Didn’t expect to see this on HN. Thanks for checking it out though. Not a developer, just a lawyer at GitHub, this repo was me playing around with the GitHub Spark tool.
It would be cool to encode the chess board state and turn into the URL so you could hurl urls back and forth over slack to play chess just by clicking on it
but there's something charming about the ascii art over slack in this project that would miss
making a move should automatically copy the new url to your clipboard. you can still keep the ascii charm by server side rendering the ascii chess board as an og description.
I thought it would at least read the slack websocket and send slack messages. Having to copy paste is horrible.
It's not very hard, you can take inspiration from my project "localslackirc", it even has a tool to extract the slack token from firefox's cookies (if you copy paste code, keep in mind it's GPL licensed).
Sometimes the best software is no software. You can do this without some emojified github readme, the text editor your messenger uses will be just fine.
(If you're not a beginner it's probably fine to just use "algebraic" notation anyway.)
I thought it was just me disliking this types of readmes. Is this aislop or are people actually wasting time to paste all those emojis everywhere? It is like we are going back to hieroglyphs -- I swear if I see another rocket launch...
It's LLMs. You can tell from the phrasing ("no X just Y", "isn't A it's B", "pure ___"), and because all of the sections except Screenshots say basically the same thing. A competent person would not write it like this and an incompetent person would not be this meticulous about grammar and formatting.
I realize this sounds exactly like "I can tell from some of the pixels and from seeing quite a few shops in my time" but I really can tell.
Edit: Also you can see this was made with https://github.com/features/spark, but that's cheating, it's more fun to make accusations based on vague feelings about the writing tone.
I wonder where LLMs got this style from, since, while you did see something like it, it wasn’t nearly as widespread before the rise of ChatGPT, so not the most statistically likely form for these to be generated in.
Was it trained into them in the supervised or reinforcement phases?
Has it emerged from prompts extorting them to be friendly, somehow sneaking in from text message training as a result?
Is it now in a self-reinforcing feedback loop as training data grows to include modern Readmes?
but there's something charming about the ascii art over slack in this project that would miss
making a move should automatically copy the new url to your clipboard. you can still keep the ascii charm by server side rendering the ascii chess board as an og description.
https://correspondence-chess-production.up.railway.app/
It's not very hard, you can take inspiration from my project "localslackirc", it even has a tool to extract the slack token from firefox's cookies (if you copy paste code, keep in mind it's GPL licensed).
(If you're not a beginner it's probably fine to just use "algebraic" notation anyway.)
I realize this sounds exactly like "I can tell from some of the pixels and from seeing quite a few shops in my time" but I really can tell.
Edit: Also you can see this was made with https://github.com/features/spark, but that's cheating, it's more fun to make accusations based on vague feelings about the writing tone.
Was it trained into them in the supervised or reinforcement phases?
Has it emerged from prompts extorting them to be friendly, somehow sneaking in from text message training as a result?
Is it now in a self-reinforcing feedback loop as training data grows to include modern Readmes?