|(1) Posted by Steven Dowd [Sunday, Feb 7, 2010 20:24]|
Can anyone tell me how to set Popeye to solve for set play only?
That is, a h=7,5 instead of a h=8 "with set play"?
My attempts have proven fruitless. I use Thomas Bark's interface with PY 4,55.
|(2) Posted by Geoff Foster [Sunday, Feb 7, 2010 21:12]|
Popeye accepts problems with an extra half-move e.g. h=7.5
Perhaps you are using "," instead of "."?
|(3) Posted by Steven Dowd [Sunday, Feb 7, 2010 21:15]|
You could be correct, Geoff. I can't enlarge the text and I often press the "," instead of the "." button, and if the font is small, my eyes don't recognize the difference.
|(4) Posted by Joost de Heer [Sunday, Feb 7, 2010 21:22]|
If you don't like dots:
|(5) Posted by Steven Dowd [Tuesday, Feb 16, 2010 17:47]|
I have another question. Why doesn't Popeye show its progress?
I have a difficult h=7 that I had to "load up" a bit first with material (5+12) due to cooks (its not a lot really just two extra pawns). It has been hanging on my computer running a quad core, Popeye 4.55 64 bit for over 2 days now. It cooked its predecessors in a day or under in intelligent mode, which I am using. It's especially frustrating because this was supposed to be a test that would tell me how to get to the h=8 that I know is there.
With other programs you know where they stand in the solution process. For all I know, Popeye may not even be running, although the little cursor is blinking.
It's frustrating because I have no idea how long to wait or whether it will EVER solve this problem. How do you work with that? Why have the developers not tried to put some sort of "progress meter" in Popeye?
Someone who might know how long it should take, I'd be happy to show you the problem and its expected solution.
|(6) Posted by Geoff Foster [Tuesday, Feb 16, 2010 21:59]|
You should use "option MoveNumbers". With intelligent mode you then have "option Intelligent MoveNumbers".
|(7) Posted by Dan Meinking [Tuesday, Feb 16, 2010 22:16]|
I suggested several 'intel' testing options on SourceForge:
If Popeye uses a square-by-square analysis method, these options would show progress on a regular and meaningful basis. Adding a 'timestamp' option (show date/time on every generated line) could also be helpful.
|(8) Posted by Geoff Foster [Tuesday, Feb 16, 2010 23:18]|
Your problem has two black pawns on the second rank, ready to promote. The presence of pawns close to their promotion square slows Popeye considerably, because there are more combinations of pieces to consider. One trick which often helps is to use the "PromOnly" condition. For example, with "PromOnly B" Popeye will only consider promotions to bishop. I used this condition on your problem and Popeye quickly found a cook. I'll send it to you in a note.
|(9) Posted by Ian Shanahan [Wednesday, Feb 17, 2010 02:16]; edited by Ian Shanahan [10-02-17]|
@Steven, who wrote:
"It's frustrating because I have no idea how long to wait or whether it will EVER solve this problem."
Strictly speaking, this conundrum is insurmountable - being akin to Alan Turing's "blank tape halting problem" (the undecidability of whether an arbitrary computer program will run forever or eventually halt given a blank input). I've seen this idea used in a proof of Kurt Gödel's Incompleteness Theorem - far too involuted to explicate here...
|(10) Posted by Steven Dowd [Wednesday, Feb 17, 2010 02:46]|
Thanks for the help Guys......
No more posts
MatPlus.Net Forum Internet and Computing Popeye