snicker7

Most newer languages will give you multiple orders of magnitude better performance than python.

Python’s main advantage was that it was easier than some of its competitors (C++/Java). But that is no longer the case with modern languages (Nim/Crystal/Julia/JavaScript) being both faster and comparably as easy (or easier).

It is now coasting off its momentum, mostly do to the vast amount of (usually poorly designed) open source libraries. That and Jupyter.

Yachts are expensive. And it’s literally impossible to live in Cali without one. What would the neighbors think?

I see a literal zero percent chance that the Dems will pass a tax bill this Congress.

NASA switched from SIMULINK to ModelingToolkit (Julia) for modeling spacecraft dynamics [0]. They found that the latter was much easier and gave a 15,000x (!!!) performance improvement.

[0]: https://www.youtube.com/watch?v=tQpqsmwlfY0

For emacs, I recommend the 'almost-mono-themes' package [0]. It supports syntax highlighting without making your editor look like Mardi Gras.

[0]: https://github.com/cryon/almost-mono-themes

Apple operates at a volume where investing thousands of person-years for small performance/revenue gains is worthwhile.

The demographics trends are interesting. It seems that Julia users ("Juliers?") are increasingly more young and more female-identifying.

Among other things, they are doing this to prevent second-hand smoking. I guess those carcinogens can travel through cyberspace.

Bill de Blasio should be made to provide accomodations for her in his own home.

Nix and Guix (my preference) are good alternatives. Guix, in particular, has its own system/application containers (not runc compatible). Highly efficient since they are essentially composed of symlinks of individual items in the Guix store.

I feel bad for all those Uyghur slaves toiling in Apple's supply chains. But if it makes Tim even slightly wealthier, it is all worth it. :D

Femtolisp is, by-the-way, a brilliant and performant implementation of Scheme. Wish it was used outside of Julia.

Bootstrappability is the reason why GNU Guix is one of the most ever important software projects. Moreover, I see Guix as the unifying force for the entire GNU project. It is the perfect playground to test and develop emerging tech (Shepherd, Hurd, &tc.).

The solution is simple: just elect government officials that (1) support the complete removal of Iran's nuclear capability (2) stop funding/training terrorists and insurgents abroad and (3) recognize the right for Israel to exist.

For Julia coders, there is Javis.jl

https://www.youtube.com/watch?v=ckvsc6ukdOc

This feels like something you'd find on linkedin.

Currently only supported by Intel.

A brutal admission that modern software is broken and that stability, predictability, and correctness is just too much to ask for.

MathML is the only accessible option. Screen readers work with it.

Tools like KaTeX output both HTML and MathML simultaneously (HTML for rendering, MathML for screen readers), leading to massive page bloat. MathJax requires client-side JS to look good, which is a big 'no' for me.

Right now, there is no way to display math on the web that: (1) is accessible, (2) works across all modern browsers, (3) is not slow or bloated in some way. MathML in Chrome fixes basically all these problems.

I looked at the source code for UnixMaps.jl. It seems to call :jl_mmap (exported by Julia's underlying C++ source code). I wonder why this style might be preferred to, say, calling libc directly?

> Not sure how exactly the lack of these skills could benefit anyone.

Lower the standards across all levels -- high school graduation, college admissions, job placements. Eventually we end up with surgeons and lawyers that are illiterate. But at least it is equitable!

Not everyone who works in NYC is a NYC resident. The entire state of NJ basically exists for people commuting to NYC. half of the entire state is basically highways and parking lots.

tl;dr -- Fox News is to blame for all of America's problem.

> It had. It's now around parity with pytorch.

In some cases, it is much faster.

Consider Neural Stochastic Differential Equations, Flux is literally over 70,000x faster than Google's PyTorch-based implementation:

https://gist.github.com/ChrisRackauckas/6a03e7b151c86b32d74b...

Julia code is (or can be) quite readable. But I think there are better examples. This package is not poorly written, but it has a lot of nested control flow (probably unnecessary).

Half-Cambodian. We have the same breakfast. We called it Bar Bar. The Chinese call it Congee. Other Asian countries have their own variants. I think even the Greeks have their own savory rice-porridge recipe (Avgolemeno).

It is good for literally any time of the day/year. Especially good for colds.

Use org-mode or die.

That's kind-of what Kafka is. See also write-ahead-logs.

I feel that we are converging with each passing day.

> I could never have done that when I was in an office. I couldn't take 20 min and play banjo or accordion and come back to my problem with a fresh eye

I mean ... you could.