nojs

7 ms·

91.
by nojs 8mo ago
> If you ask ChatGPT or Claude to write something like this through their websites, they will refuse. This OpenClaw agent had no such compunctions. OpenClaw runs with an Anthropic/OpenAI API key though?
92.
by nojs 8mo ago
I am genuinely curious - what do you see as the difference between "agent-friendly payments" and simply removing KYC/fraud checks? Like basically what an agent needs is access to PayPal or Stripe without all the pesky anti-bo
93.
Ask HN: Aggregated Weekly World News?
1 points by nojs 8mo ago|1 comments
94.
by nojs 8mo ago
“Aperture” is a pretty creepy name for this. > By collecting usage information into a single place, engineering and IT leaders can get a complete picture into both user and agent token efficiency across the organization and providers. Wh
95.
by nojs 9mo ago
> The travel to/from Taiwan was statistically riskier than the climb That is not even close to being true, haha. Probably not as risky as it looks but come on.
96.
by nojs 9mo ago
It’s more about context management, not speed
97.
by nojs 9mo ago
I mean double tapping escape, going back up the history, and choosing the “restore conversation and code” option. Sometimes bits of code are restored, but rarely all changes. It worked when first released but hasn’t for ages now.
98.
by nojs 9mo ago
> checkpoints/rollbacks are still not implemented in the VS Code GUI Rollbacks have been broken for me in the terminal for over a month. It just didn’t roll back the code most of the time. I’ve totally stopped using the feature and
99.
by nojs 9mo ago
I've noticed an uptick in API Error: 400 {"type":"error","error":{"type":"invalid_request_error","message":"Output blocked by content filtering policy"}, re
100.
by nojs 9mo ago
Run the agent in a sandbox without access to production secrets.
101.
by nojs 9mo ago
You’re right to push back.
102.
by nojs 9mo ago
It seems like everyone wants to avoid running a local VM manually and I'm not sure why. It's a very simple solution that solves all these issues. If you're on a Mac working on a linux docker containers, your Docker engine is
103.
by nojs 9mo ago
How long ago? I’m surprised you got anywhere near their servers with 10k requests per second from a single machine.
104.
by nojs 9mo ago
I thought people generally run it daily? It’s a no-op if it doesn’t need renewal.
105.
by nojs 9mo ago
This post is spot on, the combo of tailscale + Claude Code is a game changer. This is particularly true for companies as well. CC lets you hack together internal tools quickly, and tailscale means you can safely deploy them without worrying
106.
by nojs 9mo ago
> What drove that growth? It is a natural fit with component-based frontend frameworks like React. You keep the styles with the component instead of having to work in two places. And it’s slightly nicer than writing inline styles. The co
107.
by nojs 9mo ago
The same thing that stops your phone’s Face ID working with a photo of a face, I suppose
108.
by nojs 9mo ago
I question the premise that it’s not taking over. Our logs are at least 50% ipv6 now. A few years ago I feel like a barely saw it.
109.
by nojs 10mo ago
It’s a major factor in build times for Django containers for example.
110.
by nojs 10mo ago
“X: the Y paradox” is pretty indicative too
111.
by nojs 10mo ago
This. An SPA has some chance of being repackaged as an app without major surgery, a Django app really doesn’t. So you have to maintain it separately. PWAs/Webviews/etc are not really practical options.
112.
by nojs 10mo ago
This may not be as bad as it sounds. Reviews are also presumably flagged as “fully AI-generated” if the reviewer wrote bullet points and used the LLM to flesh them out.
113.
by nojs 10mo ago
What were the problems you had with Celery?
114.
by nojs 11mo ago
Gitlab*
115.
by nojs 11mo ago
I also like that there’s no ads in premium. Speaking of premium things, I also really like Brilliant for learning things quickly and Surfshark VPN because it protects my privacy when I’m on the go.
116.
by nojs 11mo ago
Surprisingly my experience has been the opposite with qwen, if you can force the thinking trace to English the results seem better. But probably just due to the amount of training data.
117.
by nojs 11mo ago
Thanks for posting this - just set it up and liking it so far
118.
by nojs 11mo ago
> Thinking blocks from previous assistant turns are preserved in model context by default This seems like a huge change no? I often use max thinking on the assumption that the only downside is time, but now there’s also a downside of con
119.
by nojs 11mo ago
Using both extensively I feel codex is slightly “smarter” for debugging complex problems but on net I still find CC more productive. The difference is very marginal though.
120.
by nojs 11mo ago
This looks very useful. I would love something similar that lets me plug in actual Claude Code/Codex with their original agent loop, prompting etc, and just handles the multiplexing, worktrees, isolation, etc automatically (it looks li

More ›