I am maintaining a PWA that utilizes WebSockets. In this app, after the web page is loaded from the server, JavaScript establishes a WebSocket connection. Because the app connects to a completely local server on a managed local network, we use http:// and ws:// (non-secure). The Issue: The app worked perfectly from older iOS versions up to iPadOS 18. However, during testing on iPadOS 26, we encountered an issue where the WebSocket connects successfully but is disconnected by the iPad approximately one second later. According to our packet captures, the iPad is sending a FIN packet to initiate the teardown. Additional Context: What makes this particularly confusing is that we have another server hosting the exact same PWA where the WebSocket connection stays stable even on iPadOS 26. This suggests it might not be a universal OS bug, but perhaps related to specific network conditions or a new security policy. Has anyone experienced similar behavior on iPadOS 26? Could this be a bug, or is there a new security restriction regarding non-secure WebSocket (ws://) connections in local networks? Are there any known changes in WebKit/Safari for iPadOS 26 that might cause the OS to send a FIN packet immediately after a handshake? I would appreciate any insights or suggestions on what might be causing this.

The crash is specific to iOS 26.2 prior versions working fine. WKScriptMessageHandler delegate func userContentController(_ userContentController: WKUserContentController, didReceive message: WKScriptMessage) Name attribute is accessible but WKScriptMessage body attribute causes crash The object seems to be not accessible(not in memory) self.webkit.configuration.userContentController.add(self, name: "sampleHandler") self.webkit.load(request) func userContentController(_ userContentController: WKUserContentController, didReceive message: WKScriptMessage) { print(message.name) // works print(message.body) // crashes }

The crash is specific to iOS 26.2 WKScriptMessageHandler delegate func userContentController(_ userContentController: WKUserContentController, didReceive message: WKScriptMessage) Name attribute is accessible but WKScriptMessage body attribute causes crash The object seems to be not accessible

Since recently, when I try to close a thread I created on an answer of mine, I get the following error: Trying later, next day, does not solve it. In the list of all threads, the post is not marked as answered. But if I look via Chrome or plain Safari, I see the answer itself is marked as correct: Which is not the case when opening in Safari Technology Preview (Release 233) where the same answer is still to be accepted:

I'm in the process of supporting Apple Pay through a 2nd payment gateway for some users, so I need to support two separate Apple Pay relationships. Both require their own apple-developer-merchantid-domain-association. I don't see how I can have both files for the one domain. Is this possible? Is there a workaround other than just replacing the old file with the new one and hoping that doesn't disrupt anything. That seems to be the approach taken by others (https://developer.apple.com/forums/thread/695538) but it is too high risk for me without any confirmation from Apple that this is ok. I'd also like to avoid having to setup a 2nd domain for these customers, since the current domain is already quite embedded in their operations.

I have a simple website, that I would like to make available as a progressive webapp, so I can somewhat customize what happens when people add the website to their homescreen (using the Share button) and maybe, in the future, use PWABuilder or something similar to publish it in some stores. My website needs authentication using OIDC, which after some redirects within the PWA opens an authenticator app on the phone. When the app returns from authenticating however, it opens a new browser tab instead of my PWA, so there is no context and the login fails. I am under the impression that this used to work, but I'm not sure. I'm also not sure what the problem is, I have no idea how, when the Authenticator app opens a URL (which is what I assume is happening under the hood), iOS or Safari or whatever would know which app to open (as the redirect URL is to the OIDC Idp, not the URL of my Web app). Any hints where I could look for a solution? Thanks!

Hello, We are experiencing a behavior change with WKWebView related to upgradeKnownHostsToHTTPS. Current application, we explicitly disable automatic HTTPS upgrades: let config = WKWebViewConfiguration() config.upgradeKnownHostsToHTTPS = false Observed behavior iOS 17.5 (built with Xcode 15.3) http:// image URLs are not automatically upgraded to https://, and the behavior works as expected. iOS 18.5 / 18.6.x (built with Xcode 16.4) http:// image URLs appear to be automatically upgraded to https:// by WebKit, even when upgradeKnownHostsToHTTPS is explicitly set to false. This behavior occurs for subresource requests such as <img src="http://..."> inside a WKWebView. Question Has the behavior of upgradeKnownHostsToHTTPS changed in iOS 18 / Xcode 16? Is this property now ignored for certain types of subresource requests (e.g. images), or overridden by new WebKit security policies such as mixed-content HTTPS upgrades? Any clarification or official guidance would be greatly appreciated!.

Problem Safari requires tabindex="0" for keyboard access to scrollable containers. Chrome (v130+) and Firefox (v4+) handle this automatically. Current Behavior Chrome/Firefox: Scrollable div with overflow: auto → automatically keyboard-accessible (Tab to focus, Arrow keys to scroll) Safari: Same element → NOT keyboard-accessible unless: Add tabindex="0", OR Container has focusable children Workaround <div style="overflow-y: auto; height: 300px;" tabindex="0"> <!-- content --> </div> Issue: Adds unnecessary tab stops on Chrome/Firefox where not needed. Request Will Safari support auto-focus for scrollable containers? (matching Chrome/Firefox) If not planned: Any official Apple guide for cross-browser scrollable accessibility? Timeline? If on roadmap, estimated Safari version? Can I subscribe for updates? Use Cases Dropdown menus Modal dialogs Tab panels Data tables Chat interfaces Reference: WCAG 2.1 Keyboard Accessible: https://www.w3.org/WAI/WCAG21/Understanding/keyboard.html Example component: https://www.radix-ui.com/themes/docs/components/scroll-area

Hi, after upgrading MacOS (MB Air M1 Version 26.2 (25C56)) & Safari (Version 26.2 (21623.1.14.11.9)) to latest versions, we are experiencing a new bug occurring with our web extension (Click & Read) on local storage writing and getting this error : Invalid call to browser.storage.local.set(). Disk I/O error. This doesn't happen on other browsers (chromium, Firefox). export const setLocalStorage = async (value: object) => { try { await browser.storage.local.set(value); } catch (error) { console.error("[Click & Read] Error setting local storage", error); } };

WKWebView has a new property "isBlockedByScreenTime" since iOS 26. But I do not yet understand when exactly this property could be used. When I setup content-based restrictions in the ScreenTime settings then WKWebView reports an error 105 via "webView:didFailProvisionalNavigation:" delegate. The isBlockedByScreenTime property still returns false in this case. If ScreenTime has a time-based limit, the App would not run at all. Under which circumstances would the property "isBlockedByScreenTime" return the value true? When exactly and for what can this property be actually used? The "problem" is that I want to find if a web page is blocked and can not be loaded, why this is the case. By simply trial and error I found out that WKWebView returns error codes 104 and 105 for blocked web sites because of content filters and Screen Time restrictions, however these error codes are not documented at all (at least I've not found any documentation or documentation for these error codes and also some other codes like 100, 102, 204 etc), so I'm not really sure if I handle all cases correctly. I hoped that isBlockedByScreenTime would at least tell me one reason for blocked pages. If there are documents which explain these error codes (100 and above), where I can find these?

Starting in iOS 26 (tested on 26.1), when I use any of the “policy” methods of WKNavigationDelegate to return an action policy of cancel I get a trace like this printed to console: 1 0x18de71bbc WebKit::WebFramePolicyListenerProxy::ignore(WebKit::WasNavigationIntercepted) 2 0x18db3dd50 WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction(WebKit::WebPageProxy&, WTF::Ref<API::NavigationAction, WTF::RawPtrTraits<API::NavigationAction>, WTF::DefaultRefDerefTraits<API::NavigationAction>>&&, WTF::Ref<WebKit::WebFramePolicyListenerProxy, WTF::RawPtrTraits<WebKit::WebFramePolicyListenerProxy>, WTF::DefaultRefDerefTraits<WebKit::WebFramePolicyListenerProxy>>&&)::$_0::operator()(WKNavigationActionPolicy, WKWebpagePreferences*) 3 0x100189e5c $sSo24WKNavigationActionPolicyVIeyBhy_ABIeghy_TR 4 0x100189d38 $s16WebkitPolicyTrap14ViewControllerC03webD0_06decideB3For15decisionHandlerySo05WKWebD0C_So18WKNavigationActionCySo0lmB0VctF 5 0x100189df4 $s16WebkitPolicyTrap14ViewControllerC03webD0_06decideB3For15decisionHandlerySo05WKWebD0C_So18WKNavigationActionCySo0lmB0VctFTo 6 0x18db255c0 WebKit::NavigationState::NavigationClient::decidePolicyForNavigationAction(WebKit::WebPageProxy&, WTF::Ref<API::NavigationAction, WTF::RawPtrTraits<API::NavigationAction>, WTF::DefaultRefDerefTraits<API::NavigationAction>>&&, WTF::Ref<WebKit::WebFramePolicyListenerProxy, WTF::RawPtrTraits<WebKit::WebFramePolicyListenerProxy>, WTF::DefaultRefDerefTraits<WebKit::WebFramePolicyListenerProxy>>&&) 7 0x18dea9848 WebKit::WebPageProxy::decidePolicyForNavigationAction(WTF::Ref<WebKit::WebProcessProxy, WTF::RawPtrTraits<WebKit::WebProcessProxy>, WTF::DefaultRefDerefTraits<WebKit::WebProcessProxy>>&&, WebKit::WebFrameProxy&, WebKit::NavigationActionData&&, WTF::CompletionHandler<void (WebKit::PolicyDecision&&)>&&) 8 0x18dea7a34 WebKit::WebPageProxy::decidePolicyForNavigationActionAsync(IPC::Connection&, WebKit::NavigationActionData&&, WTF::CompletionHandler<void (WebKit::PolicyDecision&&)>&&) 9 0x18d9cbbf4 void IPC::handleMessageAsync<Messages::WebPageProxy::DecidePolicyForNavigationActionAsync, IPC::Connection, WebKit::WebPageProxy, WebKit::WebPageProxy, void (IPC::Connection&, WebKit::NavigationActionData&&, WTF::CompletionHandler<void (WebKit::PolicyDecision&&)>&&)>(IPC::Connection&, IPC::Decoder&, WebKit::WebPageProxy*, void (WebKit::WebPageProxy::*)(IPC::Connection&, WebKit::NavigationActionData&&, WTF::CompletionHandler<void (WebKit::PolicyDecision&&)>&&)) 10 0x18d9c7728 WebKit::WebPageProxy::didReceiveMessage(IPC::Connection&, IPC::Decoder&) 11 0x18e49a0d8 IPC::MessageReceiverMap::dispatchMessage(IPC::Connection&, IPC::Decoder&) 12 0x18df1908c WebKit::WebProcessProxy::dispatchMessage(IPC::Connection&, IPC::Decoder&) 13 0x18d9dfc28 WebKit::WebProcessProxy::didReceiveMessage(IPC::Connection&, IPC::Decoder&) 14 0x18e47f72c IPC::Connection::dispatchMessage(WTF::UniqueRef<IPC::Decoder>) 15 0x18e47fac4 IPC::Connection::dispatchIncomingMessages() 16 0x199ad3758 WTF::RunLoop::performWork() 17 0x199ad4eb0 WTF::RunLoop::performWork(void*) 18 0x1804563a4 __CFRUNLOOP_IS_CALLING_OUT_TO_A_SOURCE0_PERFORM_FUNCTION__ 19 0x1804562ec __CFRunLoopDoSource0 20 0x180455a78 __CFRunLoopDoSources0 21 0x180454c4c __CFRunLoopRun 22 0x18044fcec _CFRunLoopRunSpecificWithOptions 23 0x1926be9bc GSEventRunModal 24 0x18630f0d8 -[UIApplication _run] 25 0x186313300 UIApplicationMain 26 0x18554ac38 block_destroy_helper.15 27 0x10018a70c $sSo21UIApplicationDelegateP5UIKitE4mainyyFZ 28 0x10018a67c $s16WebkitPolicyTrap11AppDelegateC5$mainyyFZ 29 0x10018a818 __debug_main_executable_dylib_entry_point 30 0x1000cd3d0 29 dyld 0x00000001000cd3d0 start_sim + 20 31 0x1002bab98 30 ??? 0x00000001002bab98 0x0 + 4297829272 This doesn’t happen in 18.6. Also, it doesn’t seem to have any negative consequences other than the console spam? But then, the navigation is being cancelled anyway, so maybe it’s trapping and just happens to have the effect of not loading the request? Anyway, I guess I can’t upload zips. But it’s pretty easy to reproduce. Just assign a WKWebView a navigationDelegate with an implementation like: func webView(_ webView: WKWebView, decidePolicyFor navigationAction: WKNavigationAction, decisionHandler: @escaping @MainActor (WKNavigationActionPolicy) -> Void) { decisionHandler(.cancel) } and then have it .load() anything. Have I been doing this wrong and 26 exposes it? Or is this a bug in 26? If the latter, any downstream consequences I should be looking out for? macOS 15.7.1 (24G231) Xcode 26.1.1 (17B100) iOS 26.1 (23B86)

If the extension uses manifest v3 and a background script in the form of a service worker, then in Safari it is not possible to open the background script debugging window. If I expand the Developer menu in Safari, there is nothing under Web Extension Background Data (or disappear after click), which is an error. In other browsers (Edge, Chrome, Opera, Firefox) this works correctly. If I switch the background script back to non-persistent script mode, everything works fine and from the Developer menu and the Web Extension Background Data submenu I am able to open the background script debugging window for the extension. Am I doing something wrong?

My iOS app uses a WKWebView with a WKUIDelegate method (webView:createWebViewWithConfiguration:forNavigationAction:windowFeatures:) to handle popup windows. This works for most cases, but during OAuth flows on certain sites (e.g., canva.com), the popup WKWebView attempts to send results back to the main WKWebView using JavaScript like window.opener.postMessage(...). However, window.opener is always null in the popup, preventing the message from being posted and blocking login completion.I've researched this and found suggestions that it's by design, as WKWebView instances are isolated for security reasons. Has anyone encountered this and found a reliable workaround (e.g., bridging communication between the main and popup WKWebViews without relying on window.opener)?

Hello WebKit Team, I’m writing to ask if iOS provides a native way to intercept AJAX (XMLHttpRequest or fetch) calls inside WKWebView. On Android, this is handled via: shouldInterceptRequest(WebView view, WebResourceRequest request) but iOS currently seems to have no equivalent. We’ve tried: WKURLSchemeHandler → works only for custom schemes URLProtocol with WKProcessPool → unreliable for AJAX in WebView JavaScript injection → partial and unofficial Could you please clarify: Is there a recommended native approach to intercept AJAX requests? If not supported, is it planned for future releases? Any official workaround or guidance? This is critical for debugging, analytics, and compliance in hybrid apps.

webkitEnterFullScreen API is supported on iOS for video element, but not for a div element. Also as a fullscreen demo website shown, Safari on macOS supports div element but not on iOS. Is there any plan to add the support in iOS? If not is there any way to fullscreen a div element or make it run as fullscreen on Safari iOS?

在 iOS 平台使用 WKWebView 通过file://协议加载本地 HTML 文件时，存储在localStorage中的数据会在 App 后台切换、进程重启后偶尔丢失；但相同代码在安卓 / 鸿蒙平台无此问题。 现在的文档 仅明确了「默认数据存储（defaultDataStore）可将网站数据持久化到磁盘，非持久化存储（nonPersistent）仅存内存」的基础规则； 未提及「file://协议内容即使使用默认持久化存储，也会被归为临时内存存储」这一关键场景限制； 仅在WKURLSchemeHandler关联说明中隐含「自定义 URL 协议可处理 WebKit 原生不支持的 URL 方案」，但未直接关联file://的存储问题。 我找不到如何处理这个问题的官方文档，仅仅有其他的博客说需要增加http/https加载就没有这个问题。 请提供给我官方文档或者官方回复 关于出现这种file:/加载html出现问题的处理办法

Hello, can't find a better Forum and after all this is Apple Support here? Seems starting with iOS 26.2 RC2 our Newsletter is no longer rendered 'responsive' so with unuseable small font. Anybody any suggestion? Or maybe @apple fix the bug? Where would I even start to debug this? I don't have any iphone at hand so i only got screenshots from users. Our where can i contact the actual apple support? correct: invalid:

Hi there I've been having trouble finding any details around how safari is supposed to behave when a FairPlay license expires. My assumption was that the video segments would stop getting decrypted and playback would stop, however I just see that the playback continues like nothing has happened. I've setup the "fps_safari_has_key_renewal.html" sample code from the Fairplay SDK and got encrypted playback working. The renewal method also appears to work. However, if I don't issue a renew call, or if I wait several minutes after the renew has succeeded the video never stops (my license is set with a 1 minute expiry so I can test this quickly). I've also observed that the MediaKeySession expiration property is always set to NaN even though my license has an expiry. I've tried with both Lease and Rental expiries set in the license (separately AND at the same time in separate tests). I'm using EZDRM as my drm provider. Just looking for some feedback on if this is supposed to work this way in safari or if license expiry isn't supported in safari. Thanks!

Subject: Unexpected system confirmation dialog when opening a Universal Link Description of the issue: We’re implementing a login flow using Native iOS apps, Universal Links, and OpenID Connect authentication. Our domain is correctly configured with the apple-app-site-association file, and Universal Links work as expected. However, under certain circumstances, the behavior differs on the same login page: The user connects to the OIDC provider in their mobile browser. Instead of automatically switching to the Native application, a popup asks the user to open the app. The behavior depends on how the login page is opened: When opened in a new browser tab, the Universal Link opens the app immediately without showing any system confirmation dialog. When the same tab is reused without a page refresh, iOS displays a system confirmation dialog asking the user to open the link in the app. This confirmation dialog doesn’t appear in the first scenario. This additional system dialog impacts the user experience, as we aim for a frictionless and seamless login flow with minimal confirmation steps. Question: Why does the app switch work automatically only the first time? Is this intentional iOS behavior? If so, what security requirement or system rule causes a confirmation dialog when reusing the same tab, but not when redirecting in a new tab or pressing a button on the reused tab? Is this expected design, or should we adjust our side to avoid this dialog? Environment details: Verified on two devices iOS versions: 18.7.2, 26.1 Thanks for your help.

I'm building a macOS extension that needs to track multi-step navigation chains (A → B → C) to adjust behavior based on where users came from. Current approach: Using webNavigation.onBeforeNavigate to detect intermediate steps, but experiencing issues in Safari that don't occur on Chrome/Firefox/Edge. Questions: Is webNavigation the right API for tracking redirect chains in Safari? Does ITP/Private Browsing affect event delivery? Any alternative approaches recommended? (Safari version 26.0.1) Any guidance appreciated!