I keep looking for a good excuse to learn it, but it does not fit my domain or my language-style preferences very well.

I primarily develop web services, and it is really difficult to beat Typescript due to the fact I can write my front-end, backend, and mobile apps in it (via NativeScript or React Native).

I use uWebsockets.js, which has both an HTTP API that is nearly-identical to Express, and a Websocket API that is identical to Socket.io, but performance ~x10 faster than Express and ~x2 faster than the top-performing Go/Rust/C web libraries due to being a C++ library exposed as Node V8 bindings.

If I really need performance outside the context of handling web requests, I will write small functions or services in Go/Rust and compile to WASM and invoke it through Node, or just throw it up on OpenFaaS.

I like a lot of what Nim does, but I have had a difficult time since finding it two years ago ever coming up with an adequate reason for using it.

Edit: I want to make one clarification. I think that Nim may actually have a bright future in ML, primarily due to the work of Mamy Ratsimbazafy. He has some libraries for HPC, Laser [0] and Arraymancer [1] that can smoke C by several orders of magnitude. Combined with Weave [2], the multi-threaded runtime he built for Nim, that opens the doors for ridiculously performant ML. I think the only the other language that would give Nim a run for its money here is Julia, I really dig the work that has been going on there for ML as well.

[0] https://github.com/numforge/laser

[1] https://github.com/mratsim/Arraymancer

[2] https://github.com/mratsim/weave

Well, Nim compiles to JS and cross-compilation to iOS and Android are both possible (I don't know how tricky those last two are).

So you could write your entire stack in Nim too the same way you do with TS. :)

Smoke C by several orders of magnitude? Citation needed.

Entirely reasonable doubt, you can find well-commented benchmark output of the Laser High-Performance Computing lib and Arraymancer lib in their source:

https://github.com/numforge/laser/blob/d1e6ae6106564bfb350d4...

https://github.com/mratsim/Arraymancer/blob/6fddfa9a734ac01c...

To qualify the statement, it would be more accurate to say that Nim can smoke C by several orders of magnitude unless you devote significant engineering efforts into replicating the equivalent Nim (as Nim targets C/C++ regardless, so this is just heavy lifting and optimizations)

While all you say is true, "on order of magnitude" mean 10x and "several orders of magnitude" mean 1000x or more.

I think it needs a "killer library". The core language is pretty easy to learn if you've used Python, but there aren't a ton of third party libraries, and the ones that do exist seem to be incomplete, poorly documented, buggy, abandoned, or some combination thereof.

I got annoyed when it didn’t have a decent standard library to handle command line flags and arguments.

The examples given so far, are mediocre.

parseopt is incredibly cumbersome.

docopt is an incredibly poor design, that I can’t believe anyone thought this was a good idea.

And failing to have such basic libraries, I realized, this language has a long way to go.

I contemplated building my own library to handle such a need, but why, this doesn’t solve my problem of building something. And if it’s missing such basic libraries, then what else is it missing? The odds were too great, and too risky for me to commit my time to it.

There are many libraries for that, but I suggest cligen. It is the simplest command line parser I have found across all programming languages.

What if the Nim maintainers made some kind of continuous popularity measurement to let good libraries bubble up? Opens other problems I know, but may fix such scenario and accelerate Nim adoption?

There was an attempt a few months ago, but it was a little weird. But if you just want to find libraries relevant for your task (e.g. parsing the command line), there is https://nimble.directory where all Nim libraries are listed

I hadn't seen docopt before, but am curious why you think it's such a bad design.

you can try cligen and then argparse https://github.com/iffy/nim-argparse

making clones of the top 50 python libraries would make it attractive. familiar syntax! way more speed!

Many of the top 50 python libraries are wrapping a lot of code that is actually c, c++ or fortran. So I doubt that there is a lot of performance to be gained. Numpy is a good examples for that.

I have applied for a job at a company where they were switching from Python to Rust, because even though they were using Numpy, there was a lot of overhead in setting the data up to send to Numpy.

It's not those libraries' performances, but allowing people to port their own code with less friction for a presumed gain in performance.

How would one find these "top 50 python libraries" list?

I’d suggest based on PyPI download numbers, see https://hugovk.github.io/top-pypi-packages/

https://libraries.io/ ?

This would be super compelling.

I'll place my chips on mratism (Arraymancer, Weave, Laser) being the killer librarian. But it will take a decade or two to unseat Python.