Hato converts the multipart parameters into a stream, which makes HttpClient to unconditionally use an InputStreamPublisher, which doesn't set content length, which results in a chunked request.
That can be problematic because e.g. Trello doesn't support it - its API returns HTTP 411 if a request doesn't have content length (and chunked requests can't have content length).
Hato converts the multipart parameters into a stream, which makes
HttpClientto unconditionally use anInputStreamPublisher, which doesn't set content length, which results in a chunked request.That can be problematic because e.g. Trello doesn't support it - its API returns HTTP 411 if a request doesn't have content length (and chunked requests can't have content length).