Leif Åstrand

As a workaround, you can use the same endpoint implementation and inject different implementations of the data provider into it. With some additional boilerplate, you can even let the endpoint...

@netbeansuser2019 It seems like I was making an invalid assumption in the forum thread. The situation described in this ticket is still something that can happen and may be confusing...

Rather than moving this to the Spring repo and only fixing it there, I would suggest somehow doing this on the Flow side so that service init listeners defined as...

I was originally thinking that the logic in `VaadinService.init` that actually invokes the listeners could omit the duplicates, but now I realize that there isn't any good way for it...

One problem with the suggested approach is how to arrange it so that the field can get validated immediately when it is no longer disabled/readonly/invisible. I'm wondering whether it would...

Got a hint that `org.springframework.boot.actuate.autoconfigure.security.EndpointRequest` might be quite close to how we would want to deal with whitelisting access to static resources.

There's supposed to be some additional information about the message payload after `Message start:` in the thing that is logged on the debug level. Looking at that data might reveal...

I don't think this is a proper solution. The purpose of `StaticFileServer` is to emulate the servlet container's "default" servlet for serving static resources in the cases where the Vaadin...

If you explicitly deploy the `default` servlet to some specific path, then it's very clear that you do that explicitly for the purpose of making static resources available below that...

Or maybe the proper solution would simply be to revert https://github.com/vaadin/flow/issues/13769?