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

DigitalOcean is such a dream to use. I also really appreciate all their guides for almost everything web server related.

How is their managed Kubernetes product nowadays? I've realized all I really use on GCP and AWS is managed Kubernetes and Postgres, and I feel like I must be overpaying particularly for GPU instances.

I have health insurance. I don't feel like I get anything out of paying for it, I still get massive bills whenever I use it.

That whole thing is so scammy that I totally see why people would rather self-diagnose.

Except the author did it wrong. You don't just ignore a huge rash that every online resource will say is lyme disease. If you really want to trust an LLM, at least prompt it a few different ways.


I personally don't find modals inherently all that bad, though they can definitely be implemented poorly. Does anyone have specific reading material on the problems with modals?

Accessibility doesn't even need to be related to any disability or unusual user requirement. A user-hostile website can be inaccessible even to users with perfect visual and motor functions.

What are Depot runners?

Founder of Depot here. We provide faster and more reliable GitHub Actions runners (as well as other build performance services) at half the cost of GitHub [0]

[0] https://depot.dev/


Is there a write up on the security of actions or equivalent that explains how they are secure both with direct and transitive dependencies? If this applies to Depot.

Ah got it, thanks. I thought there was another kind of GitHub runner (like their "large" runners) that I hadn't heard of.

Not to mention its ergonomics issues. I held onto mine as long as possible because I loved the capacitive shell. Eventually I had to ditch it though to keep my wrist healthy.

Are you talking about the linked website or Flock itself?

fixed

On the contrary, I cannot use the top Gemini and Claude models because their outputs are so out place and hard to integrate with my code bases. The GPT 5 models integrate with my code base's existing patterns seamlessly.

Supply some relevant files of your codebase in the ClaudeAI project area in the right part of the browser. Usually it will understand your architecture, patterns, principles

I'm using AI in-editor, all the models have full access to my code base.

You realize on some level all of these sort of anecdotes, though, are simply random coincidence .

But I think HN’s comment section is the only one worth reading!


On the other hand, I only use flexbox and I think grid sucks.


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

Search: