191–200 of 309 posts
Until this is resolved stay miles away from anything Andrey Sitnik has ever developed.
The colorette's creator, Jorge, tried to dcma other repos on GitHub: https://news.ycombinator.com/item?id=10271304
Use his libraries at your own risk.
Earlier quoted context omitted.
Colorette author here. I put the benchmarks in a CI workflow so anyone can check. Click Benchmarks to see the results: https://github.com/jorgebucaran/color-bench-test/runs/371243...
Wonder how “ai” would feel if you just forked postcss without attribution and opened PRs against popular projects to replace. Something tells me they wouldn’t be so calm about the whole affair
It looks like nanocolors is a fork of colorette with a few (substantial?) changes, see https://github.com/ai/nanocolors/wiki/Colorette-Changes . However, the author of nanocolors, Andrey Sitnik, only added attribution info about colorette after Jorge Bucaran went public about the fork on Twitter. If the fork had attributed colorette from the start, would there be drama here?
Colorette author here. I put the benchmarks in a CI workflow so anyone can check. Click Benchmarks to see the results: https://github.com/jorgebucaran/color-bench-test/runs/371243...
https://github.com/jorgebucaran/colorette/issues/78
Just look at performance changes history,
Another side of this story https://twitter.com/sitnikcode/status/1441824191624486924 Also at the end Colorette’s author back-ported Nano Colors performance optimizations, but this time he refuse to mention another developer. https://github.com/jorgebucaran/colorette/issues/78
Earlier quoted context omitted.
Colorette author here. I put the benchmarks in a CI workflow so anyone can check. Click Benchmarks to see the results: https://github.com/jorgebucaran/color-bench-test/runs/371243...
You have a new benchmark results after you backported Nano Colors optimizations and refused to mentioned it https://github.com/jorgebucaran/colorette/issues/78 Just look at performance changes history,
What is even going on here? https://github.com/jorgebucaran/colorette/issues/78 . They're arguing about who should have attribution for adding a single "!s &&" to improve performance? What happened with the world where open-source was about creating cool projects together , not about who has the most stars on GitHub? To be honest, this is the general feeling I get from the JavaScript community: Instead of working tog…
It's the Instagram-ification of the open source world. All of the feedback and social signaling mechanisms are there. Likes, stars, followers, and charts of "contributions". It's just an extension of LinkedIn for many people.
Let's not forget the DigitalOcean Hacktoberfest debacle a few years ago. I still stumble upon repos with pull requests that contain nothing more than a single character change to a README. Just for a stupid SWAG t-shirt.
What I find absolutely baffling about all this is that it could have been avoided by adding a throwaway line about how this library is a fork of the other one with performance in mind. A lot of drama can often be avoided if people just communicate better.
The way that git history was removed makes me wonder if they preferred to claim as their own original work. Unless erasing the history was done by accident. This doesn’t feel like a communication issue as much as an ethical question.
It couldn't have been. Forking 'properly' on GitHub is a single click. Doing what they did required cloning, erasing .git, then creating a new repo. The original history (now replaced with colorette's) even had an 'initial commit' by ai.
This was 100% malicious.
Another side of this story https://twitter.com/sitnikcode/status/1441824191624486924 Also at the end Colorette’s author back-ported Nano Colors performance optimizations, but this time he refuse to mention another developer. https://github.com/jorgebucaran/colorette/issues/78
Wow, that second link is incredibly petty. makes ai look even worse.
And he incorporated it into his benchmark to show that colorette is as fast as nanocolor.
It sounds like an important optimization.
If this wasn't copied, colorette would have been objectively worse than nanocolor.
Earlier quoted context omitted.
Javascript is the opposite of a batteries-included language, so there is and has been an enormous amount of wasted effort duplicating functionality that should have been in a standard library.
We’re talking about a terminal colorization function though. Fragmentation of packages aside, are there any other languages that have terminal text colorization built into the standard library?
https://docs.microsoft.com/en-us/dotnet/api/system.console.f...
It's even easier on UNIX-like terminals, where ANSI escape codes should work.
In all likelihood you have each of them copied a dozen times in your node_modules for some reason, all different versions and out of date (scary!) and they all give you vulnerability warnings when you run npm audit.
(Being a former front-end engineer writing Java for the last couple years, reading anything in the modern JS world is mind boggling- just my lived experienced)