This study appears cooked by 7-piece endgame db still unknown in 2005. After the solution finishes play ..Ke5 .Pa8=Q Be6 with a draw. Then again it is probably easily saved by placing Pf7 on f6 in the diagram. And I bet that has been discovered meanwhile since it is modern and worthwhile testing!
Endgame Study - A. Ornstein, 2005, Springaren.
Sort:
No I meant Pf7 to f6. The other pawn still leads to a drawn endgame while we want a win for white!
Actually, it's even worse. With a pawn on g6, black can play ..Pf5! after white .Bh3. So definitely no pawn on g6!
@lodrac91: Did you find anything about this one in HH db?
That is precisely the same cook as mine because the moves ..Ke5 and ..Be6 may be reversed! Which means that my proposed fix Pf7->f6 might work. Actually is does work for the remaining endgame but the question is whether it loses any content at an earlier stage which I am not sure of!
Plus a note. The variation given after 5 ... Bb7 is wrong but white does win!
Jul 19, 2021
0
#9
I don't understand the question. In which line?
TB gives 11 winning moves at that point. Does it matter? The original solution stopped after .a7 if I recall well. Do we miss any content here?
Ah, I see! I am not sure any line will convince anyone unless you complete it to a self-evident win. Even before the tablebases that question woud have arisen about the final diagram. Does white really win? The tablebases eased our mind on such questions as they dispelled all doubt. Which is good because it is a great endgame study!
Actually, as a composer, I wouldn't have dared to publish such a study 40 years ago feeling uncertain about its real status. TB saves it.
OK, I see what you mean. ..Qe8+ quickly wins a pawn. I agree it works in this variation but we are still left with the main line which depends on trusting the TB. In general I wonder if it is useful to provide variations in an endgame that basically leads to a difficult TB analysis. Why not simply state win by TB in all those lines.
Jul 19, 2021
0
#17
The win is technical only if black keep a2, in the main after 8.a7 it does not occur.
White cannot avoid this endgame in the main line either. After .a7 Ke5 .a8=Q Be6 the pawn is protected and white must sweat. It's part of the DNA of this study 
Jul 20, 2021
0
#19
OK I understand, but Black cannot keep g7, white play check at each move and g7 falls, example: Qa5 Bd5 Qe1 Kf5 Qf2 Ke5 Qg3, then could you build a fortress with Pf6 and B?
That's possible! I haven't dived that deep into the variations. I don't expect solvers will find your line by their own effort. If you do not include such analysis in the solution it's much simpler to refer to TB. In the end it doesn't matter a lot as I expect that consulting TB becomes as habitual to solvers as tying their shoes!
White to play and win.
https://www.chess.com/blog/lodrac91/a-ornstein-2005-springaren