This need to rewrite, of course, is what Julia is trying to avoid. My workflow is exactly the same, and I’d love to be able to write code in a high-level language like Python and then use that directly instead of having to rewrite.

However, in my case the reason for rewriting isn’t just performance, but also to be able to build compiled binaries. Julia aims to be as high-level as Python but faster - is there a language that’s as high-level as Python but AOT-compiled?

Common Lisp, Ocaml for example.

The “need to rewrite” is actually a sort of advantage with Cython. You only target small pieces of your program to be compiled to C or C++ for optimization, and the rest where runtime is already fast enough or otherwise doesn’t matter, you seamlessly write in plain Python.

Using extension modules is just a time-tested, highly organized, modular, robust design pattern.

Julia and others do themselves a disservice by trying to make “the whole language automatically optimized” which counter-intuitively is worse than make the language overall optimized for flexibility instead of speed, yet with an easy system to patch optimization modules anywhere they are needed.

The problem with cython is that to really get the performance benefits your code looks almost like C.

I agree with you on the optimize the bits that matter, often the performance critical parts are very small fractions of the overall code base.

I really don't get this. I'am fully on the side that limitations may increase design quality. E.g I accept the argument that Haskell immutability often leads to good design, I also believe the same true for Rust ownership rules (it often forces a design where components have a well defined responsibility: this component only manages resource X starting from { until }.)

But having a performance boundary between components, why would that help?

E.g. This algorithm will be fast with floats but will be slow with complex numbers. Or: You can provide X,Y as callback function to our component, it will be blessed and fast, but providing your custom function Z it will be slow.

So you should implement support for callback Z in a different layer but not for callback X,Y, and you should rewrite your algorithm in a lower level layer just to support complex numbers. Will this really lead to a better design?

It helps precisely so you don’t pay premature abstraction costs to over-generalize the performance patterns.

One of my biggest complaints with Julia is that zealots for the language insist these permeating abstractions are costless, but they totally aren’t. Sometimes I’m way better off if not everything up the entire language stack is differentiable and carries baggage with it needed for that underlying architecture. But Julia hasn’t given me the choice of this little piece that does benefit from it vs that little piece that, by virtue of being built on top of the same differentiability, is just bloat or premature optimization.

> “you should rewrite your algorithm in a lower level layer just to support complex numbers.”

Yes, precisely. This maximally avoids premature abstraction and premature extensibility. And if, like in Cython, the process of “rewriting” the algorithm is essentially instantaneous, easy, pleasant to work with, then the cost is even lower.

This is why you have such a spectrum in Python.

1. Create restricted computation domains (eg numpy API, pandas API, tensorflow API)

2. Allow each to pursue optimization independently, with clear boundaries and API constraints if you want to hook in

3. When possible, automate large classes of transpilation from outside the separate restricted computation domains to inside them (eg JITs like numba), but never seek a pan-everything JIT that destroys the clear boundaries

4. For everything else (eg cases where you deliberately don’t want a JIT auto-optimizing because you need to restrict the scope or you need finer control), use Cython and write your Python modules seamlessly with some optimization-targeting patches in C/C++ and the rest in just normal, easy to use Python.

This sounds like it might be interesting, but your later comments about overhead and abstraction costs sounds like you maybe don't understand what Julia's JIT is actually doing and how it leverages multiple dispatch and unboxing. Could you be a bit more concrete?