NorwegianDude

5 ms·

31.
by NorwegianDude 6mo ago
Just to point it out, Cursor has not made any good models themselves. Composer 2 is Kimi K2.5, and they tried to pass it as their own until people noticed that the api specified it as Kimi.
32.
by NorwegianDude 6mo ago
It's especially bad considering the US had already taken out 100 % of Iran's military capabilities, according to the official statements. What a clown show...
33.
by NorwegianDude 7mo ago
> I worked at a startup, they built their API in PHP because it was easy and fast. Now they're successful, app doesn't scale, high latency etc. What does their php code do? 95% of it is calling a DB. So PHP worked perfectly, bu
34.
by NorwegianDude 7mo ago
Qwen3.5 27B works very well, to the point that if you use money on Claude 4.5 Haiku you could save hundreds of USD each day by running it yourself on a consumer GPU at home.
35.
by NorwegianDude 8mo ago
At least for me, it's cheap. Even Claude Haiku 4.5 would cost over $60 each day for the same token amount, after accounting for electricity costs. I have the hardware for other reasons anyway, so why not use it, avoid privacy issues an
36.
by NorwegianDude 8mo ago
I guess this might be a decent way to farm data? Those with larger OSS projects usually have better code quality, making it easier to create a dataset with maybe higher quality for training. Considering how often people leak data to the LLM
37.
by NorwegianDude 8mo ago
Fun challenge, but running the benchmark on Apple hardware is a weird decision as Apple doesn't even have server hardware. Would make much more sense to run it on a dedicated Linux box as that is more accessible and more realistic.
38.
by NorwegianDude 9mo ago
I guess people are just a few days from starting to earn $30k/yr on their model 3s, as musk said it years ago? lol
39.
by NorwegianDude 9mo ago
To be fair, he's completely right. I have a lot of experience with IE6 and safari on iOS, and while IE6 was bad and did weird shit, Safari is much worse. It's amazing that things can work in any browser, without ever even thinking
40.
by NorwegianDude 9mo ago
If you're looking up values from a 3 GB DB, most would have to hit RAM. Lookups form a hash map can be fast, but SQLite does quite a bit more than just a hash map lookup, and it would usually hit RAM, not L3 cache.
41.
by NorwegianDude 9mo ago
> You could always allow users to choose the amount and just include the transaction fee. I'd be happy to do that, but it's not really as simple as it seems. Multiple of my payment providers have fees I'm not allowed to di
42.
by NorwegianDude 9mo ago
You do have some weird takes on why I'm doing things. The game is free. Not because I'm trying to exploit someone, but because it's much harder and more expensive to market a paid game. I don't have DLCs, cause all conte
43.
by NorwegianDude 9mo ago
A one time payment for an ongoing service is not sustainable. Infrastructure and content isn't free, the same way you don't pay a one time payment for netflix.
44.
by NorwegianDude 9mo ago
It's not solved if you're not allowed to bill for fees seperatly. The end result is the same, so it's a dumb rule. It just makes it less clear what you're actually paying for.
45.
by NorwegianDude 9mo ago
Didn't I just say that it wasn't part of some evil plan? Prices of things have no relattion to that the different bundle sizes are. In-game prices are constant to ensure that people get what they pay for. The price in real money f
46.
by NorwegianDude 9mo ago
40 million reads per second, on a single core? 40 million reads/s is 25 ns per read, that is faster than any RAM I know of.
47.
by NorwegianDude 9mo ago
I've been using GitLab since the early days, but a week ago I switched to Forgejo. Power usage on the server dropped by 10%, despite both being idle most of the time. As the author says, GitLab feels sluggish, and is bloated with 1001
48.
by NorwegianDude 9mo ago
I do t have time to test myself now, but it would be interesting to see a proper benchmark. We all know it's not suitable for high write concurrency, but SQLite should be a very good amount faster for reads because of the lack of overh
49.
by NorwegianDude 9mo ago
SEK, not EUR.
50.
by NorwegianDude 9mo ago
As a game developer myself, I think young kids should not be able to make purchases on their own. But some of the ideas on what needs to be done is just silly. Here is some of the ideas the Norwegian Consumer Council suggested: - All things
51.
by NorwegianDude 9mo ago
How freaking expensive do you think infrastructure is? It's not that expensive, and certainly not anywhere close to the point where it would make a noticeable impact on GDP.
52.
by NorwegianDude 10mo ago
It's not a wrong solution. It's been commonly used since forever, tens of years before the sec-fetch-site header existed, and it stops CSRF. Sec-fetch-site is not supported in old browsers, so relying on that is unsafe without any
53.
by NorwegianDude 10mo ago
The simplest way to prevent CSRF is to use the Referer header, and that has been used since forever. If the header is missing, you no-op the post. Origin is similar, and can be used with referer as fallback, but it's not needed for mos
54.
by NorwegianDude 10mo ago
To be fair, Gmail asks for a phone number, but you dont have to add one.
55.
by NorwegianDude 10mo ago
This is very misleading. The secure defaults for sqlite is changed, so commits are not actually written to the disk. Running sqlite like this will cause data loss on os crash or power loss.
56.
by NorwegianDude 11mo ago
It is an option. You can run without cloudflare, and if you ever need filtering then you just swap over with little downtime.
57.
by NorwegianDude 11mo ago
To be fair, a single server is way more reliable than cloud clusters. Just look at the most recent many hour long Azure downtime where Microsoft could not even get microsoft.com back. With that much downtime you could physically move drives
58.
by NorwegianDude 1y ago
Nobody cares if I can do it a million times faster, everyone can. It's cheating. The whole reason you have to account for the time you spend setting it up is so that all work spent processing the data is timed. Otherwise we can just pr
59.
by NorwegianDude 1y ago
This is very silly. You're not doing the challenge if you do the work up front. The idea is that you start with a file and the goal is to get the result as fast as possible. How long did it take to distribute and import the data to all
60.
by NorwegianDude 1y ago
Yeah yeah. Nobody believes what Americans say, be that Trump, Elon or Apple. They're all full of shit, and they rarely do what they say. The average junkie is a more reliable source on what Apple will do than Apple itself.

More ›