gnugo-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gnugo-devel] more bugfixes


From: Gunnar Farneback
Subject: Re: [gnugo-devel] more bugfixes
Date: Sun, 10 Nov 2002 10:44:26 +0100
User-agent: EMH/1.14.1 SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (sparc-sun-solaris2.7) (with unibyte mode)

Evan wrote:
> - findlib() was returning a more liberties than it had copied if liberties
> > maxlib.  This would cause a crash in some cases (namely, my current work
> with attack5/defend5.  I can send in the version that has problems if
> desired).

That is not a bug, it's just a failure to read the documentation.

| /* Find the liberties of the string at str. str must not be
|  * empty. The locations of up to maxlib liberties are written into
|  * libs[]. The full number of liberties is returned.
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|  *
|  * If you want the locations of all liberties, whatever their number,
|  * you should pass MAXLIBS as the value for maxlib and allocate space
|  * for libs[] accordingly.
|  */

Or is the marked sentence ambiguous in some way?

> - approxlib() was making unneccessary compares of the form liberties <
> maxlib.

Looks correct.

> - owl1:268 was incorrect (should have been written [3 A1]).

Correct.

> It is also equivalent to trevor:360.

It's not. In trevor:360 white can capture at E9 and threaten to run
out. 

> owl1:268 is commented out and trevor:360 is corrected to be more
> specific.

I see no reason for these changes. There are subtle differences
between the tests and it's not like they take a lot of time to run.

A bugfix for owl1:268 would be in order of course.

/Gunnar




reply via email to

[Prev in Thread] Current Thread [Next in Thread]