Re: [gnugo-devel] cosmic patch

[bump]

Arend wrote:

> > I don't think so.
> > 
> > I ran the regressions, applied the patch, then ran them again.
> > I don't think I ran cvs update between.
> 
> That's not exactly very helpful. Can't you give the CVS date for which this
>happened, so that I can reproduce it and try to find the bug in
> cosmic_5_1.2a?

The annotations in my log say:

current cvs 8/6

        - and -

after cosmic_5_1.2a

The results are identical except for one test.

Nicolet wrote:

> It seems that the fail/passed status of that test in essentialy
> random. I think we are in front of a weird cached memory issue
> there, or maybe hitting a floating point precision limit.
> 
> With the *same* executable (current cvs, non cosmic, without
> recompiling), I get different results when I issue the two
> identical commands in a row in a console window :

The regressions are deterministic. However if there are two
moves which in normal play are made with equal probability it
seems conceivable that a small change in the code would affect
the deterministic answer.

It may not be important to figure out the exact mechanism. Perhaps
we should not worry about this.

Dan