somebody will try to enter Ippo, Robbo or such in an ACCA event.
The ACCA does not allow clones in any of the competitions. That means clones of any kind (legal or illegal). So, even Toga is not allowed.
Also, the ACCA requires the entrant to name the author and various other info about the entrant including team members.
After reading the Ippo web page, I see this statement "Decembristas will continue being a secret society, whose population could as great
number as the Russia Mother.". Based on this alone, such an entrant would be disqualified.
I've read through some of the Ippo code (mostly the eval) and talk about obfuscation. There are typically only two reasons for that. First,
you are a commercial developer and you want to make it as hard as possible for somebody to copy/reverse engineer your code. Secondly,
you are cloning another person's code and you are trying to hide it. Obviously, the first reason is not the case. That brings us to the second reason.
There isn't any reason to obfuscate your code and release it freely in hopes that all would learn from it. If your plan is for others to learn,
then you make your code as clear as possible. That leaves me only with they are cloning.
Looking at their future plans, I see this:
Code: Select all
First Values with IvanHoe
* Clean up the code and put it more comments for general understanding.
* Add multipv and searchmoves and ponder in the UCI parsing options.
* Introduce a testing system for making further advances.
o Super fast testing (game in 3 seconds?) for evaluation changes.
* Support the Windows environment more fully.
o Initial aim is especially for RobboBases to work in it.
* Add many extra utilities, more for developers than users.
o Structure the evaluation into segments.
+ Allow printing of eval components as with Crafty.
o Put in perft and move verifiers, and search tracers, and more.
+ A wild idea, GUI support for search tree dumps?
+ A saner idea, good statistics models to churn thru search tree dumps.
o Factually, a whole DEBUG mode can be useful.
defend their code. In order to do that, you have to admit who you are.
What poorly thought through idea. If you don't claim your code and will not admit who you are, then anybody could claim it and
make it commercial. What a slap in the face that will be.
So, the code and info on the web pages suggests that it is a clone. Also, there is nothing on the web pages that states it is an original
work. Some of the experiences of people trying to use it suggest it is a partial clone. Again I state: the author(s) will have to defend their code.
I have reviewed several chess programs over the last 12 years. It is easy to spot an original work. Ippo doesn't show any signs of it being an original work.
So, why is partial cloning an issue. Lets say you only clone the eval and the search. Well, that is like somebody publishing their
memoirs that covers them through the age of 60 or so, but they are only 21. So, they copied the rest from another book. Is that
wrong/illegal? A girl from India tried that just a few years ago and yes, it is illegal.
Clone or not: if you are an author/researcher that will not name yourself or your team members then you may not enter
ACCA events and that has been a rule for years.
