On Jul 14, 2014 5:51 PM, "Greg Wilkins" <gregw@intalio.com> wrote: > Failing that, a dedicated static table bit is a good guess at a compromise for different sized header tables: > > 0 1 2 3 4 5 6 7 > +---+---+---+---+---+---+---+---+ > | 1 | S | Index (6+) | > +---+---------------------------+ The testing done thus far suggests two things: that the static table is rarely used after an initial phase; that a six bit index is somewhat, though not significantly, worse than a seven bit one at avoiding a need for two byte references. If you have good data, please share. Otherwise, I think that we need to be a little more respectful of the work that had already been done. Yes, more data could likely suggest an alternative course, that is always true. But knowing when to stop second guessing, that is the trick. If you plan to actually do thing, that is.
Received on Tuesday, 15 July 2014 03:43:38 UTC