Yeah, for me neither.
- Feed Queries
- All Stories
- Search
- Feed Search
- Transactions
- Transaction Logs
Sep 11 2025
Maybe DNS does not work in your environment. These tests work fine for me.
Sep 10 2025
Sep 8 2025
Sep 5 2025
- Use array shapes
- Improved AuthController docs
- More API docs
- More API docs, fixed regressions
Sep 4 2025
- Hide not implemented routes
- API docs
- API docs
- Introduce PlanResource, add 'plans' to device info response
Sep 3 2025
- Fix regressions, more api docs for the SignupController
There's some regressions I'm working on fixing.
Sep 1 2025
I think for Roundcube and most other services the startup hook is early enough. Of course, the logon phase is a separate problem, as we need an authenticated user to get the config from cockpit.
Aug 29 2025
Aug 25 2025
I guess, separately we should make sure user does not do something stupid when the form is submitted. I.e. he can still manually set mail.inbox to any subfolder. But that's a separate issue.
Aug 22 2025
- Remove wrong TODO comments
Aug 20 2025
Using ->withoutMiddleware(['allowedHosts']) would be shorter, but I'm not sure which I'd prefer here.
Aug 19 2025
Aug 18 2025
Aug 14 2025
Aug 11 2025
Aug 8 2025
- Hide Wallet and Policy in the Dashboard for wallet controllers that aren't owners
- Enable wallet controllers to create users
- Enable wallet controllers to create shared folders
- Enable wallet controllers to create groups
- Enable wallet controllers to create domains
- Make wallet controllers able to create resources
Aug 6 2025
Aug 1 2025
Jul 30 2025
- iTip REQUEST notifications
- Tests
Jul 29 2025
- Small refactoring
- Refactor
- Check recipient in a REQUEST
- Cleanup
Jul 28 2025
- Cleanup FIXME/TODO comments
- Proper SEQUENCE handling with tests
I would probably not incorporate an obsolete features into an existing code, but ok.
I wonder if 403 in other circumstances would not be more generic. E.g. authentication issue.
This is kinda mess. Maybe a better approach would be to just check that the request comes from the "internal" location? E.g. by having list of IP networks (or IPs). But that potentially may be more complicated. Another possibility, a "client secret", but this probably would be event more problematic.
Jul 25 2025
Jul 22 2025
Jul 21 2025
Jul 18 2025
Jul 17 2025
- Use lower-case queries for consistency with the rest
- Add user:debug-start and user:debug-stop commands
Jul 16 2025
Nitpicking.
Jul 15 2025
It makes it more standard if we use 204 when the reply is indeed empty. So, it might make sense to use both 200 and 204 according to the response body being empty or not. Also, the header name might be better as X-Kolab-Mailfilter-Action.
In D5487#71508, @mollekopf wrote:probably needs a test adjusted someplace.