231–240 of 540 posts
Earlier quoted context omitted.
This isn't a web development concept. It's the unix philosophy of "write programs that do one thing and do it well" and interconnect them, being taken to the extremes that were never intended. We need a different hosting model.
In my mind the unix philosophy leads to running your cloud on your own hardware or VPS's, not this.
Earlier quoted context omitted.
In my mind the unix philosophy leads to running your cloud on your own hardware or VPS's, not this.
exactly this, write - not use some sh*t written by some dude from Akron OH 2 years ago”
And my own kernel. Can't trust some shit written by a Finnish dude 30 years ago.
And my own UEFI firmware. Definitely can't trust some shit written by my hardware vendor ever.
Claude Code defaulting to a certain set of recommended providers[0] and frameworks is making the web more homogenous and that lack of diversity is increasing the blast radius of incidents [0] https://amplifying.ai/research/claude-code-picks/report
The other day, I was forcing myself to use Claude Code for a new CRUD React app[1], and by default it excreted a pile of Node JS and NPM dependencies. So I told something like, "don't use anything node at all", and it immediately rewrote it as a Python backend, and it volunteered that it was minimizing dependencies in how it did that. [1] only vibe coding as an exercise for a throwaway artifact; I'm not endorsing vib…
I once made a golang multi-person pomodoro app by vibe coding with gemini 3.1 pro (when it had first launched first day) and I asked it to basically only have one outside dependency of gorrilla websockets and everything else from standard library and then I deployed it to hugging face spaces for free.
I definitely recommend golang as a language if you wish to vibe code. Some people recommend rust but Golang compiles fast, its cross compilation and portable and is really awesome with its standard library
(Anecdotally I also feel like there is some chances that the models are being diluted cuz like this thing then has become my benchmark test and others have performed somewhat worse or not the same as this to be honest and its only been a few days since I am now using hackernews less frequently and I am/was already seeing suspicions like these about claude and other models on the front page iirc. I don't know enough about claude opus 4.7 but I just read simon's comment on it, so it would be cool if someone can give me a gist of what is happening for the past few days.)
I've been part of a response team on a security incident and I really feel for them. However, this initial communication is terrible. Something happened, we won't say what, but it was severe enough to notify law enforcement. What floors me is the only actionable advice is to "review environment variables". What should a customer even do with that advice? Make sure the variable are still there? How would you know if a…
> Environment variables marked as "sensitive" in Vercel are stored in a manner that prevents them from being read, and we currently do not have evidence that those values were accessed. However, if any of your environment variables contain secrets (API keys, tokens, database credentials, signing keys) that were not marked as sensitive, those values should be treated as potentially exposed and rotated as a priority.
https://vercel.com/kb/bulletin/vercel-april-2026-security-in... as of 4:22p ET
Earlier quoted context omitted.
Cloudflare’s developer experience doesn’t come close, it is terrible. Cloudflare are working on it, and hopefully they’ll be a real competitor to Vercel on ease of use someday, but right now, it is painful when compared to Vercel. Cloudflare is infrastructure first, Vercel is developer experience first.
Yes, CloudFlare's full of bugs and sharp edges. Not to mention the atrocious 3MB worker size limit (especially egregious in the age of ML models). They don't mention this up front in the docs and the moment you try to deploy anything non trivial it's oops time to completely re architect your app.
That's for the free plan.
Limits are documented here:
https://developers.cloudflare.com/workers/platform/limits/#w...
Earlier quoted context omitted.
exactly this, write - not use some sh*t written by some dude from Akron OH 2 years ago”
That's why I wrote my own compiler and coreutils. Can't trust some shit written by GNU developers 30 years ago. And my own kernel. Can't trust some shit written by a Finnish dude 30 years ago. And my own UEFI firmware. Definitely can't trust some shit written by my hardware vendor ever.
Claude Code defaulting to a certain set of recommended providers[0] and frameworks is making the web more homogenous and that lack of diversity is increasing the blast radius of incidents [0] https://amplifying.ai/research/claude-code-picks/report
Interstingly, a recent conversation [1] between Hank Green and security researcher Sherri Davidoff argued the opposite. More GenAI generated code targeted at specific audiences should result in a more resilient ecosystem because of greater diversity. That obviously can't work if they end up using the same 3 frameworks in every application. [1] https://www.youtube.com/watch?v=V6pgZKVcKpw
It is true that "more diversity in code" probably means less turnkey spray-and-pray compromises, sure. Probably.
It also means that the models themselves become targets. If your models start building the same generated code with the same vulnerability, how're you gonna patch that?