if you have any prove that OS/4 is based on leaked Merlin, pls, show them to us.
Without wanting to demote developments possibly based on "open source": the party which has to prove something heavily depends on the legal system, and IBM may not even be allowed to disassemble e.g. my source code to prove that I've used leaked source code.
They wouldn't have to. You can look for patterns of code without regard to physical offsets (something anti-virus software does, for example), and without disassembling it, but simply comparing some known patterns in your code to what you see in the other code.
We could probably do this ourselves, by the way, to scan the existing kernel and find patterns that exist in the OS/4 kernel. If there are many more than several, it's a good sign it's a direct derivative work. If someone can post a copy of the true blue OS/2 kernel and the OS/4 kernel somewhere, I would be willing to write an analysis tool which looks for those patterns. It would be a fun project. :-)
x86 encodes immediate data and memory addresses directly into many of their instructions. So, we'd have to look for computer instruction patterns and skip those portions because they may not even be the same between separate compiles from the same source code (depending on the variables of compilation).
And copyrights may not be restricted to source code. Surely U2 will have a problem with a commercial "Without or with you"-song which, by luck, almost sounds like an improved version of their copyrighted song "With or without you". If it sounds like a clone, then a judge may decide that it's a clone. Even if it's a honest new song.
The whole thing about OS/4 right now is ... it's suspicious. For me, that's enough to avoid it completely and I'm content to leave it at that. We need an open-source OS/2 kernel anyway.
I also applaud Martin's position on this (wanting to do the right thing, to make sure he speaks rightly on the subject when he's in a teaching position). It's admirable.