Hey @brutalbirdie is there a specific reason why # Where will social media icons be saved - local or cloudflare. STORAGE_PROVIDER="local" is set in the start.sh? I'd like to test this with Cloudflare R2

For anyone having those problems, here's quick update for this one: After a few weeks someone from the Postiz Discord (terrible way and platform for bug reporting btw) suggested to set up Cloudflare R2. It took two restarts, but after that I managed to connect my LinkedIn pages, because apparently it wasn't able to store the profile images and thus created that redirect above. Although, now that I'm writing this, the bucket it still empty, but there are files in /app/data/data/2025/11/27 (the second data folder wasn't there yet, I needed to create that one manually, maybe that solved it after all?) Also, even with the working connection, this error appeared in the logs, maybe we could let that point to somewhere it can write: Nov 27 19:37:40 ⨯ Failed to write image to cache x5s7nobKRgx-CJ5Nq4z9X1PXnOjxJPxqtCDckv+9gKk= Error: ENOENT: no such file or directory, mkdir '/app/code/apps/frontend/.next/cache/images' Nov 27 19:37:40 at async Object.mkdir (node:internal/fs/promises:860:10) Nov 27 19:37:40 at async writeToCacheDir (/app/code/node_modules/next/dist/server/image-optimizer.js:190:5) Nov 27 19:37:40 at async ImageOptimizerCache.set (/app/code/node_modules/next/dist/server/image-optimizer.js:574:13) Nov 27 19:37:40 at async /app/code/node_modules/next/dist/server/response-cache/index.js:121:25 Nov 27 19:37:40 at async /app/code/node_modules/next/dist/lib/batcher.js:45:32 { Nov 27 19:37:40 errno: -2, Nov 27 19:37:40 code: 'ENOENT', Nov 27 19:37:40 syscall: 'mkdir', Nov 27 19:37:40 path: '/app/code/apps/frontend/.next/cache/images' Nov 27 19:37:40 }

@nebulon said in Postiz Isn't Responding: Published a new package, you have to update to that first, otherwise the app won't even start up. Thank you @nebulon now its working normally.

@nebulon said in Open registration: The env file used to be called env.sh but got renamed since then to not use .sh extension to match the other apps. Adding to that, this also means export is no longer needed in the /app/data/env file. So you can just do: DISABLE_REGISTRATION=false

Hello @badmonky If you have chosen on installation that Cloudron should handle the user management, then Cloudron does handle the users. I am not sure how Postiz acts if ODIC (Cloudron) is enabled and you invite local managed users.

Update: I restarted the app again, and now it worked. I don't know what may have been the issue, but something seems to have a mind of it's own.

@andreasdueren I published this one but it has a bunch of issues from some testing. Each SSO user is 'independent'. They can all separately invite users. If registration is enabled, the app becomes open for all If registration is disabled, invited users cannot login For this reason, even in SSO mode , it is still left open Even if registration is disabled, one user can be registered initially. This is probably a feature. I guess we should report this upstream , if we collect the issues

I had a browser stuck looping and no api or URI would log out. Deleting the browser cookies and site data was the only way it stopped.

@plog I think this might be the wrong forum for your post. This forum is about Cloudron and Cloudron is actually not using docker compose at all.

Hello @shrey The latest version was released https://forum.cloudron.io/post/111048

@visamp it doesn't looke like postiz has any 2fa support. Maybe you can ask upstream https://github.com/gitroomhq/postiz-app/discussions

@sponch you have to add "export" . so maybe export MASTODON_URL=https://my.instance.social (watch out for quotes and spaces)

@jordankrueger Excellent work, thanks for letting us know!

Ahh I just assumed that was fixed from reading this forum. That was it set to 1MB still. Thanks

This will be picked up by the package meta info once that version is released and picked up by your cloudron