Re: [gnugo-devel] Joseki implementation question

[Evan Berggren Daniel]

On Thu, 24 Jul 2003, SP LEE wrote:

> Actually I got this question from the regression test nando:212. I didn't
> understand why gnugo doesn't know such simple joseki. Then I see in
> hoshi_keima.db, the sequence for invading the corner is only included for
> ikken basame (not sure if this is the right word). However, this sequence is
> valid for many positions of basame. It seems we need more implementation of
> these.

This patch adds another variation to hoshi_keima.sgf.  I don't see this
line in Ishida, so this is really just my own work here (that is, someone
stronger should make sure it looks ok).

This produces three PASSes at nando:210, nando:211, and nando:212.
nando:210 should be considered basically accidental; the other two are
properly solved.

I haven't run the rest of the regression tests, but I don't anticipate any
changes.

Evan Daniel


diff -u patterns/hoshi_keima.sgf patterns/hoshi_keima.sgf
--- patterns/hoshi_keima.sgf    23 Jul 2003 19:46:05 -0000
+++ patterns/hoshi_keima.sgf    25 Jul 2003 05:08:03 -0000
@@ -585,4 +585,18 @@
 (;B[pf];W[pg]MA[mh])
 )
+
+(;W[qc]MA[lh];B[qd]MA[ii];W[pc]MA[lh];B[od]MA[lh]C[U
+];W[rd]MA[lh];
+B[re]MA[lh];W[rc]MA[lh];B[qe]MA[lh]C[U
+];W[nc]MA[lh]
+(;B[pf]MA[lh])
+
+(;B[tt];W[of]MA[ij]C[j
+];B[nd]MA[ij]C[U
+];W[mc]MA[ij]C[U
+];B[md]MA[ij]
+C[j
+])
+)
 )