Obviousness is not a subjective thing. Without prove of it's obviousness nothing is obvious.
If an invention is new, meaning there are differences to a single piece of closest prior art, the entitiy alleging obviousness needs to find prove that the differentiating element was known and known or suspected to be combinable with the piece of prior art. Thats needs to be proven. It cannot be proven without a document or (really) common knowledge. But common knowledge in this case would be "a car can have 4 wheels", not "I can use algorithm x to do y".
And therefore, nothing is obvious unless the prior art teaches all of it's limitations in as few documents as possible (in most cases 2) from at least neighbouring fields.
But that doesn't answer your question, does it? That's because that is the part about combining references, but the references are combined and/or read by a person skilled in the art. And wouldn't that person skilled in the art know a little more than "a car has 4 wheels"? Maybe which algorithm to use to achieve thing y?
Yes! But! As I said, obviousness needs to be proven. You need to prove that the person skilled in the art would know that.
If you find a standard textbook teaching the missing limitation, the invention was obvious for the person skilled in the art.
Which is the answer to your question. Anything that can be found in standard literature for the field is obvious. But that needs to be proven. And proving that would require a reference (even though some standard textbook or something similar), which would actually be "missed prior art". Therefore I stand by my answer: "Nothing is obvious without prior art teaching it."