Komodo 12 and MCTS

[lkaufman]

@mark and larry, now that K12 is out will you make 10 free?

Probably we will before too long, but not right away.

[MikeB]

Larry and Mark,

Thank you for the new release - it is always a joy having a new Komodo to play with.

best regards,

Michael

[MikeB]

Christ, dude, we heard you the first time...

WTF IS it with this forum? Why is a civil conversation so damn difficult?

Well you can now just use the "add foe" option to the profile of the user that does not know how to be civil. Then all post from that user will be hidden from you.

Great option, one suggestion for all - when responding to someone being hostile, no need to quote the hostile remarks. They will soon be posting to themselves.

[Kaj Soderberg]

Thank you for this new engine, Mark and Larry.

MCTS is great fun to play around with, though K12 "normal" indeed seems stronger.

I notice that occasionally MCTS freezes, both on Windows and on Mac OS.
On my quad, it does not matter if 1 use 1, 2 or 3 threads or if i use more or less multi-pv.
The thing just sometimes freezes: the clock does not "tick" and the search depth remains the same.
Odd behavior or something fully explainable?

Thanks,
Kaj

[lkaufman]

Thank you for this new engine, Mark and Larry.

MCTS is great fun to play around with, though K12 "normal" indeed seems stronger.

I notice that occasionally MCTS freezes, both on Windows and on Mac OS.
On my quad, it does not matter if 1 use 1, 2 or 3 threads or if i use more or less multi-pv.
The thing just sometimes freezes: the clock does not "tick" and the search depth remains the same.
Odd behavior or something fully explainable?

Thanks,
Kaj

Does this happen after leaving it to analyze in infinite mode for a fairly long time, like an hour or so? If so it is because memory limits are reached. This isn't a bug, but perhaps we'll have a fix for it in a future update.

[Kaj Soderberg]

Hi Larry,

Indeed this occurs in infinite analysis mode, sometimes after 10 minutes, sometimes much later.
So it's a memory issue. Maybe to be fixed later on. OK.
What puzzles me a bit, not being a techie, is that it only occasionally occurs!?
Any practical idea for a workaround for the time being, settingswise or so?

Cheers, Kaj

[Jesse Gersenson]

Does this happen after leaving it to analyze in infinite mode for a fairly long time, like an hour or so? If so it is because memory limits are reached. This isn't a bug, but perhaps we'll have a fix for it in a future update.

Isn't a bug?! Is it expected behavior?

Komodo's causing his computer to freeze, that's a serious bug.

[Kaj Soderberg]

Does this happen after leaving it to analyze in infinite mode for a fairly long time, like an hour or so? If so it is because memory limits are reached. This isn't a bug, but perhaps we'll have a fix for it in a future update.

Isn't a bug?! Is it expected behavior?

Komodo's causing his computer to freeze, that's a serious bug.

Just a bit of clarification. The engine freezes and causes the GUI to stop giving information. The PC is still responding.
Cheers, Kaj

[mjlef]

Does this happen after leaving it to analyze in infinite mode for a fairly long time, like an hour or so? If so it is because memory limits are reached. This isn't a bug, but perhaps we'll have a fix for it in a future update.

Isn't a bug?! Is it expected behavior?

Komodo's causing his computer to freeze, that's a serious bug.

Just a bit of clarification. The engine freezes and causes the GUI to stop giving information. The PC is still responding.
Cheers, Kaj

It seems odd to me that the GUI would stop working. Nothing we do in the engine should make the GUI stop. The GUI has its own clock. In MCTS mode, when the memory for MCTS nodes fills up, it simply stops searching and returns the best move. If you find a position where the GUI freezing is repeatable, please send me the position, the GUI you are using and your settings and I can try to reproduce on my machines.

[Joerg Oster]

Isn't a bug?! Is it expected behavior?

Komodo's causing his computer to freeze, that's a serious bug.

Just a bit of clarification. The engine freezes and causes the GUI to stop giving information. The PC is still responding.
Cheers, Kaj

It seems odd to me that the GUI would stop working. Nothing we do in the engine should make the GUI stop. The GUI has its own clock. In MCTS mode, when the memory for MCTS nodes fills up, it simply stops searching and returns the best move. If you find a position where the GUI freezing is repeatable, please send me the position, the GUI you are using and your settings and I can try to reproduce on my machines.

Do you send the best move only after the GUI has sent the 'stop' command?
If not, this might eventually confuse the GUI?!