I agree wholeheartedly with all of the points made in the article, but this website kinda sucks as an example.
- The website is "one binary", but the deploy strategy involves compiling it instead of running a binary artifact
- Won't rely on github pages, but depends on external resource at openlibrary.org
- Writing your own stuff lets you take advantage of open standards, but the .html files in thoughts/ are structured as plaintext with a false file extension
- When people go this route, they usually try to cram everything into the initial page load. This has several static files served dynamically.
- The Golang code does not cache any responses, and does not store templates in memory.
- Several dependencies in the go.mod file, all of them seem unused?
Most importantly:
No discussion about the technical benefits of the "single-binary" ethos compared to modern infrastructure.
It's my humble opinion that it's a great idea, and this is a poor example.
The code uses `//go:embed` which actually embeds a fake filesystem into the binary. It’s really easy to miss that part, but it truly is one single binary, despite seeming to reference files by paths.
I think you're right that the world of executives is less like "employment" and it's kind of insulting to pretend it is.
But I also think it's fine for executives to sue the companies they represent. I think the government should have a role in corporate governance. I would even support having a government appointed representative like the Federal Reserve has.
The replies are missing the point, I agree. “Executives” are part of the bourgeoisie, employment laws bitterly won by the organised proletariat aren’t really relevant.
At the same time, people can sue each other based on laws, why not? If you don’t like that, you should join those working to overthrow the current system.
I think it’s also potentially useful, at least rhetorically, for gender discrimination to be an issue even among the ruling class.
These things are made of neural networks. Literally everything about them is weights and biases.
I agree and all, but it's weird to claim these models have general bias without testing them on a variety of inputs. These models have a lot of minute details. They're capable of differentiating a lot of specific things. They don't lack information about Indian women.
The industry agreed to standardize on Python for the task of describing compute-graphs that get executed by compute engines implemented in something other than Python. Python is not meant to be used for the computation itself.
And then other people came along, using other languages, and realised we could do the same compute, with a fraction of the hardware, in about the same time, in the languages they’re already using, which really diminishes the appeal of Python, which boils down to:
“Setup some infra, and get someone who knows how to operate it, and then write new code in python, and babysit it as it ossifies into a core piece of infrastructure”
_or_ optimise it in a language you’re already using, or is semi-common in your stack already, and then move on with things.
This should only affect C# repositories, and we're working on a fix - C# support is relatively new, and looks like there are some teething issues with the caching mechanism we use to run tests (we aim to keep median response times <3s).
I get your point, and I agree with it for most cryptocurrencies.
In this case, the proof-of-work would be peer-review instead of computation, and the "blocks" would be as frequent as "minor releases". The need for computation would be pretty light.
As it is, I go on npmjs.com, look at the number of downloads, and hope for the best. Then ‘npm audit’ runs automatically, spitting out anywhere from a few dozen to several hundred warnings which I then ignore, still hoping for the best.
If you can't find the kind you're looking for, you can make one!