Greg Wilkins

@stuartwdouglas also can you clarify how you handle something like: ```java public void onWritePossible() { ... out.write(IlikeBigBuffers); assert(isReady()==false); out.close(); } ``` Ie if `isReady()` has been called and returned false,...

+1 `close()` should not block +1 `flush()` is a write and a call to `isReady()` returning true is required before another write. I'm not sure I agree that `isReady()` is...

The API already has https://github.com/eclipse-ee4j/servlet-api/blob/master/api/src/main/java/jakarta/servlet/ServletRequestListener.java, which allows an app to be called when a thread dispatched for a request is about to leave scope. I think this is sufficient for...

I think such a new API is out of scope for 5.1... however we should clarify our existing APIs in similar circumstances such as clarifications of HTTP/2 behaviour with regards...

I have created a test that does a POST with query params, body params and forward query params as follows: POST query | POST body | forward query -- |...

@markt-asf If you agree Jetty is wrong here, then please close this one.

Whilst adding Local and Remote variants of getScheme would bring the API into line with Host/Addr/Port, I'm not so sure that is sufficient. Those methods have never been either fully...

@stuartwdouglas I guess with current forwarded header handling, the container needs to be configured to even pay attention to any of those headers. So I agree having a raw getForwardedHeader...

@stuartwdouglas yeah I think the existing methods should continue to work the way they are (but perhaps with better documentation). Depending on the exact headers and configuration, Jetty can overwrite...

My thoughts are as follows: + checkstyle must pass + CI must pass + PRs can come from branches in any repo. Committers must cleanup branches in the main repo....