INim VS mosdepth

- `nim secret`[0]: code is interpreted with Nim's vm, it is limited to compile-time (e.g. no C/C++).

- `nlvm r`[1]: nlvm backend supports JIT compilation and repl-like interface was added in latest release.

- `inim`[2]: supports all nim code, but it's not really a true repl. It adds code to a file and recompiles it (you can see the source file with `ctrl+x`). I'd recommend to use it with clang compiler, because it's a bit faster than gcc in my experience. It's the best option right now. And I use it almost daily.

First two options have rudimentary input system. But you can get command history and left-right navigation by wrapping them with rlwrap (should be preinstalled on most *nixes): `rlwrap nim secret`, `rlwrap nlvm r`.

There are plans to support incremental compilation for the next Nim release (it is currently broken). That would improve inim and other repls experience significantly.

[0]: https://nim-lang.org/docs/manual.html#restrictions-on-compil...

[1]: https://forum.nim-lang.org/t/10697

[2]: https://github.com/inim-repl/INim

Go must be terrible to google too! Nim's easy to google though, it's how I got to rust really. I'm actually the maintainer of inim, the nim repl: https://github.com/inim-repl/INim

I'm not sure why that happens, but an alternative to bedtools genomecov is mosdepth. Maybe you can try that tool instead as it's very reliable.

I would give mosdepth a look. It can provide depth across a chromosome/contig at each base, or a mean depth for a given window size (eg 1kb), or regions provided within a bed file, or a simple summary of the mean depth across each chromosome/contig.