sourcecodeplz

Running LLMs local is fun and powerful but if you want to get work done... it is a big headache. You have to pre-plan and plan, and make specs, etc... The big OpenAI, Claude models just get you with just a few sentences..

It is really weird reading things but I guess normal? It seems many feel this, including me. AI just compounds this behavior even more! Darn.

Not all webhosting companies are using cpanel. Cpanel increased their prices exponentially in the last few years.

Why is everyone on HN so fascinated with calculators? Was that your first real pc that you could program?

Sometimes the degree is needed in order for the company to be allowed to pay him over a specific amount/scale.

Great project!

This is also a fine example of a vibe-coded project with purpose, as you acknowledged.

Wow, I love it when people discover these NEW ways of doing things.

Think it is because they focus on what they know best. Coding a LLM harness is nothing spectacular.

What a great read, so detailed and with references! They are all migrating to cloud agents it seems.

Looks awesome. Wonder what happens if one engine fails, it will adjust itself probably.

> Use USB charging for the heart rate monitor, step tracker and notifications. Time display is powered by solar charging alone when the battery runs low.

Isn't this how AWS also started? They built internal devops tools for them that were so good and expandable that they decided to give others access to them.

Yeah, opencode was/is like this and they never got caching right. Caching is a BIG DEAL to get right.

Wonder how Canva were roped into this, Claude Design will just replace them from what I read?

Ah WordPress, the ever growing security nightmare

I do it even more simpler: build in PHP and webhosting from Hetzner. All managed: email, sub-domains, name-servers, OS updates/patches etc.

I really started to enjoy managed servers/instances.

I like that they kept limited access to Codex even on free tier.

LE: Someone said this is how the tiers are now counted:

"Essentially if old plus is 1x then new limits are: Plus - 0.3x Pro $100 - 1.5x Pro $200 - 6x (unchanged)"

But you still get the reduced context-window.

How much did it cost for how long?

Which chinese models do you use and do you use any for specific tasks?

Gives me Fight Club vibes.

Haven't seen a jump this large since I don't even know, years? Too bad they are not releasing it anytime soon (there is no need as they are still currently the leader).

Everyone has started either hiking their prices or limiting the tokens, gravy train is over. Glad we have open models that we can host; Sad RAM is so expensive..

Yep, haven't tried 5.1 but for my PHP coding, GLM-5 is 99% the same as Sonnet/Opus/GPT-5 levels. It is unbelievably strong for what it costs, not to mention you can run it locally.

This part is interesting:

"Plugin security is the root of this problem. Marketplace businesses provide trust when parties otherwise cannot easily trust each other. In the case of the WordPress marketplace, the plugin security risk is so large and probable that many of your customers can only reasonably trust your plugin via the marketplace. But in order to be part of the marketplace your code must be licensed in a way that forces you to give it away for free everywhere other than that marketplace. You are locked in."

There was much drama with wordpress some time ago and the plugin marketplace.

Nice presentation. The reality is there is nothing really special about the claude code harness?

I really appreciate the free options, without even needing a login. Wish they would also keep the small free weekly allowance for Codex.

I've tested many open models, Deepseek 3.2 is the only SOTA similar.

RIP legend

Ah just let it go already, why keep ruining peoples memories...