Yea, they aren't random. They can be connected to something "random" and some languages will produce the same sequence after that. Yet nothing is really random, though there are people that can tell you what side of the coin it will land after you throw it(air, muscles etc. there are components that aren't random either, some can do it with dice as well).
I think its more about variations then randomness.
However, I got the impression AKs problem went beyond that, as he mention user interaction also provide the same result at least in the apps they tried.
What we did is trying to use normal random numbers, but going through normalization that provides output that fits to patterns found in nature. As programming languages will change them so they are even in all aspects, they aren't even in nature. This is the only logical solution to all problems. As logically, if user interaction with delay still provide wrong input then only logical thing is that the somewhere along the way something else gets messed up in the way the programming languages handle that. And the way numbers are normalizaed in programming language seemed to be the only possible problem(even distribution etc).
Either that, or the apps they tested that were with user interaction(shaking the phone it was I think), didn't really used their interaction and stil just made random numbers.
Overall is very complex topic and until we have enough practitioners with high enough accuracy none of this can be tested in a equal level for all of us for WWG... All problems will have a solutaiton but without being able to test, its wise to assume there either isn't a problem or we fixed it,already.So I guess moving on seems the best idea. QMDJ doesn't have that problems, no random numbers in there for example. : )
There are many solutions though. There are appis that use special hardware that used movement of air, there are libraries that use date of last post in reddit and all other kinds of stuff... And depending what it is, if its the 'same sequence' problem we could just get all 3 numbers from the first set...
But again - until we know the system well enough to test it, all is blind attempts to fix problem that we can't really see. And that is waste of time. : )
So I just added the most logical thing, and time will tell if it changed anything. I'm still aiming at other systems now, though, this was just a temporary step backwards as there was another project that I tried to do with the WWG calculators, so seemed good idea to at least fix that on the way, as
the other thing didn't really worked out with the current code(it doesn't really matter, in any case).