> > This requires new application protocol verbs "STARTCOMPRESSION", > > "STOPCOMPRESSION", and underlying support in the TLS layer.
> I wonder if it would have been possible to do this via renegotiation, though > this has overhead. Intriguing, but moot of course, since renegotiation is gone. :) Interesting corner-cases to think about: is compression restarted, or do you preserve all state to pick up where you left off (e.g., the zlib dictionary)? Either either one probably has some security issues to think about. _______________________________________________ TLS mailing list TLS@ietf.org https://www.ietf.org/mailman/listinfo/tls