tmikaeld

I was the original poster and was just about to post the same thing!

We really need to celebrate and bring out the Champagne :-)

I second that, everything "just works" as you'd expect it to.

If StreamOS can get widespread and they start to introduce desktop apps to the store, I think they can take some market share.

About that, it's clear that thousands of people read articles on HN but extremely few of them actually upvote anything. Please do that, it only takes a second but makes a huge difference.

All modern business is built on trust, since Americans are keen on throwing that out the door now, every EU business is going to re-evaluate who to trust.

Which means, no longer trusting American.

Probably SEO farm. Clearly not what it's purposed to be..

It's rightfully flagged

Mistral uses Cerebras for their Lé Chat, which is even hosted in the US.

I really don't think that the amount of applications demonstrate a saturated market, instead, since this is developer positions, It's probably developers better utilising AI/Automations to cast a wider net of job applications to have options if anything should happen to their current position. Considering the state of the US right now, I would definitely be doing this.

It may also be deliberate, I know a lot of people that are very dyslexic and are using AI for making themselves understood online.

This is the way.

Seriously, this is the only useful flow I've found for AI coding in general..

As an outsider trying to make sense of all this.

Is America taking the hammer down on all types of spending because America is desperate to bring down debt from 120% of GDP?

” The models will be developed within Europe's robust regulatory framework, ensuring alignment with European values while maintaining technological excellence.”

They may release something, but i doubt it will be more useful than what already exists.

Honestly? I find that teens have lost trust in basically everything in modern society, not just tech.

Neither deno nor bun do tree shaking because they can’t make the same assumptions as front end code. They do cache the compiled code though, which makes some difference. But the reason S3 SDK is at least 5X slower than bun, is partly due to code size, since JavaScript always have to parse the code on every run. I worked 3 years on cloudflare workers and scripts even 1MB in size are slow and cpu intensive to run (before using it!). Now, even the most common auth for node with basic features is 20MB of scripting. That put a perspective on things like this.

Batteries included makes a huge difference, this is why i love that Web APIs (Fetch API, Service Workers, Web Components, and ES6+, WebRTC etc) are now native on both V8 and Webkit runtimes.

But it has to be to a certain degree, maybe S3 is too far, but SQL drivers makes sense - but again, to which degree? There are _many_ databases out there, should there be drivers for half of them? Even at that level it's a lot of added code which means slower executable.

Also, I think Bun is missing out on security by adding such sensitive APIs to Bun, imagine bun taking all your source files and uploading it to your private S3 due to some script or path issue that allowed eval to run! It's game over right there.

AI in general doesn’t understand scope

It’s just horribly slow though, using sonnet can often take 5 min to apply

The exact reason that I use Deno is the sandbox, it will not run if something is calling an unknown endpoint.

Ah, I see they do have that now!

But it's 1Gbit public and max 5Gbit (Plus double the price)

While most of these are great, none of them seem to have support for 4x NVMe drives which Hetzner has.

Would be good to have a fall-back solution, is there something similar in dedicated server price in the EU as Hetzner? Or does no on else come close?

That's the irony of it all isn't it - since it's bound to traditional currencies for it's valuation, it's essentially the same thing as a traditional currency except you can't inflate it by printing more of it.

It is "digital gold" at this point.

It doesn't make sense to limit HEAD requests though as all S3 implementations store this metadata inside a KV database which is _built_ to handle much more than 750 reads per second.

Edit: There's an ongoing discussion here https://forum.hetzner.com/index.php?thread/31569-will-per-bu...

Yes, the amount of operations per second is too small for directly serving files like for hosting. (We didn't have this issue since we proxy/cache all files via dedicated hardware)

Though, it's still too small if a HEAD request counts as an operation, since we need to check if files updated or not.

Finally!

And the technical limits are really good now: https://docs.hetzner.com/storage/object-storage/overview#lim...

Just curious, what can a 0.5B model be used for?

Love the idea, is there an iOS version planned?

Momento? This? https://www.gomomento.com/

Seems to be cloud hosted only.

... with 4-bit quant and 4000 token context.

"Please note that normal charges for CDN bandwidth are billed separately"

Compared to cloudflare workers, which has free bandwidth, bunnys bandwidth is not that cheap at 0.01$/GB

So while their example suggests stream-encoding video is possible, it would probably be cost-prohibitive.

Impressive! Definitely going to try this, thanks!