Mutation protection opinion

[Numsgil]

cond, start, and else aren't metadata.  They are actual base pairs directly in the DNA.  When we say metadata, we mean information that isn't coded in base pairs.  Mutation rates at present are an example of metadata.  Each bot's DNA has a mutation rates structure, how likely point mutations are, etc., but it isn't coded in the DNA at all.

[jknilinux]

Well, even if they're not metadata, they still illustrate the problem with the metadata idea in evolution, being that you require two randomly inserted lines of code to be after each other in a specific order for it to work.

[Peter]

Well, even if they're not metadata, they still illustrate the problem with the metadata idea in evolution, being that you require two randomly inserted lines of code to be after each other in a specific order for it to work.

You seem to miss the point. Metadata is not affected by evolution. Metadata can only be inserted by the creator(a human) and it is usefull for have in some bots. This way bots have a amount of genes that aren't affected by mutations.

The discussion about metadata or no metadata was pretty much about, ''may this protection evolve inside a bot or not''.

[jknilinux]

But when Numsgil first brought up metadata, he didn't mean for it to be impossible to mutate- did he?

I like the idea of having the amount of mutation protection given also affect the instruction, like in my example.

"

1: 10 .protect store 'protects self and next 9 bps
...
10: 100 .mkprotenzyme store 'The protection is 100%


"

So, the instruction can act like metadata, in that it is never mutated, if that's why you want metadata. You should also be able to turn off mutation protection costs.

[Peter]

That doesn't take into account the data that comes into the dna. It can't get out that way, but it can get in. And with metadata it can't be inserted.

I really haven't got a powerful opininion about this. Can't we just wait with doing one or the other. There are better suggestions that will do more good for DB in my opinion.
Anyway I suggest the priority here is set low.

Anyway can sharv explain why evolving protection in dna could be bad. (apart from, it doesn't happen mich in nature)