[Retros] A rhetoric question

[andrew buchanan]

Hi Andrey,


> More

> specifically, does it make sense to produce SPG shorties hundreds of

> which can be computer-generated in close to no time at all? Shouldn’t

> we indeed concentrate now on longer, more complicated and more

> esthetic-looking records, taking our chance while the PCs being used

> are not powerful enough yet?

> 

> Best,

> Andrey


Well, I certainly don't want to produce stuff which a computer can
produce today. That's why I encouraged Mario & Francois to run their
programs.

But having done that, and established the boundary of what their
software & hardware can tackle today (about 12 ply), I can't (and
others don't seem to) resist the temptation to work on some of these PG
challenges which are just outside the known range. 

It's the same kick that I imagine some people get from solving a sodoku
puzzle (which I don't do).

To me, it's actually a bonus that some time, programs are going to be
able to verify whether our answers are the best ones.

It does trouble me than there are so many of these little challenges
(88) and that there is no clear boundary to stop us sliding towards the
situation of looking at the cases of *all* last moves. For me, I like
the 8th rank ones, but that's as far as I'll go.

cheers,
andrew.