Abstract: An invention is disclosed for classifying a graphic—e.g. as text or non-text. In embodiments, machine learning is used to generate a solution for classifying graphics of a graphic based on providing the machine learning system a plurality of graphics that are already classified. The way to determine a classification is then used by a remote presentation session server to classify tiles of frames to be transmitted to a client in a remote presentation session. The server encodes the tiles based on their classifications and transmits the encoded tiles to the client.

Abstract: In an embodiment, techniques for sustaining session connections are provided. The techniques send heartbeat messages when not sending a message may cause the session connection to close because of a timeout condition. Heartbeat messages are valid transport layer messages that will be ignored by protocols at higher levels of a data communications stack. As an example, the techniques may send a TCP message containing only a carriage return and line feed (“CRLF”) in its payload. Because the TCP layer considers a message containing only a CRLF to be a valid TCP message, intermediary computing devices such as proxy servers may not interpret heartbeat messages as “keep alive” messages, and may sustain session connections.