Re: new draft for the minimum value setting mechanism of HTTP2.0 Window and Window_update

Hiya,

To add to what Martin asked, I'd like to understand if there is really
_any_ minimum value that is safe to use. It seems that an attacker could
always try to game the system, so implementations should be defensive.

Picking too granular a minimum value also seems prone to problems. I can
see the possibility for some deadlocks to happen here unless endpoints can
commit to buffering.

Cheers
Lucas