Dependabot neither supports pnpm nor yarn:
https://github.com/dependabot/...
https://github.com/dependabot/...

The intention from GitHub is clear, Microsoft acquired npm and the fancy new supply-chain-security is just a lousy way of walling people inside the ecosystem.

GitHub is great, github.dev is amazing, VS Code is sick. But no, this one guy of Isaac Schlueter makes me hate this whole supply chain.

pnpm, renovatebot and GitLab: I choose you!

Today I tried: pnpm.

Following up my hateful rant against Isaac Schlueter and his decisions on npm.

I went all the way out and tried so many alternatives, honestly found a developer experience much greater than the "official" one.

What triggers me the most is the explicit statement of "other creative means" in this commit https://github.com/npm/arborist/..., you don't talk like that when all other package managers are making creative workarounds for the design failure of node_modules.

I don't know what it is but I really hate this guy.

When I reach my short term milestone, I might start dedicating my full-time to one of the following projects:

1. deno
2. brain.js
3. The forking of TypeORM
4. The teardown of Isaac Schlueter
5. Create a full on movement to rescue open source

I believe there is a sizable community suffering the current situation of TypeORM, a blackmailing author without introducing new members.

Why isn't there a succeeding fork yet? We have Sequal Ace now, people can do it!

Ok this is my first try in machine learning, this is DQN for Pong!

I want to add my own flavor by adding pressure touch to paddle movements, such that the harder I press in a direction it moves faster to a max speed.

Does it makes sense if I apply the floating point outputs of sigmoid directly to speed control? Or should I make multiple outputs to represent different "steps" of pressure/speed?

From the look of https://github.com/yarnpkg/berry/..., Microsoft is not (yet) planning to hug "npm audit" as a great evil plan of asserting dominance to the open source community by raising everything into NatSec level and force shortcut releases.

If that's the case alternatives like yarn and pnpm will be removed from the scene, VS Code will be intentionally made incompatible with Yarn's PNP just like how NPM sneaks https://github.com/npm/arborist/... through, under the name of security.

I am still not convinced, it is Microsoft after all. We'll see.

P.S. I will laugh menacingly if that turns out to be ONLY a stupid dream and a poor decision of one single genius businessman.

It takes courage to use npm as a product. The inability of its leader, Isaac Schlueter, to communicate the reasoning behind decisions pose a risk as a choice for long term toolchain.

My company will move to yarn for now and jump ship to Deno as soon as it reaches all of my check marks.

Joke: npm audit

https://github.com/lodash/lodash/...

The genius businessman Isaacs Schlueter, who sold a common tool npm to GitHub, forced a breaking change to block yarn in a minor upgrade in August.

Every decision of trading his users for his own business growth is so obvious, so hostile and so badly executed.

Let us give a moment of silent to the damage he's done to the community, I still see comma-first and colon-skipping as an intentional habit in some open source libraries.

This very commit https://github.com/npm/arborist/... breaks so many things at once.

How to do typically avoid the necessity to rebase a WIP branch after every other merge on the default branch?

The webpack production mode can't even recognize destructuring. It can only naively replace literal occurrence of process.env.NODE_ENV.

What is this? A f'ing template engine?

So `aws-sdk@^2` happens to include an ancient polyfill `url@0.10.3` which breaks webpack, specifically fork-ts-checker-webpack-plugin, causing it to warn about the inexistence of class "URL".

And this is the default included package for AWS Lambda node.js runtime.

wow, just wow.

Before I dip my toes on machine learning, let me leave some silly comments so I can laugh at myself in the future.

Let's make geth.

1. The model will spit out layer definitions and the size of sample data for training, children models are trained with limited computational resources.
2. Child models are voters that only response in terms of yes/no. A simple majority wins and then the action is taken.
3. The only goal for master models are to survive. i.e. To prevent me from killing them.

Questions:
1. How do models respond to a random output size? (Study GPT-3, should take weeks/months but worth it.)
2. How to define actions for voters to vote? Sounds like the boundary between actions should be blurry and votes can be changed from tick to tick (i.e. responding to something in a split second). Therefore
3. Why I haven't seen this yet? Is this design a stupidly complex way of achieving the same thing done by a simple neural network?

I am full of curiosity and stupidness.

Is devRant pay to win? I don't think it's possible to organically get +20k ever

The cake is a lie

Ignorance is bliss or ignorance is guilt?

The overwhelming code smell when you finally got a chance to do a quarterly project-level code review as a team lead.

It's my fault for feeling like they are teaming up against me, and I don't know how to fix it.

npm install
iCloud Sync goes boom
webpack
iCloud Sync goes boom
git anything
iCloud Sync goes boom

Why does every other tool in npm wants to litter their own stupid configuration file at my project root?

Have to work on Windows for a DLL integration project, how does Microsoft managed to build something so disgusting to human eyes for all these years?

Dark mode in Visual Studio Installer is actually pitch black, hello? Common sense? Basic human-computer interfacing? Basic color scheming?

All these installations, frameworks, version management...

Why am I dipping myself into this bloatware shitshow?

gqless just woke up from a long sleep

Took me by surprise that Google Docs apps are using Vue, but then thinking it as the better Angular makes a lot of sense.

I am a React fanboi.

The struggle of startup founders

You: Ask the right question at the right moment
Also you: I don't care just make it happen

Someone in GitHub made a good pose.

A paid solution like AppSync should be complete and does not require hacking, using the AppSync JavaScript client is just a pain in the ass.

Someone posted a fix to a 5 years old problem that Docker as a company thrown into dumpster (moby) for years.

From the README there it's well researched and it seems they know what they're doing.

The whole daemon is one single file with only 300 loc, totally manageable for everyone if they want to scrutinize.

https://github.com/moby/moby/...

Where would you draw the line between knowledgeable and annoying?

NEVER maintain multiple release branches as a maintainer, period.

Given enough PRs, nobody, not even god-tier maintainers, is able to keep track of all commits and the whole context/story behind it.

Remember that people ALWAYS PR against your master and nobody even cares about your whatever-named-next-branch buried deep inside your feature branches.

Please make it easy for others if you want actual contributions.

master -> main, classic Euphemism Treadmill.

So, entropic has been up for a year now. Anyone walking on their feet now?

The talk: https://youtu.be/MO8hZlgK5zc

The bipolar nature of leading a startup tech team:

It's either no one needs you, or the problem is big enough to have you push 5 fundamental fixes from abyss to upstream.