You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm noticing this in Chrome 154.0.8037.98 on MacOS.
Doing a hard reload:
Console:
?php=8.4&wp=latest&networking=yes&language=&multisite=no:1 <meta name="apple-mobile-web-app-capable" content="yes"> is deprecated. Please include <meta name="mobile-web-app-capable" content="yes">
wordpress-DLEoGjgB.js:1 Failed to update service worker. TypeError: Failed to update a ServiceWorker for scope ('https://playground.wordpress.net/') with script ('https://playground.wordpress.net/sw.js'): A bad HTTP response code (403) was received when fetching the script.
Ke @ wordpress-DLEoGjgB.js:1
playground-worker-endpoint-blueprints-ysD0hoY3.js:5268 Uncaught (in promise) TypeError: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.js
opfs-site-storage-wmmpyS_c.js:2 TypeError: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.jsCaused by: Error: Comlink method call failed
at Object.deserialize (wordpress-DLEoGjgB.js:14:13356)
at L (wordpress-DLEoGjgB.js:14:3884)
at async Object.boot (wordpress-DLEoGjgB.js:14:33970)Caused by: Error: Comlink method call failed
at Object.deserialize (opfs-site-storage-wmmpyS_c.js:15:13486)
at gr (opfs-site-storage-wmmpyS_c.js:15:3958)
at async SA.bootPlayground (blueprint-editor-B3tSK9LY.js:2:2887)
at async pq (blueprint-editor-B3tSK9LY.js:2:7018)
at async main-CI1utP6L.js:58:4919
qu @ opfs-site-storage-wmmpyS_c.js:2
opfs-site-storage-wmmpyS_c.js:2 Error creating saved site. Falling back to a temporary site. Error: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.js
at Object.effect (blueprint-editor-B3tSK9LY.js:80:27629)
at a (blueprint-editor-B3tSK9LY.js:74:31573)
at blueprint-editor-B3tSK9LY.js:74:32484
at blueprint-editor-B3tSK9LY.js:74:36608
at blueprint-editor-B3tSK9LY.js:74:17826
at dispatch (blueprint-editor-B3tSK9LY.js:74:6557)
at main-CI1utP6L.js:58:5827
qu @ opfs-site-storage-wmmpyS_c.js:2
wordpress-DLEoGjgB.js:1 Failed to update service worker. TypeError: Failed to update a ServiceWorker for scope ('https://playground.wordpress.net/') with script ('https://playground.wordpress.net/sw.js'): A bad HTTP response code (403) was received when fetching the script.
Ke @ wordpress-DLEoGjgB.js:1
playground-worker-endpoint-blueprints-ysD0hoY3.js:5268 Uncaught (in promise) TypeError: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.js
opfs-site-storage-wmmpyS_c.js:2 TypeError: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.jsCaused by: Error: Comlink method call failed
at Object.deserialize (wordpress-DLEoGjgB.js:14:13356)
at L (wordpress-DLEoGjgB.js:14:3884)
at async Object.boot (wordpress-DLEoGjgB.js:14:33970)Caused by: Error: Comlink method call failed
at Object.deserialize (opfs-site-storage-wmmpyS_c.js:15:13486)
at gr (opfs-site-storage-wmmpyS_c.js:15:3958)
at async SA.bootPlayground (blueprint-editor-B3tSK9LY.js:2:2887)
at async pq (blueprint-editor-B3tSK9LY.js:2:7018)
at async main-CI1utP6L.js:58:4919
qu @ opfs-site-storage-wmmpyS_c.js:2
opfs-site-storage-wmmpyS_c.js:2 Error: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.js
at Object.effect (blueprint-editor-B3tSK9LY.js:80:27629)
at a (blueprint-editor-B3tSK9LY.js:74:31573)
at blueprint-editor-B3tSK9LY.js:74:32484
at blueprint-editor-B3tSK9LY.js:74:36608
at blueprint-editor-B3tSK9LY.js:74:17826
at dispatch (blueprint-editor-B3tSK9LY.js:74:6557)
at main-CI1utP6L.js:58:5827
qu @ opfs-site-storage-wmmpyS_c.js:2
blueprint-editor-B3tSK9LY.js:80 Uncaught (in promise) Error: Failed to fetch dynamically imported module: https://playground.wordpress.net/assets/php_8_4-BewWylNE.js
at Object.effect (blueprint-editor-B3tSK9LY.js:80:27629)
at a (blueprint-editor-B3tSK9LY.js:74:31573)
at blueprint-editor-B3tSK9LY.js:74:32484
at blueprint-editor-B3tSK9LY.js:74:36608
at blueprint-editor-B3tSK9LY.js:74:17826
at dispatch (blueprint-editor-B3tSK9LY.js:74:6557)
at main-CI1utP6L.js:58:5827
Then doing a soft reload:
Console:
?php=8.4&wp=latest&networking=yes&language=&multisite=no:1 <meta name="apple-mobile-web-app-capable" content="yes"> is deprecated. Please include <meta name="mobile-web-app-capable" content="yes">
?php=8.4&wp=latest&networking=yes&language=&multisite=no:1 <meta name="apple-mobile-web-app-capable" content="yes"> is deprecated. Please include <meta name="mobile-web-app-capable" content="yes">
opfs-site-storage-wmmpyS_c.js:2 Unable to clean up autosave site-kind-quiet-garden: NotFoundError: A requested file or directory could not be found at the time an operation was processed.
qu @ opfs-site-storage-wmmpyS_c.js:2
logMessage @ opfs-site-storage-wmmpyS_c.js:5
warn @ opfs-site-storage-wmmpyS_c.js:5
(anonymous) @ opfs-site-storage-wmmpyS_c.js:2561
Promise.catch
list @ opfs-site-storage-wmmpyS_c.js:2561
await in list
(anonymous) @ main-CI1utP6L.js:108
Mf @ blueprint-editor-B3tSK9LY.js:43
xs @ blueprint-editor-B3tSK9LY.js:43
vw @ blueprint-editor-B3tSK9LY.js:43
Uo @ blueprint-editor-B3tSK9LY.js:41
qD @ blueprint-editor-B3tSK9LY.js:43
ni @ blueprint-editor-B3tSK9LY.js:43
NP @ blueprint-editor-B3tSK9LY.js:43
b @ blueprint-editor-B3tSK9LY.js:28
D @ blueprint-editor-B3tSK9LY.js:28
postMessage
$ @ blueprint-editor-B3tSK9LY.js:28
I @ blueprint-editor-B3tSK9LY.js:28
e.unstable_scheduleCallback @ blueprint-editor-B3tSK9LY.js:28
zP @ blueprint-editor-B3tSK9LY.js:43
pn @ blueprint-editor-B3tSK9LY.js:43
or @ blueprint-editor-B3tSK9LY.js:43
jf @ blueprint-editor-B3tSK9LY.js:43
If.render.fy.render @ blueprint-editor-B3tSK9LY.js:43
(anonymous) @ main-CI1utP6L.js:108
opfs-site-storage-wmmpyS_c.js:2 Error reading site site-kind-quiet-garden: NotFoundError: A requested file or directory could not be found at the time an operation was processed.
qu @ opfs-site-storage-wmmpyS_c.js:2
logMessage @ opfs-site-storage-wmmpyS_c.js:5
error @ opfs-site-storage-wmmpyS_c.js:5
list @ opfs-site-storage-wmmpyS_c.js:2561
await in list
(anonymous) @ main-CI1utP6L.js:108
Mf @ blueprint-editor-B3tSK9LY.js:43
xs @ blueprint-editor-B3tSK9LY.js:43
vw @ blueprint-editor-B3tSK9LY.js:43
Uo @ blueprint-editor-B3tSK9LY.js:41
qD @ blueprint-editor-B3tSK9LY.js:43
ni @ blueprint-editor-B3tSK9LY.js:43
NP @ blueprint-editor-B3tSK9LY.js:43
b @ blueprint-editor-B3tSK9LY.js:28
D @ blueprint-editor-B3tSK9LY.js:28
postMessage
$ @ blueprint-editor-B3tSK9LY.js:28
I @ blueprint-editor-B3tSK9LY.js:28
e.unstable_scheduleCallback @ blueprint-editor-B3tSK9LY.js:28
zP @ blueprint-editor-B3tSK9LY.js:43
pn @ blueprint-editor-B3tSK9LY.js:43
or @ blueprint-editor-B3tSK9LY.js:43
jf @ blueprint-editor-B3tSK9LY.js:43
If.render.fy.render @ blueprint-editor-B3tSK9LY.js:43
(anonymous) @ main-CI1utP6L.js:108
opfs-site-storage-wmmpyS_c.js:2 Error reading site site-kind-quiet-garden: NotFoundError: A requested file or directory could not be found at the time an operation was processed.
qu @ opfs-site-storage-wmmpyS_c.js:2
logMessage @ opfs-site-storage-wmmpyS_c.js:5
error @ opfs-site-storage-wmmpyS_c.js:5
listMetadata @ opfs-site-storage-wmmpyS_c.js:2561
await in listMetadata
(anonymous) @ blueprint-editor-B3tSK9LY.js:80
await in (anonymous)
(anonymous) @ blueprint-editor-B3tSK9LY.js:74
a @ blueprint-editor-B3tSK9LY.js:80
createNewSavedSite @ blueprint-editor-B3tSK9LY.js:80
L @ main-CI1utP6L.js:108
(anonymous) @ main-CI1utP6L.js:108
Mf @ blueprint-editor-B3tSK9LY.js:43
xs @ blueprint-editor-B3tSK9LY.js:43
qD @ blueprint-editor-B3tSK9LY.js:43
ni @ blueprint-editor-B3tSK9LY.js:43
vw @ blueprint-editor-B3tSK9LY.js:43
Uo @ blueprint-editor-B3tSK9LY.js:41
(anonymous) @ blueprint-editor-B3tSK9LY.js:43
opfs-site-storage-wmmpyS_c.js:2 Error creating saved site. Falling back to a temporary site. Error: The browser storage worker failed while writing /sites/site-determined-noisy-mountain/wp-runtime.json at /assets/opfs-site-storage-worker-for-safari-pJM5slyx.js.
at r.onerror (opfs-site-storage-wmmpyS_c.js:2561:15289)
qu @ opfs-site-storage-wmmpyS_c.js:2
logMessage @ opfs-site-storage-wmmpyS_c.js:5
error @ opfs-site-storage-wmmpyS_c.js:5
L @ main-CI1utP6L.js:108
await in L
(anonymous) @ main-CI1utP6L.js:108
Mf @ blueprint-editor-B3tSK9LY.js:43
xs @ blueprint-editor-B3tSK9LY.js:43
qD @ blueprint-editor-B3tSK9LY.js:43
ni @ blueprint-editor-B3tSK9LY.js:43
vw @ blueprint-editor-B3tSK9LY.js:43
Uo @ blueprint-editor-B3tSK9LY.js:41
(anonymous) @ blueprint-editor-B3tSK9LY.js:43
wordpress-DLEoGjgB.js:1 Failed to update service worker. TypeError: Failed to update a ServiceWorker for scope ('https://playground.wordpress.net/') with script ('https://playground.wordpress.net/sw.js'): A bad HTTP response code (403) was received when fetching the script.
Ke @ wordpress-DLEoGjgB.js:1
logMessage @ wordpress-DLEoGjgB.js:4
error @ wordpress-DLEoGjgB.js:4
Zt @ wordpress-DLEoGjgB.js:14
await in Zt
(anonymous) @ wordpress-DLEoGjgB.js:15
wordpress-DLEoGjgB.js:15 Error: WebWorker failed to load at https://playground.wordpress.net/playground-worker-endpoint-blueprints-ysD0hoY3.js.
at t.onerror (wordpress-DLEoGjgB.js:14:25963)
(anonymous) @ wordpress-DLEoGjgB.js:15
I have the same issue and cannot access playground, neither via custom link, nor simply by accessing playground.wordpress.net without any custom configuration. Confirmed by a peer from a different machine in a different office.
Unfortunately we're unable to reproduce this so far. We had one occurrence ourself but nothing reliably reproducible so far. Could any of you try to go to DevTools → Application → Service Workers → Update and tell us if it makes a difference? Thanks!
It might also be useful to get the HTTP headers of those 403 requests. For example through a HAR file (right click in the network tab on a request: copy > copy all as HAR).
Thanks, @akirk! This resolved the issue for me in Chrome and other Chromium-based browsers. After clicking “Update” under DevTools → Application → Service Workers and reloading the page twice, Playground loaded correctly.
I’m attaching a screenshot to help make the steps easier to follow.
We are working on changes in #4557 that should reduce the possibility of this issue. After those are applied, if we still see the issue, we will have to follow up with our hosting provider's system folks.
Hi @FTWAgency, @srtfisher, a fix was deployed on the hosting side for the browser-check responses on service worker requests. Could you try again, including the embedded case, and tell us if it loads now, with your browser version? If it still fails, please post a screenshot of the console error or the network error. 🙏
Reproduction details from a Windows desktop Chrome, in case they help narrow the hosting-side fix mentioned above. The failure still reproduced on 2026-11-18 / 11-19 (after the hosting-side change was announced).
Environment
Windows 11 (build 26200), Google Chrome 158.0.8127.37. No enterprise policies, no proxy/VPN, no chrome://flags overrides.
Reproduces in the normal profile, Incognito, Guest, and a brand-new disposable headed Chrome user-data directory (no extensions). Another (non-Chrome) browser and an Electron-based Chromium 156 (different user agent) on the same machine load Playground fine.
Reproduce
Open https://playground.wordpress.net/ in headed Chrome 158 with a default user agent and a fresh profile.
Playground shows "Ooops! WordPress Playground had a hiccup!". Console: Failed to register a ServiceWorker for scope ('https://playground.wordpress.net/') with script ('https://playground.wordpress.net/sw.js'): ServiceWorker cannot be started.
What is failing (DevTools Protocol ServiceWorker.workerErrorReported and a Chrome net-log)
GET /sw.js returns 200. Currently it is a 55-byte stub: import "/releases/<build>/sw.js";.
The service worker then fetches that import. This request is Sec-Fetch-Mode: cors, Sec-Fetch-Dest: serviceworker, Origin set, Referer: https://playground.wordpress.net/sw.js, and sends no cookies. It receives HTTP 403, content-type: text/html, set-cookie: _hcc and _hcr (names only), server-timing: a8c-cdn, ...cache;desc=BYPASS. That matches the "Checking your browser..." / HASHCASH_CONFIG page quoted earlier in this thread.
Chrome reports it as: "A bad HTTP response code (403) was received when fetching the script." The page-level error is only the generic "ServiceWorker cannot be started".
A service worker has no way to run the challenge JavaScript, so a challenge page on this request cannot be solved.
What passes / fails
Fails: registering /sw.js (module type) in headed Chrome 158 (3 of 3 fresh profiles).
Passes: registering the versioned worker /releases/<build>/sw.js directly. It is one self-contained file with no import, so it never makes the second request.
Passes: the same layout (root stub importing a versioned worker) served from a local test server in the same Chrome profile, so Chrome's module service worker support is fine.
Passes: headless Chrome 158 (HeadlessChrome/158 user agent), and headed Chrome 158 with a HeadlessChrome user-agent override. So the response appears to depend on request characteristics such as the user agent. I have not identified the exact CDN rule, and plain curl with the same headers got 200, so it is not a simple header match.
Interpretation (hypothesis)
This looks like an interaction between the CDN browser check and the service worker's import request, not something in a WordPress plugin: it happens on the plain Playground URL before any blueprint or plugin loads. I am not claiming a browser defect.
Possible fix
Make sure requests that load service worker scripts and their imports (Sec-Fetch-Dest: serviceworker, or the /sw.js and /releases/*/sw.js paths) are exempt from the HTML browser challenge and always get the JavaScript, while keeping the check on other routes. Alternatively, serve a root /sw.js that does not import another script. Happy to retest or share more sanitized details.
Would you be up for using the browser dev tools to capture and share a HAR file of your error? You could DM via WordPress.org Slack if you are up for sharing but prefer to do so privately.
@ajitbohra, thank you for the HAR file and for the update that Playground is loading for you now.
I'm very sorry for the trouble. It appears to have been due to some hosting defense code that conflicted with the special way browsers request Service Worker resources.
Activity
fellyph commentedon Oct 12, 2026
Hi @uzariimalik12-ship-it,
Can you share the following information:
westonruter commentedon Oct 13, 2026
I'm noticing this in Chrome 154.0.8037.98 on MacOS.
Doing a hard reload:
Console:
Then doing a soft reload:
Console:
westonruter commentedon Oct 13, 2026
Interestingly, I have no errors in Firefox 157.0.1 (aarch64).
hiroshisatoy commentedon Oct 13, 2026
I'm experiencing the same issue in my environment, so I'm sharing my system and browser details
macOS 27.0.1
Since it works in Safari and Firefox, the issue may be related to Chrome (Chromium)-specific features or behavior.
apermo commentedon Oct 13, 2026
I have the same issue and cannot access playground, neither via custom link, nor simply by accessing playground.wordpress.net without any custom configuration. Confirmed by a peer from a different machine in a different office.
michakrapp commentedon Oct 13, 2026
Tested on several Browsers, seem to fail only for Chromium
System: macOS 26.6.2
akirk commentedon Oct 13, 2026
Unfortunately we're unable to reproduce this so far. We had one occurrence ourself but nothing reliably reproducible so far. Could any of you try to go to DevTools → Application → Service Workers → Update and tell us if it makes a difference? Thanks!
It might also be useful to get the HTTP headers of those 403 requests. For example through a HAR file (right click in the network tab on a request: copy > copy all as HAR).
hiroshisatoy commentedon Oct 13, 2026
Thanks, @akirk! This resolved the issue for me in Chrome and other Chromium-based browsers. After clicking “Update” under DevTools → Application → Service Workers and reloading the page twice, Playground loaded correctly.
I’m attaching a screenshot to help make the steps easier to follow.
westonruter commentedon Oct 14, 2026
It seems to be working now for me in Chrome.
apermo commentedon Oct 14, 2026
@akirk that
DevTools → Application → Service Workers → Updatedid the trick for me too.And the browser is: Version 154.0.8037.98 (Offizieller Build) (arm64) on MacOS 26.6.2
michakrapp commentedon Oct 14, 2026
Also had to do a hard refresh, Mac:
Cmd + Shift + RThen it worked with the Service Workers Update
8 remaining items
srtfisher commentedon Nov 2, 2026
Happening for us on https://repl.alley.dev/ where
remote.htmlreturns a challenge required:fellyph commentedon Nov 2, 2026
@adamziel try running on a slow connection; when I was traveling in Brazil, I saw the error a few times.
brandonpayton commentedon Nov 6, 2026
We are working on changes in #4557 that should reduce the possibility of this issue. After those are applied, if we still see the issue, we will have to follow up with our hosting provider's system folks.
mho22 commentedon Nov 18, 2026
Hi @FTWAgency, @srtfisher, a fix was deployed on the hosting side for the browser-check responses on service worker requests. Could you try again, including the embedded case, and tell us if it loads now, with your browser version? If it still fails, please post a screenshot of the console error or the network error. 🙏
Qaisarkhursheed commentedon Nov 18, 2026
There is still there
Qaisarkhursheed commentedon Nov 18, 2026
I tried to updated the service worker but still again after reload same shit
erfanteymourii commentedon Nov 18, 2026
Reproduction details from a Windows desktop Chrome, in case they help narrow the hosting-side fix mentioned above. The failure still reproduced on 2026-11-18 / 11-19 (after the hosting-side change was announced).
Environment
Reproduce
https://playground.wordpress.net/in headed Chrome 158 with a default user agent and a fresh profile.Failed to register a ServiceWorker for scope ('https://playground.wordpress.net/') with script ('https://playground.wordpress.net/sw.js'): ServiceWorker cannot be started.What is failing (DevTools Protocol
ServiceWorker.workerErrorReportedand a Chrome net-log)GET /sw.jsreturns 200. Currently it is a 55-byte stub:import "/releases/<build>/sw.js";.Sec-Fetch-Mode: cors,Sec-Fetch-Dest: serviceworker,Originset,Referer: https://playground.wordpress.net/sw.js, and sends no cookies. It receives HTTP 403,content-type: text/html,set-cookie: _hccand_hcr(names only),server-timing: a8c-cdn, ...cache;desc=BYPASS. That matches the "Checking your browser..." /HASHCASH_CONFIGpage quoted earlier in this thread.What passes / fails
/sw.js(module type) in headed Chrome 158 (3 of 3 fresh profiles)./releases/<build>/sw.jsdirectly. It is one self-contained file with no import, so it never makes the second request.HeadlessChrome/158user agent), and headed Chrome 158 with aHeadlessChromeuser-agent override. So the response appears to depend on request characteristics such as the user agent. I have not identified the exact CDN rule, and plaincurlwith the same headers got 200, so it is not a simple header match.Interpretation (hypothesis)
This looks like an interaction between the CDN browser check and the service worker's import request, not something in a WordPress plugin: it happens on the plain Playground URL before any blueprint or plugin loads. I am not claiming a browser defect.
Possible fix
Make sure requests that load service worker scripts and their imports (
Sec-Fetch-Dest: serviceworker, or the/sw.jsand/releases/*/sw.jspaths) are exempt from the HTML browser challenge and always get the JavaScript, while keeping the check on other routes. Alternatively, serve a root/sw.jsthat does not import another script. Happy to retest or share more sanitized details.brandonpayton commentedon Nov 18, 2026
@Qaisarkhursheed thank you for letting us know.
@erfanteymourii, thank you as well. The details are very helpful and we are looking into it.
I'm sorry for this troublesome experience.
brandonpayton commentedon Nov 19, 2026
@Qaisarkhursheed how does your failure path compare with @erfanteymourii's?
Would you be up for using the browser dev tools to capture and share a HAR file of your error? You could DM via WordPress.org Slack if you are up for sharing but prefer to do so privately.
ajitbohra commentedon Nov 19, 2026
@brandonpayton sharing log and network HAR seeing same issue.
This is chrome on macos! tried clearing application data to start fresh.
playground.wordpress.net-1795094455750.log
playground.wordpress.net.har.zip
ajitbohra commentedon Nov 19, 2026
Update:
Playground is now working as expected
brandonpayton commentedon Nov 19, 2026
@ajitbohra, thank you for the HAR file and for the update that Playground is loading for you now.
I'm very sorry for the trouble. It appears to have been due to some hosting defense code that conflicted with the special way browsers request Service Worker resources.
@erfanteymourii @Qaisarkhursheed Are you still experiencing the issue?