Are engine author in favor of book limits in tournament ?

[mathmoi]

Hi,

I'd like to know how many author are in favor of a limit to the number of moves an engine could get from it's book. If you're not an engine author, please don't vote, maybe you can give us your opinion in a comment.

Here is the discussion that sparked the idea : http://talkchess.com/forum/viewtopic.php?t=32821

[mathmoi]

For the record, I'm against the idea.

[swami]

After some thought I decided "Engines should be allowed to bring any book they can get"

I like the opening of Spark against Ikarus in CCT Blitz. It was great fun to watch.

[jshriver]

I voted no, simply because limiting the opening book is just a path to limit other forms of precomputed data like learning and maybe even egtb's.

Code is code, the the overall strength of an engine should be tested not only it's software design but it's ability to learn and grow and reuse that knowledge.

imho.
-Josh

[metax]

I wanted to vote 'No, engine should be allowed to bring any book they can get.' but unfortunately, I seem to have chosen the second radio button.

[garybelton]

My comment Mathieu is related to the game played between Sjeng and Hiarcs in the CCT12 main event. The game was at least 50 moves in book, At some point, probably around move 25-30, the remaining moves played by both engines were all moves found by Rybka. The TD goes to great length to make sure all the contestants kib their PV's to ensure that all the engines are legit and not copies of the strongest engine and it's derivates, but for moves 25-50 both Sjeng and Hiarcs might as well have been Rybka. Doesn't seem sensible to make sure Rybka clones are not in the tourney but at the same time allow engines to play Rybka moves ad infinitum, does it?

[Togga]

one tourney without books
one tourney with limited books
one other tourney without book limits

its free to decide if you take part in such tourney or not.