Peter C wrote:BTW, are you sure Fabien didn't invent LMR? Or am I just misinformed?
![Confused :?](./images/smilies/icon_e_confused.gif)
You are misinformed.
The first strong, popular open source program I am aware of that used LMR was Pepito. The idea is a lot older than that, its origins must be considered lost in antiquity. Back in 2004, the idea was widely known, but nobody apart from Sergei Markoff and myself seemed to believe in it. I used to pester the CCC with lots of posts about the technique, desperately trying to convince people to at least give it a try. Fabien finally gave in and implemented it in Fruit 2.0 (initially in a simplified version which didn't re-search in case of a fail high on a reduced depth search). At the time, Fabien didn't like the idea and wanted it to be disabled (there was a UCI parameter for switching it on and off) by default. Enabling it was a last-minute decision, caused by the stubborn insistence of beta tester Joachim Rang.
In my opinion, most of the discussion in this thread is silly and 15 or 20 years outdated. Carefully guarding secrets with the hope of gaining commercial advantage is pointless, because strength is no longer an important selling point. All reasonably bug-free modern chess engines are so much stronger than the average human player that they for all practical purposes become almost completely equivalent. Apart from an extremely tiny minority (which for obvious reasons is overrepresented on this message board),
people just don't care about strength. If you want to make money from computer chess, target the computing platforms people actually use (i.e. cellphones, not PCs), and focus on features, beautiful graphics, and good ways to dumb down the program, not on maximum playing strength. Glaurung on the iPhone has been downloaded more than 200,000 times, and in the countless e-mails I have received with wishes for future versions,
noone has complained that it's too weak.
Also, even if you do care about strength, openness makes it so much faster and easier to improve, both on the community level and on the individual level. Stockfish isn't strong despite being open source, it's strong
because it's open source.