Nobody is saying GCHQ are idiots and I fail to see why you mention them.
This is not about some "toy encryption schemes" it is a observation that as this law stands it there is no real way to say what is random and what is encrypted or in the case I point out - compressed. Now the whole argument of making the data intelligible is a completely different argument and gets back to how do you prove random data is actualy just that. You can't.
Good encyption with have entropy akin to random data. Also a compressed file will have the entropy of poorly encypted data.
Data is just that, data. Intelligble data is information and is not data. Big difference and in that any data set is random without meaning/interpritation.
> Now question is - compression can be views as encryption. How does that pan out if you use a non-standard form of compression that does not require a key as the compression formula is the key in itself!
You then ask why I mention GCHQ. I mention GCHQ because they control NTAC (National Technical Assistance Centre) - this is who will attempt to decrypt the data. This will happen in parallel to RIPA notices being issued.
If a person uses a non-standard form of compression and the police are interested there are two actions from police:
1) GCHQ trivially 'break the crypto'
2) A RIPA notice to make the data intelligible is issued, forcing the user to un-compress the data.
> Now the whole argument of making the data intelligible is a completely different argument
No, it really isn't. If you've encrypted it or compressed it or used steganography or used some simple code system to hide data they issue a notice and you have a limited amount of time to make the data intelligible.
> and gets back to how do you prove random data is actualy just that. You can't.
This is a different argument, and is not what you said.
This is not about some "toy encryption schemes" it is a observation that as this law stands it there is no real way to say what is random and what is encrypted or in the case I point out - compressed. Now the whole argument of making the data intelligible is a completely different argument and gets back to how do you prove random data is actualy just that. You can't.
Good encyption with have entropy akin to random data. Also a compressed file will have the entropy of poorly encypted data.