Hacker Newsnew | past | comments | ask | show | jobs | submit | amichail's commentslogin

For a new kind of snake game, check out PluriSnake.

PluriSnake is a snake-based color matching daily puzzle game.

Color matching is used in two ways: (1) matching circles creates snakes, and (2) matching a snake’s color with the squares beneath it destroys them.

Snakes, but not individual circles, can be moved by snaking to squares of matching color, as long as their paths are not blocked by other snakes.

The goal is to score as highly as you can. Destroying all the squares is not required for your score to count.

Of course, there is more to it than that as you will see.

Try it out:

https://testflight.apple.com/join/mJXdJavG

Any feedback would be appreciated. Have fun!


I tried it, but really shouldn't have as I am colorblind and so matches were quite difficult. Some color games have an option for a colorblind-safe palette, maybe consider that?

The iPhone has built-in color filters. Have you tried using one of them?

Sure, I have them turned on for most of my devices.

Colorblindness isn’t exactly fixed by filters, for a few reasons. They help for some colors, but not others. And each person’s experience is different based on their specific deficiencies.

Anyway, this is a color-based game and that’s probably what it needs to be (unless size or shape could vary enough to substitute), so I’m not in the target audience.


PluriSnake is a snake-based color matching daily puzzle game.

Color matching is used in two ways: (1) matching circles creates snakes, and (2) matching a snake’s color with the squares beneath it destroys them.

Snakes, but not individual circles, can be moved by snaking to squares of matching color, as long as their paths are not blocked by other snakes.

The goal is to score as highly as you can. Destroying all the squares is not required for your score to count.

Of course, there is more to it than that as you will see.

Try it out:

https://testflight.apple.com/join/mJXdJavG

Any feedback would be appreciated. Have fun!




So it's neither ignorant nor reluctant to tell you about vibe coding.


It didn't know what it was at the beginning of the conversation.


When you say:

"Should schools have a chess variant class where students invent and vibe code novel chess variants and play each other's variants?"

I don't know what you are saying. Sounds more like your objective is to have students brainstorm new ways to play chess. Why is there any indication that the LLM should assume this is a programming question?

In fact, it has divined that programming might be involved and offered that certain python libraries might be leveraged. You frame your input as a question about schools offering a certain class and the LLM responded by assuming that is what your objective was and started to help build such a curriculum.

Did you expect with your input that the LLM would simply jump into writing some kind of code? Should it have written a multi user chess playground? Should it have written a python chess variant? I personally think it did the reasonable thing and from the context you provided assumed your task was to build such a class pedagogically not produce some python code.

If you are focused on chatGPT lumping other forms of "hacking" or "cowboy coding" into the same bucket as "LLM Assist" and calling it all "Vibe Coding" I personally don't have an issue with that. They all fall into the bucket of fast and loose coding techniques. That is just my take on it though.


To an LLM all coding it does is vibe coding.

In this case you’re asking it a closed question with a {yes, no, maybe} answer set.

It tries its hardest to give you more than that.

You’re asking a machine to have an opinion. If someone asked you that question, what would your answer be?

Think about your goals and state them more clearly.

“What are the pros and cons of …”


If you submit a prompt with "vibe coding" in it, they will reply in a way that seems to imply they have no idea what the term means and are just faking a reply based on surrounding context in the prompt.


"they" - which ones ?


ChatGPT for one.


Share a screenshot of what you're seeing because it seems like everyone else is seeing something radically different.



Request is not allowed…


I just asked it about vibe coding, and it happily generated numerous paragraphs and bulleted list on the subject.


That is not my experience, at least when I tried it just now.


I've noticed a few drops on the surface of my m3 MacBook Pro after going indoors following a 30 minute walk in the cold.


This one involves a grudge against one human.


So maybe the training data has a lot of old English writing and overcoming the model's tendency to use em dashes everywhere with custom instructions would use up more electricity.


This is for passengers so they don't get bored during the drive. It's a virtual drone not a real drone.


What's wrong with looking out the window? I'm also sure that passengers will have their own devices so they probably won't get bored, even on long trips.


You never see the numbers. The idea is to transform the partial order into a total order (shown as vertical chain of nodes).


You still haven't described fully what the user sees before and after a step.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: