On Mon, Jan 27, 2003 at 12:32:58AM +0200, Jarkko Hietaniemi wrote: > > perlbench, that well known reputable supplier of reliable timing comparisons, > > Don't do that to me, my belly is hurting from laughing. > > > thinks that blead is only 14% slower than 5.004_05 (both compiled with the > > same compiler and same options on the same machine) > > > > Can we kill pseudohashes yet? > > Their head is rolling in 5.10. In the 5.8 series we can't. Is there anything stopping them coming out of blead today? That would let us see what perlbench (and the vapourware reliable benchmark) thinks the speedup is. Nicholas ClarkThread Previous | Thread Next