateful that at least they keep privacy at the forefront of conversation. If they abandoned it, there'd be nobody to pick up the mantle.
Back in the days Google notoriously launched turn-by-turn navigation on Android only. They bet on this being a big enough differentiator for people to use Android over iPhones.
Apple then launched Apple maps - which at some point became quite good. Google quickly learned that they can't afford to make Android specific features in their apps or they risk losing large percentage of iOS users if Apple makes a competing product
If Apple didn't respond with making their own maps, then maybe we would see more and more Android specific features, to the point where Android would become the dominating platform
But this is also exactly the same game Apple plays against Android users. It's the same reason why iMessage bubbles are green for Android. Google won the maps round, but such wins are vanishingly rare against Apple.
There are non-Android devices that can send texts as well; they also appear as green. It's probably more accurate to say that encrypted messages are blue and unencrypted are green. Look at the recent AT&T hack to see why the difference matters.
Even if that was more accurate (I don't think it is), it's certainly not the way users see it.
In fact that's NOT the way Apple describes it, either (see the Apple article cited above), because Apple doesn't actually want to enable E2EE -- it only wants to be able to say it offers it.
In practice, ensuring that other users are pressured into choosing iMessage on iPhone is the only thing that matters to Apple.
And, this very simple trick works extremely well: at least 87% of teenagers in the U.S. (https://mashable.com/article/apple-messages-green-doj) are pre-programmed to buy an iPhone, even though they have the lowest disposable income of all. Meanwhile, less than a third of the overall global population owns an iPhone.
Is that because iPhones are better? As an owner of both a recent Pro Max and Pixel Pro, I can unequivocally answer, "no", but I do find all of the annoyances between cross-device communication accrue to the point of just wanting to switch to my iPhone full-time, even though it's arguably a worse experience in many ways.
and services like e.g. SMS text reminders from Internet services do no run on Android. The green is not a signifier of Android, just of non-encrypted. Or non-Apple, if you want to be less precise. (Apple devices where encryption is disabled also appear as green.)
"Is it Apple Maps bad?" --Gavin Belson, Silicon Valley
After the fiasco from their initial app launch, I'm sure they would have preferred not to be a meme in a sitcom if possible on this go round. It is possible to release too early
Sorry, i didn't mean to be disingenuous. i meant, ads are not the main source of its income.
And in this context, that's why it is not a foolish choice to spend money on something that it's hard to sell ads on as long as it helps sell more iPhones.
Google Maps had a total monopoly and Google could have leveraged that in the competition between Android and iOS. Maybe they even tried asking Apple for a lot of money to be able to use it on iOS.
It takes years, even a decade to get maps to a good quality (Apple maps launched in 2012). So I think it's a good thing that Apple started early enough. I'm sure it's crazy expensive to build and maintain. Apple can fund it from iPhone sales, and ensure that their ecosystem has an alternative for Google maps.
I don't think it's meant to turn a profit, I think it's meant as protection of their iPhone revenue.
> I never understood the value proposition of Apple Maps.
They ship their operating systems with all the "common" apps pre-installed (e.g.: Email, Calendar, Reminders, Notes, Maps, etc). For the maps to work, they need some data source. That's what Apple Maps is.
Apple doesn't make money with the Email app directly, but its existence likely improves how users perceive iOS. This probably translates to return customers and more people recommending it.
Privacy and a vastly better navigation experience is what makes me prefer Apple Maps for turn by turn nav. For finding local businesses Google Maps is better
Google Maps on iOS works terribly where I am. Current and previous phone. Going through the Caldecott tunnel would fast forward all the stops. Switched to Apple Maps and I’ve been very happy. Just a single glitch noticed (a light appears before a freeway onramp).
One of the selling points of Apple devices is that their software is [1] just _nice_ to use, letting you do what you need to do, without having to keep you in and monetize you otherwise.
Is Mail.app the most powerful client on earth? No; but it is Good Enough, and I don't have to download and pay for a third party app.
Is Weather.app the best weather app with all the bells and whistles possible? No; but I don't care about weather apps to download and trial fifteen other ones and It Just Works.
Maps are (orders of magnitude) more complicated; but arguably are also on the baseline level of functionality for a modern mobile OS.
And Maps.app is just so much _nicer_ to use than Google Maps. It has the same problems that all Apple products like it does (search is atrocious, POI db is bad); but it is just a much more pleasant product. It looks nicer, it _feels_ nicer, it has best-in-class transit directions, and doesn't shove ads in front of my face.
[1]: Arguably getting worse and worse at it every year; but still miles ahead of everyone else.
The Maps application on iOS used to use Google Maps. But then Google started to collect too much user data and withholding features like turn-by-turn navigation (while making it available on Android).
At the time Apple Maps came out, Google Maps on iOS was limited to bitmap tiles and had no turn by turn directions, whereas Google Maps on Android had both dynamic vector based maps and turn by turn directions.
Apple Maps forced Google to improve Google Maps on iOS.
Apple Maps data was definitely substandard when it was released, but it has improved considerably since then. I vastly prefer it to Google Maps, especially for turn by turn directions when I'm driving.
I checked with curl and no matter the user agent it will answer with a 302 to google maps, I don't know how it works that others say it will redirect them to the Apple Maps App.
Yeah what it's showing is that on iOS/Safari (user agent) it returns a maps:// protocol address. This makes sense of all the reports here - it's trying to direct you to whatever maps app you've selected. Redirecting to google maps makes sense if the server doesn't think the maps:// result will do that.
If you have a link to maps.apple.com on a platform with app links (or claimed HTTPS URLs on android platforms), a native Apple Maps app can display the results. Otherwise, they haven't had a web version, so they would change it to a redirect to Google Maps.
I can't speak as to whether Chrome or Firefox on Mac, or alternate browsers on iOS, support this OS feature to launch native apps which have claimed a URL path.
The defacto query parameters from Google Maps (at least historically) are supported by Apple Maps as well.
This made migration to Apple Maps easy when it launched - you can just change the domain in the URL and prefer the native experience, knowing it will fall back to Google Maps in other cases.
Once Apple's web experience is finished, I imagine they will stop redirecting to Google Maps. Google has somewhat compromised the privacy posture of Google Maps, which may very well be how Apple internally justified building a web experience.
Firefox is being marginalised mostly by Mozilla corporation others seem to just don't care much about dying browser, it is weird that it is broken on Safari with iOS for some folks as it is Apple product.
You need to be specific. It's very much location based in terms of data quality and how current it is. It's great in the SF Bay Area, which isn't surprising given where Apple is based.
First problem: the searchbox does not get focus upon opening. Looking for a place is the main thing you do. Why does it require mouse handling and handling at all?
No keyboard response to Escape. It's basically the maps widget with a user unfriendly, but nice looking, drawer
Same with the Contacts app on macos, it's slow, crashes, and I doubt anyone uses it
Other example: voice memos. Takes 20 seconds to start recording if you have a long list (30+) of recordings. It needs to find the next available file name. Instead of just listing them, it probably loads every files and tries to read some xml metadata instead.
It appears to check viewport width alongside user agent. Shrinking my browser down in responsive mode results in the same "browser unsupported" screen.
Good job apple team! Very smooth experience. Fyi you may want to sanitize some of your response headers because one can easily tell Envoy is running at the edge. Upstream service latency looks healthy though :)
Why is that? Having Google's ability to navigate with live traffic data isn't a valuable feature to you? Apple's traffic-flow is mostly a joke to me, and I've never seen anyone trust it.
Anecdotally: I use Apple Maps when I need directions (mostly because it’s native/integrated and not google), for drives over an hour my experience is that the ETA time is +/- 5min even when there is lots of traffic.
Except in one edge case where my girlfriend and I were doing a 7 hour drive traveling east late at night on an empty highway and our eta increased by an hour then a little while later another hour, we were so confused and thought we might be driving in the wrong direction! Until we figured out we had crossed a time zone and it was also day light savings!
my colloquial evidence …
apple maps is the most accurate in predicting time to destination and handles network instability in a way i prefer (keeps you on the track and just notifies you’re in offline mode)
google maps suggests more alternative routes that may save me time but their predictions are generally less accurate. network instability seems to cause the application to “panic” and it just starts spinning around - especially when walking through downtown areas
while google has a sleeker presentation of traffic and shows the “red highlighter of misery and frustration” on my map more precisely, it’s timing is generally wildly incorrect and apple has already routed me around the problem and with more accurate time to destination estimates
Apple gives better verbal instructions, e.g. "go past this light, then at the next, turn right" and it neatly shows which lane to be in. I can get where I am going even without looking at the map. Last time I used Google Maps it would give you no clues until it was basically "MAKE A HARD RIGHT NOW".
562 comments
[ 2.0 ms ] story [ 578 ms ] threadWhat year is it?
What's with the browser restrictions too?
The most annoying part is they CDN from California for UK/EU users. Start playing something not downloaded, hit skip a couple of times and enjoy 10 seconds of buffering.
The browser restrictions are probably because developing hardware accelerated map rendering engines for the web is a bit of a project and the support for things like WASM and Web GPU in Safari is probably requiring dealing with some Apple specific quirks. Maybe they'll get around to that eventually. I think for most web developers, no Firefox support would be a show stopper. There's no point to this strategy unless they address that.
KeyNote, Mail, Numbers, Pages and Photos have been available as web apps for years (and minor ones such as Notes and Reminders)
I think a main reason Keynote, Numbers and Pages web apps exist is for sharing with non-Mac users.
I wouldn't be surprised to see apple try and release some expanded subscription that includes mapping features. Not sure what TBH and there have been no leaks, but they're searching for revenue streams, and the App Store is getting eyed by regulators.
Oh and the browser restriction is probably temporary - MapKit.js works on all major platforms, even Firefox, so its safe to assume this will get there too.
edit: Just tested it. Nice! Faster than Google Maps in my estimation. (panning and zooming the map builds and focuses faster). Google, please don't delete my account for criticism! hahaha! :)
Supported on your Windows PC: Edge Chrome
https://x.com/jxeeno/status/1814975093380116783
What I have noticed is that the satellite view for Sydney is over 3 years old.
Oddly, until recently Brisbane's satellite maps on Google were over a decade old even though it showed a current year copyright. More up to date on Apple Maps, except it doesn't have Brisbane in 3D, even though it does have Gold Coast.
This also makes me wonder how much does it cost to run a Map services. I assume the actual server and bandwidth cost are negligible. But the updating and Data would be the most expensive part. But what incentive does Apple have to open this up?
Seems you already figured it out: Access to more data / updates. Hence the "Have a Business on Maps?" being a prominent feature.
The way protomaps does this is by serving a single large file with all the map data via bucket storage and then using lambda functions + CDNs to extract tiles from there on demand. So, they don't pre-calculate the tile files and this simplifies the update process to replacing a single file. The CDN caches the extracted tiles so this is relatively cheap and doable even for small startups. So, this minimizes compute and storage.
Generating the map tiles requires a bit of compute obviously but it's a constant overhead; and they have to do this anyway for their native apps.
Probably the hardest part for them was building a hardware accelerated render engine for the web. Similar to Maplibre, Google Maps, etc. That would explain why it doesn't work on Firefox as well. And obviously Safari is a bit lagging with things like web GPU and WASM that I imagine would be useful for this.
https://appleinsider.com/articles/22/08/21/apple-maps-could-...
Makes more sense if you consider that.
[1] - https://news.ycombinator.com/item?id=40908310
You need quite a lot of infrastructure:
1. map tiles
2. satellite view
3. geocoding. Where you have several services like forward, reverse, IP2coord. Likely also different services for different countries.
4. A-B routing. Again with several services like car, bike, walking and transit. Especially transit is a completely different thing. Also traffic data requires a different data pipeline.
5. ratings / reviews
6. user data (when logged in) for preferences etc
The button in the bottom left explains why this exists: It's a gateway to get more information about businesses and other attractions from entities out there in the world that don't live in the Apple ecosystem. Apple Maps ultimately needs a direct line to the real world to be maximally useful, especially against its competitors, and this is their attempt to build that bridge.
What's especially odd is that Apple acknowledges Firefox's existence in their WWDC videos about web features, when they mention browser compatibility or who they're working with.
I’m a disappointed Firefox user but I also know what Beta means.
It's extremely hard to retrofit compatibility onto products. Case in point: all the "we only work in Chrome" sites that use Chrome-only APIs.
H.265 is what they don't support. I'm not an avid enough user to know where Apple Maps makes use of media, but the source code contains media player controls, so it must somewhere. Retrofitting compatibility by launch may be as simple as re-encoding the H.265 content. Not at all worth the effort for beta 1, but with an obvious path forward.
Do codecs need to be supported by the browser itself? I thought this was unloaded to some media decoding framework. Linux does have h.265 support at least in mpv.
Not necessarily. The browser could defer to licensing established by the operating system vendor, but Firefox places the expectation upon itself to have parity across platforms and to not support encumbered technologies.
> Linux does have h.265 support at least in mpv.
And if you've negotiated the licensing fees you can even use it, but chances are... Microsoft and Apple have dealt with the licensing for you on their platforms, so the ballgame is different there.
> To start, Maps on the web is available only in English. Maps on the web will be available for additional browsers, platforms, and languages soon. Published Date: July 24, 2024
This is indeed how many bad/junior engineers approach this issue but it's backward - anyone with any experience doing launch QA knows well that browser compat needs to be built in from Day 1 - retrofitting it is disastrously expensive from a launch-delays perspective.
Same for native application ports, ship them as early as possible. Just mark them beta or alpha. At least you collect bugs. Bonus, you filter which are generic issues and which are platform dependent issues.
If it doesn’t work at all in a web-browser which handles HTML5 and modern CS it is probably not a website - but a proprietary protocol which needs a special client-application.
https://imgur.com/LtS3jXD
https://developer.apple.com/documentation/mapkitjs/
https://developer.apple.com/documentation/mapkitjs/
Not all Apple users are clueless.
A user can easily enable the developer tools if needed, same way I'm not a mechanic but can open the hood of my car.
Point is, this is not a binary choice. Between user and developer there are many people with varying skills that will use a user-agent switcher if needed.
Most people could do a bit under the hood of a car, but they simply don’t care.
It’s the same thing with computers. Most users are savvy. They just don’t care.
No web developer should be allowed to “block” webbrowser. Test for features and say “this thing doesn’t work because of and I don’t care about another solution”. Same shitty experience with Microsoft Teams which blocked - at least some months ago - the call buttons for Firefox, despite everything works fine. And Confluence which claims they don’t block but started, Epipany is now hiding as Safari and…surprise…everything works.
> Support for additional languages, browsers, and platforms will be expanded over time.
This is a beta. You have to start somewhere.
You don’t want broken screenshots shared. Also, people will click through and still open support tickets.
Not surprising it got to this point, Mozilla has been stagnant on features most users care about and catered exclusively to the privacy crowd for years - which isn't a large group and competes with Chromium offshoots (giving it a smaller niche, privacy but demanding an alternative rendering engine).
Not even that. Firefox on iOS doesn’t have an integrated adblocker. It’s been requested for years at this point, and browsers like Brave do have one. Pure unwillingness. It’s why I got all my non-techie family and friends to switch to Brave.
So you cannot view maps from Safari browser on iPhone.
I am curious what they use that Firefox is not supported…
It sucks when companies restrict normal access to a website when it's uncalled for. It's not the first time I've gotten "Use Google Chrome" for no reason.
Its dwilling 3% market share means it no longer makes into project delivery acceptance browser matrices.
It is the good will of some that keeps it around.
Ugh.
It may or may not work, but since that is a bad user experience, they disable it.
"Supported" means that they provide a certain effort for make the configuration operational for their users, by designing said support if needed and providing assistance as required.
"Hey look, I can save my PowerPoint as a web page! And it even has the animations!" Except it's 2024, and we have standards, and for them to say, "Oh we don't adhere to the standards" is shockingly bad.
Welcome to 2007 I guess!
Which strongly suggests that it makes use of H.265 content somewhere (the source code corroborates such functionality), likely as a carry over from content created for the iOS/macOS versions of Apple Maps where support is granted.
> It's embarrassing for a company such as Apple.
To be fair, it is still in beta. There is still plenty of time for them to recreate the content in a format with wider support before release.
Much more embarrassing is that we enable this state of affairs. The situation that keeps Firefox and Linux from jumping all over H.265 is not some natural property of the universe, it's just a social construct that we uphold by willing choice.
Can you elaborate and/or link me to anything related to this?
For codecs, they are not flat fee[1], but per piece shipped. Which obviously, presents a problem for linux distributions. Even if they had money, they cannot count how many instances there are.
[1] Well, there is a ceiling, if you ship a insanely huge number of them. Linux isn't it. Cisco is, which is why we have openh264 binaries by them.
Basically, H.265 is based on some patents and you would have to license them to be allowed to implement the Codec. Mozilla categorically doesn’t want to do that until they can implement it without any patents.
I don't understand why there isn't "continue at your own risk" button. Maybe with a scary warning. It's kind of stupid that I have to spoof UA for a website to let me in. And in most cases, everything just works fine.
Maybe one day I'll create a website to inform about the issue.
I like that it tells me what lane to be in, so it's my main mapping app. Also presumably better privacy than Google Maps.
Yeah you might say that.
My Android-owning Irish mate got hammered one night. Had no idea where he’d been.
We launched Google Maps and it had a GPS track of his entire night. Like a dotted map with every step he’d taken.
https://www.theverge.com/2024/6/5/24172204/google-maps-delet...
Google maps does that too.
https://www.androidauthority.com/google-maps-pop-up-ad-34581...
The iPhone launched with Google Maps. Then Google decided to push feature updates skewed towards android phones, leaving iPhone users behind. Apple saw that a vendor could screw their users over (and potentially cause defectors), and decided to invest to ensure they don't have a dependancy.
The best part is that they can now offer it to App Store developers as a free iOS SDK (and paid API on web). Meanwhile the same developers would have to pay an exorbitant cost to use Google Maps. It's part of the moat that makes iOS the more profitable platform to develop for. You can also see this playbook with the release of free Weather APIs.
Yea Apple/Google maps has to be expensive to build and maintain, but at least for apple, they were able to buy their way to bootstrapping the map. What's impressed me is all the fly-over and custom 3D modeling they've done. It does really feel like they just wanted to make a good map at some point, even beyond what people needed or expected. That said, mapping products probably has good caching and fault tolerance you can design in to reduce cost - maps don't go out of sync that fast (for caching) and you'd never know if their "suggested routes" data was out of date occasionally, because you can never drive both routes at once.
Apple Mapkit is free up to 25K api call a day, after that you have to contact Apple for more (and pay I guess?).
Rather than sell out their users, Apple dropped Google Maps as the backend and launched their own maps, and then let Google write their own Maps app where they could do anything they wanted.
Then, on top of that, there is value in the data you're able to collect. Traffic data is really valuable. Tracking the movement of vehicles and pedestrians lets you create very accurate maps based on "real world" data, you could use it to figure out really specific things like traffic light timings, diversions, pedestrian crossings, parking space, layout of private roads...
At one point, Apple was working on a car, if you were making a self driving car, all that data would be useful for you, and beacuse of the value of it, competitors may not even sell it to you. So your only option is to generate it yourself.
As for transit data, that is fairly simple, most transit agencies will publish their timetables in GTFS format, there are tools to automatically export this in scheduling software. That will probably get your 90% of the way there, so you might have a few on the groud people in major cities to tweak and make it more accurate, which is nothing for a company on the scale of Apple.
...but then they decided to market themselves as "privacy-focused", so they can't really do that, right? Or are they actually doing it?
> At one point, Apple was working on a car
...but then they killed the car project, so that goes out of the window too.
When Apple built maps, the car project was still alive, so it would have been a factor in deciding on the investment. They could still partner with a car manufacturer and use the data.
I do suspect that my first point was key in green lighting Apple Maps. Google could have asked for more and more money to provide maps for Apple, or they could pull out completely, and force users to use the App Store app, which would have left the product direction of Maps completely out of Apple's hands.
[1] https://www.idownloadblog.com/2019/03/13/apple-maps-navigati...
That’s just because your phone sends it, but your phone also has a list of “significant locations,” around which it could avoid sending data for a mile or less.
Here's the genius behind Apple's marketing: when they say "privacy" they (mostly) don't mean from them! They are mainly talking about third parties. Apple collects a ton of first-party data, and nobody seems to be concerned about that. I also the pond Apple swims in (big tech) is so disgusting and polluted that even their minor effort at cleanliness seems pretty good.
As an example, location data is shared with Apple, but it’s associated with a random unique identifier rather than your account. When your trip ends, your device switches to a new identifier. Traffic information is only shared if a certain threshold of users travel on a route [1].
Other examples include the entirely on-device photo scanning, the same rotating identifier system for transcripts of Siri interactions, etc. and, of course, being the only major cloud provider to offer E2EE on everything.
Not perfect, but a huge difference from their competitors.
[1] https://www.apple.com/privacy/docs/Location_Services_White_P...
On the flip side, I tend to think a company so large would have at least one whistleblower or something on the inside, and/or would be so concerned about legal fallout that they wouldn't risk it.
On the flip side of the flip side, Apple is notoriously secretive (even among insiders) and very tight-fisted around employees sharing/leaking information. They also have some of the best lawyers in the world and a near infinite ability to fund any legal action, so may feel (and in fact, be) untouchable. And should Apple go evil, there aren't really great alternatives anyway for the average person, and they're generally so invested in the walled garden that walking away would entail a major disruption to their life.
I agree though, while not perfect, they are certainly much better than their competitors (not counting small players, e.g. GrapheneOS), and I'm grateful that at least they keep privacy at the forefront of conversation. If they abandoned it, there'd be nobody to pick up the mantle.
Apple then launched Apple maps - which at some point became quite good. Google quickly learned that they can't afford to make Android specific features in their apps or they risk losing large percentage of iOS users if Apple makes a competing product
If Apple didn't respond with making their own maps, then maybe we would see more and more Android specific features, to the point where Android would become the dominating platform
https://explodingtopics.com/blog/iphone-android-users
But this is also exactly the same game Apple plays against Android users. It's the same reason why iMessage bubbles are green for Android. Google won the maps round, but such wins are vanishingly rare against Apple.
https://support.apple.com/en-us/105087
There are non-Android devices that can send texts as well; they also appear as green. It's probably more accurate to say that encrypted messages are blue and unencrypted are green. Look at the recent AT&T hack to see why the difference matters.
In fact that's NOT the way Apple describes it, either (see the Apple article cited above), because Apple doesn't actually want to enable E2EE -- it only wants to be able to say it offers it.
In practice, ensuring that other users are pressured into choosing iMessage on iPhone is the only thing that matters to Apple.
https://www.npr.org/2024/03/28/1241443505/green-bubble-shami...
And, this very simple trick works extremely well: at least 87% of teenagers in the U.S. (https://mashable.com/article/apple-messages-green-doj) are pre-programmed to buy an iPhone, even though they have the lowest disposable income of all. Meanwhile, less than a third of the overall global population owns an iPhone.
Is that because iPhones are better? As an owner of both a recent Pro Max and Pixel Pro, I can unequivocally answer, "no", but I do find all of the annoyances between cross-device communication accrue to the point of just wanting to switch to my iPhone full-time, even though it's arguably a worse experience in many ways.
I was really just pointing out that devices like this:
https://www.hmd.com/en_us/nokia-2780-flip?sku=16WNDL11A01
and services like e.g. SMS text reminders from Internet services do no run on Android. The green is not a signifier of Android, just of non-encrypted. Or non-Apple, if you want to be less precise. (Apple devices where encryption is disabled also appear as green.)
SMS messages are green, no matter if it's sent from an Android phone or an iPhone or an authentication service or a marketing service, etc.
After the fiasco from their initial app launch, I'm sure they would have preferred not to be a meme in a sitcom if possible on this go round. It is possible to release too early
https://www.youtube.com/watch?v=tVq1wgIN62E
1. less reliant on your worst competitor
2. get to give your users something everyone else has to pay for (with money or data)
Not all of it, but it's disingenuous to say Apple doesn't make money from ads.
And in this context, that's why it is not a foolish choice to spend money on something that it's hard to sell ads on as long as it helps sell more iPhones.
How many people would still buy an iPhone without Apple or Google maps ?
It takes years, even a decade to get maps to a good quality (Apple maps launched in 2012). So I think it's a good thing that Apple started early enough. I'm sure it's crazy expensive to build and maintain. Apple can fund it from iPhone sales, and ensure that their ecosystem has an alternative for Google maps.
I don't think it's meant to turn a profit, I think it's meant as protection of their iPhone revenue.
Like, if McDonalds didn't pony up every year, they drop out of the list for Fast Food searches.
They ship their operating systems with all the "common" apps pre-installed (e.g.: Email, Calendar, Reminders, Notes, Maps, etc). For the maps to work, they need some data source. That's what Apple Maps is.
Apple doesn't make money with the Email app directly, but its existence likely improves how users perceive iOS. This probably translates to return customers and more people recommending it.
I dunno, have you actually used Apple Mail?
Is Mail.app the most powerful client on earth? No; but it is Good Enough, and I don't have to download and pay for a third party app. Is Weather.app the best weather app with all the bells and whistles possible? No; but I don't care about weather apps to download and trial fifteen other ones and It Just Works.
Maps are (orders of magnitude) more complicated; but arguably are also on the baseline level of functionality for a modern mobile OS.
And Maps.app is just so much _nicer_ to use than Google Maps. It has the same problems that all Apple products like it does (search is atrocious, POI db is bad); but it is just a much more pleasant product. It looks nicer, it _feels_ nicer, it has best-in-class transit directions, and doesn't shove ads in front of my face.
[1]: Arguably getting worse and worse at it every year; but still miles ahead of everyone else.
https://en.wikipedia.org/wiki/Apple_Maps#Initial_release
Becoming independent from Google for such a core feature was an important move.
Apple Maps forced Google to improve Google Maps on iOS.
Apple Maps data was definitely substandard when it was released, but it has improved considerably since then. I vastly prefer it to Google Maps, especially for turn by turn directions when I'm driving.
The entire reason that Apple Mapd was introduced a year early was that Google held back turn by turn directions on the iOS version of Google maps
So apparently, any URL with a query parameter (e.g. https://maps.apple.com/?ll=41.77708546284588%2C-122.51487365... ) will redirect you to ... their competitor, https://www.google.com/maps . I have no idea why.
For me, it triggered the opening of the (Apple) Maps.app desktop application. Which makes a lot more sense. On safari and chrome.
....
That's on my linux laptop, but I'm in the UK, I suspect geo and akami makes a difference~$ curl -v -A "Mozilla/5.0 (iPhone; CPU iPhone OS 17_5_1 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Mobile/15E148 Safari/604.1" "https://maps.apple.com/?ll=41.77708546284588%2C-122.51487365..."
...
With no query param it redirects to Apple sites. With query param, "Invalid protocol: maps:"
It shows 302 to maps.google.com here even using the website you give: https://i.imgur.com/y4jUdOn.png
If you have a link to maps.apple.com on a platform with app links (or claimed HTTPS URLs on android platforms), a native Apple Maps app can display the results. Otherwise, they haven't had a web version, so they would change it to a redirect to Google Maps.
I can't speak as to whether Chrome or Firefox on Mac, or alternate browsers on iOS, support this OS feature to launch native apps which have claimed a URL path.
The defacto query parameters from Google Maps (at least historically) are supported by Apple Maps as well.
This made migration to Apple Maps easy when it launched - you can just change the domain in the URL and prefer the native experience, knowing it will fall back to Google Maps in other cases.
Once Apple's web experience is finished, I imagine they will stop redirecting to Google Maps. Google has somewhat compromised the privacy posture of Google Maps, which may very well be how Apple internally justified building a web experience.
https://news.ycombinator.com/item?id=41022408
No keyboard response to Escape. It's basically the maps widget with a user unfriendly, but nice looking, drawer
Same with the Contacts app on macos, it's slow, crashes, and I doubt anyone uses it
Except in one edge case where my girlfriend and I were doing a 7 hour drive traveling east late at night on an empty highway and our eta increased by an hour then a little while later another hour, we were so confused and thought we might be driving in the wrong direction! Until we figured out we had crossed a time zone and it was also day light savings!
google maps suggests more alternative routes that may save me time but their predictions are generally less accurate. network instability seems to cause the application to “panic” and it just starts spinning around - especially when walking through downtown areas
while google has a sleeker presentation of traffic and shows the “red highlighter of misery and frustration” on my map more precisely, it’s timing is generally wildly incorrect and apple has already routed me around the problem and with more accurate time to destination estimates
So I think it’s more likely it’s just beta bugs (site not OS) and reducing their support matrix down during that time.