Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Elevated error rate across multiple models

status.claude.com

Post reply on HN
Print view View original

251–260 of 293 posts

Re: Elevated error rate across multiple models

#251

Post by tovej »

arbll wrote: post #228
pluralmonad wrote: post #224

Earlier quoted context omitted.

Is this stance gate keeping users? Isn't a pkg manager installation also a one liner? This seems more like gate keeping lazy distributors.

A lot of those scripts are wrappers around package managers. Creating them is extra work for distributors, but they still do it because package-manager installs are not truly one-liners and offer far less control over the installation experience. Users need to figure out which of the 10+ package managers they should be using, then run several commands. If something fails, the error messages are often cryptic and not…

Your arguments do not make even a little sense.

In what world does a user have to choose between 10 package managers? Each distro has exactly one. There are also only about three, maybe four main package managers out there.

A shell script being piped into bash has so many more ways to break than a package. And if yhe theory is that package managers are fickle (they aren't), then how does adding more complexity help?

It is much simpler, much safer, and easier to maintain a package than an install.sh, eapecially for a big project.

Configuration can be handled by a script, yes. Here's a crazy idea: Your package can include scripts for configuring the software. It's almost as if most packages do. The scripts/utilities could even restart a systemd service for you.

Unless you're talking about configuring your build, in which case we're dealing with an experienced developer who will have no trouble just cloning the repo and building from source.

My biggest issue is: if we're dealing with someone who can't use a package manager, we're dealing with someone who doesn't have the capacity to judge how safe a script downloaded off the internet is. This does not drive linux adoption, it drives botnet adoption.

Re: Elevated error rate across multiple models

#252

Post by arcanemachiner »

agentdev001 wrote: post #217
Carrok wrote: post #209

Earlier quoted context omitted.

If you care about privacy at all, you can route your Opencode requests through an inference provider that does not retain any logs or data. It is also much cheaper. So if your boxes include `Privacy` and `Affordability`, then no, they don't tick the same boxes.

You can use the Codex harness with non-openai providers if you want.

Pretty sure you need to use an older version of Codex for this to work.

Re: Elevated error rate across multiple models

#253

Post by badlibrarian »

It started failing two days ago, when it suddenly couldn't access gmail threads reliably. Then it started popping up warnings that I was over quota when I wasn't. It even let me use Fable briefly, or pretended to. Meanwhile search finally started working, so there's that.

This video, wow: https://www.threads.com/@founder__growth/post/DZz_9Ikj3Wx

Out of desperation, I moved to ChatGPT and it's working better than I remember. All these companies are playing games under load, under failure. No wonder we can't agree on what's good for what.

Re: Elevated error rate across multiple models

#254

Post by dgellow »

robertsconley wrote: post #124

I have been developing software since the late 80s, mostly CAM software for metal cutting machines, and I have been refereeing tabletop roleplaying games like Dungeons & Dragons since the late 70s. I get the power of LLMs, and I do find them useful. But I find them useful in much the same way I find a really good set of random tables useful, or a good set of rules for procedurally generating something like a star sec…

> So far, I am failing to see how the inherent probabilistic nature of the technology can be fully overcome

I deeply hope we never reach the point where that’s overcome. What we’ve seen over the past few years is how AI will destroy humanness from pretty much the entire digital realm. It’s by far the most evil, anti-human technology ever created, corrupting everything it touches. The last thing we need is for it to become reliable

Re: Elevated error rate across multiple models

#256

Post by cromka »

trollbridge wrote: post #221
cromka wrote: post #215

Earlier quoted context omitted.

I use Go plan precisely with Opencode IDE (and also Jetbrains IDE suite), but now also have access Gemini Pro and Claude Pro. And wonder which tooling to invest my time into, especially that MCP servers also potentially come into play here, and I want at least some models/tools to handle private tasks, like handling my increasingly-complex Home Assistant setup. And I also want to start using models according to needs…

I use “all of them”. My primary harness is oh-my-pi. I probably use 10 different models on a regular basis. I occasionally use OpenCode. I try to use Codex and Antigravity as much as I can, often using it as a secondary agent (due to different usage pricing models than API). The same skills and MCPs work across harnesses. Edit: I don’t use Claude Code simply because I already have enough to deal with and don’t see a…

Nice, thanks for the write-up!

Re: Elevated error rate across multiple models

#257

Post by kridsdale1 »

Legend2440 wrote: post #240
robertsconley wrote: post #124

I have been developing software since the late 80s, mostly CAM software for metal cutting machines, and I have been refereeing tabletop roleplaying games like Dungeons & Dragons since the late 70s. I get the power of LLMs, and I do find them useful. But I find them useful in much the same way I find a really good set of random tables useful, or a good set of rules for procedurally generating something like a star sec…

The trouble is: there is no deterministic algorithm that can do the things neural networks can do. For many of these problems, I think it is likely that no deterministic algorithm can exist because the problems are fundamentally underspecified. E.g. a common task in computer vision is generating a 3D depth map from a 2D image. This is inverting a lossy projection, so any solution must be a least partially a hallucina…

I feel the same way. Analogy: we’ve been geologists this whole time, building our dynamic and interesting mechanical planet.

Now, biology exists. It’s wet and messy and impossible to understand (we haven’t invented the microscope yet). That doesn’t mean biological study is not worth doing.

Re: Elevated error rate across multiple models

#258

Post by msdz »

cyberax wrote: post #229
NekkoDroid wrote: post #206

Earlier quoted context omitted.

I really hate the `curl | sh` specifically because if your connection drops at a specifically unlucky point in time you are left with a partially executed script which if you are unlucky enough may just have been executing `rm -r ~/.cache/ /download` but it stopped at `rm-r ~/`. Is it likely? No. Can it happen? Yea. Just make it `curl -o && sh ` and this entire problem is gone.

Most scripts now put all the code into a shell function and call it in the last line of the script, so this bug can't happen.

Correct, and/or in addition, most nowadays prepend something like `set -euo pipefail` to the scripts in the line immediately after the shebang which results in stopping on errors, including things such as syntax errors stemming from e.g. incomplete installer transmission over wire.

(At least for bash scripts, I’m not sure whether these are POSIX syntax to be frank.)

Re: Elevated error rate across multiple models

#259

Post by arbll »

tovej wrote: post #251
arbll wrote: post #228

Earlier quoted context omitted.

A lot of those scripts are wrappers around package managers. Creating them is extra work for distributors, but they still do it because package-manager installs are not truly one-liners and offer far less control over the installation experience. Users need to figure out which of the 10+ package managers they should be using, then run several commands. If something fails, the error messages are often cryptic and not…

Your arguments do not make even a little sense. In what world does a user have to choose between 10 package managers? Each distro has exactly one. There are also only about three, maybe four main package managers out there. A shell script being piped into bash has so many more ways to break than a package. And if yhe theory is that package managers are fickle (they aren't), then how does adding more complexity help?…

It's crazy to me that even after seeing so many major software distributors choose `curl | sh` as their entry point, people like you will still argue to the ends of the earth that there’s no problem with the package manager ecosystem.

I'll stop there. I'm not interested in continuing this discussion when it's being conducted in bad faith.

Re: Elevated error rate across multiple models

#260

Post by rvz »

root-parent wrote: post #42
freshtake wrote: post #24

I hear that 100% of code at Anthropic is coded by Claude, so this was caused by Claude. And also, no one but Claude can fix Claude

>> I hear that 100% of code at Anthropic is coded by Claude, so this was caused by Claude. And also, no one but Claude can fix Claude Claude is down....

Claude is taking a hydration break.
Post reply on HN