(from TD 100)  by György Bakcsi 
The Computer  enemy and friend (In the 60ies and the 90ies)
by György Bakcsi, Grandmaster in composition 1980
The appearence of the solving programs  more and more developed  was sometimes fatal
for the old problems. We could see, that very nice prizewinners, which were published
a lot of times, and even in the FIDEAlbums, were cooked, or the merciless machine found
heavy duals in them. This happened especially with the moremovers, where the fantasy of
the earlier solvers was not at the level of the computer of the 90ies.
So, the program became an enemy  but it could be our best friend too, if we can rightly
ask him! In my youth I was instinctively afraid of "dangerous positions" (free white
queen in helpmate, too heavy material etc.), and therefore, maybe, I myself "censored"
my problems. Now I ask the computer, in very many cases he "laughs", but there are
some  unforgettable  moments, when the idea is sound, "we have a position", and now we
ask further the program for the best realization.
I tell you three stories.
Diagram 1. Here is a complete cycle of four white moves in the 2nd, 3rd
and 4th moves*). The idea was exceptionally difficult in the 60ies, without any program.
You see 27 pieces on the board, the scheme itself demands more then 20 pieces, and we
must always account for the correctness of the problem. Nevertheless, I succeeded (?),
the problem won 1st prize, it appeared in the FIDEAlbum (No. 376), and was published
many times.
Only in the 80ies my friend, Matti Myllyniemi sent me the terrible dual (is was found
by his program): after 1  Bh6 there is the illogical (for the normal brain) move:
2. Bxe8!, and in all cases white mates in the 4th move.
At that time I felt myself defeated. It seemed impossible to correct the problem. But in
the beginning of the 90ies, having a new and speedier program, I tried again (probably
three decades after the birth of the idea). I fought about three months with the
computer, but now I defeated him. Here is the final, C+ version diagram 2.
Somebody can say: the Bd6 is a promoted man. He is right, but I think, in our days this
is an excusable fault. If somebody finds a better position, I accept it. But at the same
time I was glad saving one of my most important achievements.
Diagram 1 György Bakcsi 1 PR Nagy Ödön Mem. Ty. 1967

Diagram 2 György Bakcsi Correction

Show solution 
Same solution! 
The second story was almost the same, but the position is more simple.
Diagram 3. Here is a cyclic interference of four black pieces. However,
the problem was cooked after some years, after the second white move it is possible to
play 3. Re4 Kg6 4. Rf4 f3# too. At that time this dual seemed logical, and I
gave up the correction. But in 1994 I tried again, "rightly" asking the
computer. Here is the final position diagram 4. The computer was a real
friend, I found this correct position after 2025 "questions". (For instance: the
computer found the sly new dual without the Sa1: 1. Rb3 Kd1 2. Ba1 Ke2 3. Rb2 Kf3
4. Rd8 Bxd8#).
Diagram 3 György Bakcsi 3 PR Ty. of Chesscircle in Bologna 1965

Diagram 4 György Bakcsi Correction

Show solution 
Show solution 
Diagram 5 György Bakcsi 2 Comm. Problem 1960

Diagram 6 György Bakcsi 1 PR Sakkélet 1997

Show solution 
Show solution 
Somebody can say after these stories: it is very easy, "the computer makes all". This is not right: the computer "makes" only unnecessary and banal "problems"  on the first place remains the idea and the composer. But the computer can be a very good friend  for the true composers.