7 ms·
I suspect it might have been motivated by antitrust concerns, but safari is really not that bad. Check out Interop 2025: https://wpt.fyi/interop-2025 https://wp
by ChadNauseam 10mo ago
I suspect it might have been motivated by antitrust concerns, but safari is really not that bad. Check out Interop 2025: https://wpt.fyi/interop-2025 https://wpt.fyi/interop-2025
They generally are pretty caught up on features. They have webgpu, they support the web notifications API (once a PWA is installed), lots of stuff. My main gripe is that they make it too hard to install PWAs, but we're still waiting for an actual API for that. (Maybe in 2027? [0])
> And Apple purposely will never implement lots of APIs that only their native apps allow (which other browsers implement)
Can you give an example?
[0]: https://blogs.windows.com/msedgedev/2025/11/24/the-web-install-api-is-ready-for-testing/ https://blogs.windows.com/msedgedev/2025/11/24/the-web-insta...
- kelthuzad 10mo ago>> And Apple purposely will never implement lots of APIs that only their native apps allow (which other browsers implement) >Can you give an example? Web Bluetooth, Web USB, Web NFC, Web Serial... Of course Apple will uphold its usual charade to claim that it's about pRiVacy & sEcuRiTy to maintain plausible deniability. They could easily implement it and keep it disabled by default, such that users could make the conscious choice to enable it or keep it disabled. Any adequate analysis of Apple's behavior and motivations must mention Apple's conflict of interest, because Apple will be biased against technology that could diminish the value proposition of "native" apps which Apple has been taxing so unchallenged for all these years.
- troupo 10mo ago> Web Bluetooth, Web USB, Web NFC, Web Serial Chrome-only non-standards. Note that Firefox is against these, too. > Any adequate analysis of Apple's behavior and motivations must mention Apple's conflict of interest I've yet to see an adequate analysis that doesn't pretend that anything Chrome shits, sorry, ships is immediately a standard that must absolutely be implemented by everyone immediately.
- kelthuzad 10mo agoYou're right that Firefox also opposes some of these specific implementations in its current form, and that Google often rushes features. However, that doesn't diminish Apple's conflict of interest at all, so sometimes their arguments happen to align with reality just as a broken clock is correct twice a day. Apple applies many double standards e.g. they allow native apps to access these hardware features (where they happen to collect a 30% tax) but block the Web from doing the same (where they collect 0%). If privacy was the only concern, they would work on a safe standard, but instead they block the capability entirely to ensure that any of the App Store's rivals remain constrained and thus inferior such that the App Store's revenue isn't threatened.
- troupo 10mo ago> You're right that Firefox also opposes some of these specific implementations in its current form, and that Google often rushes features. However, that doesn't diminish Apple's conflict of interest at all Funny how you agree that Firefox opposes these non-standards, and how Google rushes things. And immediately turn around and basically say "no-no-no, Apple is to blame and Safari (and, by extension Firefox) must absolutely implement these non-standard features from Chrome". The rest of demagoguery is irrelevant. BTW literally the moment Firefox relented and implemented WebMIDI they had originally opposed, they immediately ran into tracking/fingerprinting attempts using WebMIDI that Chrome just couldn't care less about.
- kelthuzad 10mo ago>Funny how you agree that Firefox opposes these non-standards, and how Google rushes things. And immediately turn around and basically say "no-no-no, Apple is to blame and Safari (and, by extension Firefox) must absolutely implement these non-standard features from Chrome". There is nothing "funny" about me acknowledging facts, that's what a reasonable person should always do, try it. What's not funny though, is how you're butchering and misrepresenting my arguments to such a gross degree. I've never stated that everybody "must implement these non-standard features from Chrome", instead I've made a much more nuanced argument about how Apple's conflict of interest is motivating them to reject entire feature sets for competing technology instead of helping to implement a safe standard, which is indicative of their bad faith motivations. That anti-competitive strategy has been essential for Apple in collecting billions in app taxes by systematically hobbling any competition before it can emerge. >BTW literally the moment Firefox relented and implemented WebMIDI they had originally opposed, they immediately ran into tracking/fingerprinting attempts using WebMIDI that Chrome just couldn't care less about. So? Just as native apps give users certain freedoms that can have problematic aspects, web apps should have _equal rights_ and be able to play on a level playing field. The choice and freedom should be the users' and not that of Apple's finance division. None of this gives Apple the right to uphold its anti-competitive strategy with its corporate double speak. And the fact that you're so hyperfocused on specifics while failing to grasp the broader argument, so you can cheerlead for Apple's anti-competitive behavior, is revealing a clear bias.
- jauntywundrkind 10mo agoPersonally my feel is Safari at least isn't dead in the water any more, does ship some stuff. It's much better than 2 years ago. 4 years ago it was a travesty. But there's still all sorts of wonkiness they just makes Safari non viable. If you don't PWA install, your storage gets cleared alarmingly quickly. If you do install it's still cleared wicked fast. Notifications seem to have incredibly unreliable delivery issues and require PWA installs to work at all. The features are closer to parity than before but the base functionality is still sabotaged deeply. 'The user is secure' with Apple is amazing doublespeak (the second meaning being securely in Apple's pocket with no where to go). It's worth noting that Interop participants meet and decide via unanimous consent what they are going to work on each year. The anti-trust case against Apple would be far stronger if they didn't show up & find some stuff to work on, to agree to. And with apologies as I break out the tin foil hat, showing up also gives them some leverage to shape what doesn't get worked on too.
- leptons 10mo agoSafari is the worst browser by far, especially on iOS. Apple also does things their own way, ignoring standards, so that I have to have a real actual iPhone to debug their platform-specific problems, especially around touch interactions. >Can you give an example? Web Bluetooth API, and lots of others. My product could use bluetooth but we're forced to work around Apple's Safari limitations and use Wifi instead, which drains the battery faster. We do not want to write a specific app for iOS (which costs us money to build and maintain), which then allows Apple to extort us for a percentage of sales through the app. Bluetooth would be the better option, but Wifi works although is a bit more cumbersome to deal with. So sorry Apple fans, you have to use wifi with our product because Apple reasons. I am going to open a bottle of champagne when the DOJ finally forces Apple to allow other browsers on iOS. https://www.justice.gov/archives/opa/media/1344546/dl?inline https://www.justice.gov/archives/opa/media/1344546/dl?inline
- concinds 10mo agoNo, it's just not true. Interop 2025 is a subset of web features, but Apple gets a veto on which features get included in each Interop round, and vetoes heavily. It doesn't reflect interoperability in general. Safari also consistently starts out the worst each year, and improves the slowest. They don't support notifications correctly, they have a semi-broken implementation. Only a subset of sites will work, even though they'll work perfectly on Chrome or Firefox or even minor browsers. Even if you put the site on the homescreen.