See http://stackoverflow.com/questions/3107299/how-do-clojure-pr...
But there are also interpreted interpreters aren't there? Which aren't real executables, and aren't native. It would be possible to write an interpreted interpreter for Perl, maybe in a language like Ruby. Jython is a real example of an interpreted interpreter (if you ignore that the JVM has a JIT, but it isn't AOT, which you've said that you think is an important criteria).
And so it isn't useless information.
The distinction is particularly relevant here because the RPython technology they are using to build their interpreter means they can either interpret their interpreter using the Python interpreter, or they can make a native interpreter by compiling their interpreter to native AOT. That's probably why they used that particular wording.
So, again, not only is their terminology consistent with the rest of the industry, they are also making a specific and interesting point here, and it isn't useless information.
Sorry, no, this is nonsense. Scala source code is compiled to Java bytecode, which is interpreted and JITed by the JVM. Pixie source code is compiled to the Pixie bytecode which is interpreted and JITed by the Pixie VM. chrisseaton claims that "native" refers to the interpreter, but this is rubbish ... that's not what the industry means by "native".
I mean, not in use.... What would be the point? Every other interpreter worth using is AOT compiled, it's hardly something to boast about.
One exception might be JRuby, I suppose. Kind of a strange comparison to go out of your way to disabuse.
Pixie programs are translated into Pixie bytecode, which runs on the Pixie VM. Bytecode is, by definition, not "native".
IIRC with RPython/pypy the two techniques are interleaved, but most users I've seen opt for some JIT optimization.
Sure, you can for example run a Prolog interpreter on top of a Lisp interpreter. It's just not fast, but may have other qualities.