Arend wrote: > > I wrote: > > > Arend wrote: > > > > The obvious question: How big is the performance penalty? > > > > That's of course s.th. we need to take into account when discussing > > > > whether we want this in mainline. > > > > > > I don't know, because I cannot run "real" tests yet: they depend on all > > > the GNU Go at once and I still haven't converted owl and > > > `value_moves.c'. I don't expect the performance hit to be over 3% > > > though. > > > > Unfortunately, I was wrong. Performance hit is around 10%: 6885 seconds > > CPU time against 6106 seconds. Timing may be not very precise as I used > > the box during regression runs, but the main picture is clear. > > Which CPU, compiler? I may try to run some tests myself.
Athlon XP 2500+, GCC 3.2. > > In addition, there is some subtle bug that causes node counters > > differences on some test files. I was unable to track it down. > > > > So, the question is: should I abandon this project? I myself is > > uncertain. 10% or even more seems like somewhat too high a price. > > You should definitely put the work you have already done on the branch. > If it is really 10%, it feels like a bit too much to me. Maybe forking > would actually be a better idea? It is on the branch already. I'm not sure there is much sense in forking. It seems unlikely that we can support two branches at once given current development rate. Paul _______________________________________________ gnugo-devel mailing list gnugo-devel@gnu.org http://lists.gnu.org/mailman/listinfo/gnugo-devel
