Ippolit 2nd release

[George Tsavdaris]

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

I remember recently Dann Corbit and GCP(Sjeng) saying that what you say is just pointless and a matter of just some minutes of delay to the ambitious hacker.

[Gian-Carlo Pascutto]

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

I remember recently Dann Corbit and GCP(Sjeng) saying that what you say is just pointless and a matter of just some minutes of delay to the ambitious hacker.

I never said it was pointless. I said it wouldn't stop anyone but would increase the leadtime of the original programmer by making it harder. Which matches perfectly what the person you are replying to said.

Even delaying Ippolit by a month might have been enough to get Rybka 4 out the door and not have Vas be in the situation where he is now.

[George Tsavdaris]

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

I remember recently Dann Corbit and GCP(Sjeng) saying that what you say is just pointless and a matter of just some minutes of delay to the ambitious hacker.

I never said it was pointless. I said it wouldn't stop anyone but would increase the leadtime of the original programmer by making it harder. Which matches perfectly what the person you are replying to said.

Even delaying Ippolit by a month might have been enough to get Rybka 4 out the door and not have Vas be in the situation where he is now.

OK sorry then for paraphrasing you.

Would increase the leadtime. How much can you say an approximation? Because my opinion was that the delay is considerable(1-2 months) but Dann has corrected me saying it's a LOT lesser something like 5 minutes or so to do it(i hope i'm not paraphrasing him too).

[playjunior]

For a chess program it is very hard to really protect the code from reverse engineering.

The reason is that making the code hard to break makes it also slower. For a chess engine this is a no-no.

[Gerd Isenberg]

For a chess program it is very hard to really protect the code from reverse engineering.

The reason is that making the code hard to break makes it also slower. For a chess engine this is a no-no.

Make the search iterative rather than recursive. Make it huge spaghetti, some indirect jumps anyway. Mix SSE, MMX and gp instructions, some punpcklqdq and punpckhqdq opcodes make the code a bit harder to understand

[oreopoulos]

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

I remember Vas saying exactly the same thing for Rybka 3, when Strelka was out.

[oreopoulos]

Even delaying Ippolit by a month might have been enough to get Rybka 4 out the door and not have Vas be in the situation where he is now.

He is not in any situation. Ippolit is a nice engine (crippled in features). Its a good as a second opinion, but this will not hurt even a tiny bit Rybka sales.

I even think that some competition, and some free code with new ideas, will probably do good to everyone.

Think positive.

[bigo]

Rybka 4 will be out in about a month or so. It will be all that Ipollit is and much much more. It will be the chess engines hardcore users have been dreaming about, with lots of goodies yet not seen on any engine ...

How do you know all that?

The only negative thing this time though (for hackers that is) is that it will have a heavy padlock on its code. Hopefully that will not be intrusive enough to hurt performance.

If this is the true(and unfortunately it won't be easy at all to find out as we would not have the unlocked version of R4 to compare) then i guess many people will not buy the program. I will be one of them.

There is no copy protection that can't be broken. All it does is hurt the legit users, not the thieves. I used to take great delight in doing this, just to see if I could. One can do wonders with a debugger, if he knows assembly language fairly well.

My only comment: I absolutely refuse to buy _any_ software that requires that I insert the CD every time I want to use the program (and I am not talking chess engines here since I don't buy/use commercial ones). On my laptop, I often have the CD/DVD drive removed to make room for an extra battery. I will not buy software that requires any sort of "dongle" (a device plugged into a printer/serial/usb port needed to "unlock" the software). If I found something I _really_ wanted, I would buy it and defeat the copy-protection to make it actually usable again. I would not give away copies, but I do refuse to deal with draconian copy-protection methodologies.

Bob can I ask why you don't use or buy chess engine software? Don't you need the best to test Crafty? I heard you say this several times here, it sounds kind of arrogant as if you can't learn anything from commercial programs, what's your beef with the commercials. I don't really believe that you would admit to hacking a program for fun but not test it. Even if you did it for fun don't you feel somewhat dishonest?

[bob]

Nobody said anything about dongles or CD/DVD's ... I agree that doesn't work and I would also not buy anything that requires that. I remember early Chessbase software where I had to put a stupid dongle ... total FAIL. Also the entire thing with requiring you to put a DVD or CD just to get the program to run is a total turn off. Fortunately now with CB products if you run another engine in the GUI you don't need to insert any DVD to get the engine to run.

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

You can try the obvious and encrypt the executable, and use another executable to load and decrypt the engine. But that fails, you just wait until the real engine is running and then use the debugger to start poking around. You can't decrypt on the fly instruction by instruction as it would be too slow. As I said, this is a hopeless assignment.

[bob]

What I meant is in actual coding of the software so that it would not be so easy for someone to decompile and steal the code. I am sure that there are ways to make things more difficult, with Rybka 3 it apparently was too easy as no steps were done to prevent that.

I remember recently Dann Corbit and GCP(Sjeng) saying that what you say is just pointless and a matter of just some minutes of delay to the ambitious hacker.

I never said it was pointless. I said it wouldn't stop anyone but would increase the leadtime of the original programmer by making it harder. Which matches perfectly what the person you are replying to said.

Even delaying Ippolit by a month might have been enough to get Rybka 4 out the door and not have Vas be in the situation where he is now.

OK sorry then for paraphrasing you.

Would increase the leadtime. How much can you say an approximation? Because my opinion was that the delay is considerable(1-2 months) but Dann has corrected me saying it's a LOT lesser something like 5 minutes or so to do it(i hope i'm not paraphrasing him too).

Depends on what is done. But someone with a good debugger might have to waste a day or two or three to get past the protection, before diving in to the disassembly and then de-compiling tasks. It wouldn't be a huge / significant delay in general. I have even hacked up an operating system to help me figure out how to break copy protection so that I could run a program without the CD installed on my laptop. Been quite a while since I needed to do this since I now use linux and nothing is copy-protected like that (at least nothing I use).