Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
eli
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
211.
▲
by
eli
9mo ago
Doesn't seem like it has much to do with LLMs at all, just typical vendor lock-in nonsense like how Apple's own apps get entitlements not available to any other developer.
212.
▲
by
eli
9mo ago
Claude Code only trains on data if you opt in
213.
▲
by
eli
9mo ago
Significantly complicates getting a mortgage though
214.
▲
by
eli
9mo ago
An upgrade process involves heavy CPU use, disk read/writes, and at least a few power cycles in short time period. Depending what OP was doing on it otherwise, it could've been the highest temperature the device had ever seen. It&
215.
▲
by
eli
9mo ago
I'm not sure archiving and publishing are different things.
216.
▲
by
eli
9mo ago
I think if you’re going to complain about how mean and unfair the progressives are to you then you should have to include a few examples of the posts that provoked their ire. Like any social media site there are people on bluesky I find ann
217.
▲
by
eli
9mo ago
I was working at a startup that was being acquired and as part of due diligence they sent someone to do a "walk through" of our "data center," which was tricky because it was in fact a small coat closet with 2 beige towe
218.
▲
by
eli
10mo ago
I don’t think CSRF has anything to do with those?
219.
▲
by
eli
10mo ago
list-unsubscribe header sends a POST. Probably makes more sense to just use a token from an email anyway.
220.
▲
by
eli
10mo ago
Nah, you can get a plan for a couple dollars a year as a one-off https://www.digikey.com/en/resources/iot-resource-center/iot...
221.
▲
by
eli
10mo ago
" We examine whether firms have an informational advantage in selecting arbitrators in consumer arbitration [...] We first document that some arbitrators are systematically industry friendly while others are consumer friendly. Firms ap
222.
▲
by
eli
10mo ago
Two problems with that argument: 1) Amazon would also have to reimburse small claims court fees if you win, and 2) arbitration is worse for the consumer in pretty much every other way.
223.
▲
by
eli
10mo ago
Is TikTok fundamentally different from HN in some way? Seems like you could say that about any platform you don't like.
224.
▲
by
eli
10mo ago
Just need to outlaw binding arbitration
225.
▲
by
eli
10mo ago
And GPT costs 4x as much
226.
▲
by
eli
10mo ago
I was just correcting the parent post that implied the passkeys were only stored on the device. Personally I do not use that feature. I'm also pretty sure I don't have any accounts that can ONLY be accessed via passkey.
227.
▲
by
eli
10mo ago
I don't think this is correct
228.
▲
by
eli
10mo ago
You omitted my favorite feature: virtually immune to phishing. You can't accidentally submit a passkey to a lookalike domain. For phishing protection, passkey as a single factor is better than memorized password + TOTP/SMS two fac
229.
▲
by
eli
10mo ago
Android syncs them to your Google account and iPhone to your iCloud account by default. Which isn't a perfect solution but, again, is pretty good for most people.
230.
▲
by
eli
11mo ago
The Range header isn't for requesting a subset of a resource from the server?
231.
▲
by
eli
11mo ago
An HTTP Range request asks the server to send parts of a resource back to a client. Range requests are useful for various clients, including media players that support random access, data tools that require only part of a large file, and d
232.
▲
by
eli
11mo ago
Isn't that Gemini 3 and not 2.5 Pro? But nondeterministic algorithms are gonna be nondeterministic sometimes. Surely you've had experiences where an LLM is full of shit?
233.
▲
by
eli
11mo ago
> Most people use Cloudflare because they have been scared into the idea that you need DDoS protection I don't think that is correct that's why most people use Cloudflare
234.
▲
by
eli
11mo ago
Aren't all LLMs instructed to provide responses within the boundaries of their rules? How else could you have "rules"?
235.
▲
by
eli
11mo ago
Should be a bit faster if you run an MLX version of the model with LM Studio instead. Ollama doesn't support MLX. Qwen3-Coder is in the same ballpark and maybe a bit better at coding
236.
▲
by
eli
11mo ago
Because if you build an agent you'll need to host it in a cloud virtual machine...? I don't follow.
237.
▲
by
eli
1y ago
Sure, you could charge in hotdogs if you wanted to.
238.
▲
by
eli
1y ago
Assuming they updated the crawled training data, just having a bunch of examples of specifically pelicans on bicycles from other models is likely to make a difference.
239.
▲
by
eli
1y ago
They kinda have that already, no? https://openrouter.ai/apps?url=https%3A%2F%2Faider.chat%2F
240.
▲
by
eli
1y ago
To be fair it also ships with the ability to install community plugins disabled.
More ›