[cyberpunk]

Why would that make sense?

The gains you get for free using otp and the beam vm depend heavily around language integration of vm features. Porting python to a vm it would work slowly on and would be unable to use most of, if not all of, the benefits without changing the language would give you what exactly?

[thecity2]

Have you heard of Mochi?

"Mochi is a dynamically typed programming language for functional programming and actor-style programming. Its interpreter is written in Python3. The interpreter translates a program written in Mochi to Python3's AST / bytecode."

https://github.com/i2y/mochi

If all that can be done in Python 3, surely it could be ported to run on BEAM to good effect.

[35bge57dtjku]

No more GIL problems?

What couldn't you do, say by making BEAM/OTP -specific features be a library you import in Python, and by mapping Python language features to however Erlang does them?

I get there will likely be problems, but what specific problems would you run in to?

[cyberpunk]

I'm not going to get too into this as I hope most readers understand why this couldn't work without needing much depth....

Python has no concept of a process like Erlang does. The standard library does even know about, let alone is able to deal with, the difference between cast and call messages. It can't do messages even. There is no way to translate monitors vs links. What's a node? .....

You could implement python syntax maybe, but python will never make sense on beam.

[toast0]

> Python has no concept of a process like Erlang does. The standard library does even know about, let alone is able to deal with, the difference between cast and call messages. It can't do messages even. There is no way to translate monitors vs links. What's a node?

I don't know that Python on the beam is a good idea / worth the work, but I don't think there's a fundamental issue here:

A process is more or less a thread, which Python already knows about.

Cast is just sending a message, which is erlang:msg(Dest, Msg); call is send a message plus selective receive. General receive is easy, call a function (not already available) and return the message. Selective receive is trickier, because python doesn't have a construct for matching like Erlang, so you'd have to figure out how to specify the possible message signatures you'd like to receive; plus extra work to get the make_ref mailbox marking optimization.

Monitors and links are just calling functions, and maybe being killed or getting specific messages later -- it's a VM feature, not a language feature. Same with a node.

You would have to make some hard decisions about any pythonic state though. It would be more functionally pure to pass that around as a function argument, but it might feel more like python if it was stuffed into the process dictionary. Python state shared between multiple processes/threads would be forbidden, of course. Mutability would also need some soul searching to fit into Beam's garbage collection.

EDIT to add: of course, you would need to restrict I/O to message passing -- anything with file descriptors should be rewritten to be message passing with beam ports.

[cyberpunk]

Well, I was a bit dishonest there and you've called it.

Monitor/Link and Cast/Call are all built ontop of the regular message passing functionality of the language. The point I 'whitehat fibbed' to make stands though, python might be feasable as a syntax implementation but there's a _MASSIVE_ difference in the architecture of helloworld as a flask app and the same under cowboy; trying to get one to behave as the other makes no sense to me.

Implementing a python on beam makes no sense as to actually benefit from beam you'd need to write a lot of libraries to actually play with the environment and why would anyone bother?

And sorry but no, processes are not more or less threads. My laptop can do this:

     3> processes:max(250000).
     Max allowed processes: 262144
     Process spawn time=3.408 (15.644) microseconds
     Process teardown time=3.156 (1.552) microseconds

That created 250k procs then stored the pids of each in a list, then iterated though that list and killed those processes; I don't think you can do that with threads..

Preemptive scheduling is a big deal also.

[toast0]

Yes, I know processes are different than threads, but the basic idea is similar -- a process is an execution context with its own stack and program counter and etc; so is a thread, but the etc differs. That python knows how to run in threads means it's at least got some of the basics that might enable it to run in multiple contexts, was my only point.

I agree that it would be a lot different architecture than a normal python application; but where there could be value, is if you wanted to write a distributed thing in python with all the nice properties you get from an erlang system, your choices now are either drop python (my preference!); write dist/otp yourself, but in python (hard, because how do you bolt on monitors, etc); write just enough erlang to coordinate python running as port commands (probably doable, but could be fragile).

I personally would love to run a little bit of perl on the beam, but i know it's crazy, so I don't mention it very often ;)

[saurik]

The "etc" is absolutely critical as it is shared memory: BEAM works the way it does specifically because it does not want to model shared memory, which means any implementation of threads over it will be an incredibly painful and certainly ludicrously slow simulation, and there is nothing in Python that really maps to the primitive BEAM offers.

[35bge57dtjku]

You can't accomplish all of that by making users import a library to use those things, like I said? Or at least enough of it to make it reasonably useful as a Python?

[cyberpunk]

Not really. Erlang/OTP applications get the advantages they have by being what they are and that's not really something you can import as a library to a language which doesn't even have primitives those systems speak.

If it sounds like a good idea to do that when you are solving something horrible, then your better bet is to write that functionality in a beam language instead of trying to play beam in a language which can't do it.

This is a bit of a different world than most of us are used to. Porting python to the jvm or making it run on v8 are different goals than making something play in the beam space.

You could run single node python with a port to beam, but there would be no advantages to doing it. It would be slower and no one would use it. Beam langauges only get you something when you're doing distributed things really and python isn't equipped for playing those natively.

You could wrap it in some libs maybe, sure. But why would you?

[tormeh]

BEAM is for distributed systems and MySQL is for databases. Porting Python to run on BEAM like Erlang does is like porting Python to run on MySQL like SQL does: Possible, but slightly surrealistic.