[Retros] A SPG admitting a false skeleton.

[Nicolas DUPONT]

Dear retro friends,

Let's consider the following problem :

tc1d1tr1/pppcp2p/3p1pP1/8/8/1C1P4/PPf1PP1P/T2DRFCT

SPG in 18. (C+).

The solution is of course obvious. The interesting fact is that the
"extracting" problem in 16 crashes.

Does such a phenomenon already has been observed ?

Nicolas.