Hacker Newsnew | past | comments | ask | show | jobs | submit | seaal's commentslogin

Quota for flash 3.5 is terrible, they just reduced quotas once again. $20 plan gives you 4x usage, previously it was 33x.

I get significantly more usage from $20 Claude plan using only Opus 4.7 - but at least the google plan gives me 5TB of storage now.


And considering before you had only 5 hour Flash quotas and now the same quota applies to Pro and Flash and AG CLI (meanwhile Gemini CLI which had independent 24h quotas is getting killed in a month), once you run out of a couple 5 hour windows that's it for all usage (pro or flash) for a week.

https://huggingface.co/google/gemma-4-31B-it/commit/e51e7dcd...

There was an update to tool calling 3 days ago. I haven't tested it myself but hope it helps.


Wow, that is so much better! I didnt exactly test it extensively but my issues are gone.


Hmm.. is there an updated onnx?


I mean, just setup redirector extension and never think about it again.

Redirect: https://x.com/*

to: https://xcancel.com/$1


NVIDIA driver progress has been massive over the past year, I wouldn’t consider it much less stable/supported than AMD cards.


Definitely better now with their new "opensource" driver.

I still ran in a few snags:

- DKMS can break, e.g I had a kernel bump to 6.18 or 6.19 and the nvidia driver wasn't ready yet so the build failed. A mainline driver will always win this one.

- Suspend almost always works, but sometimes fails on lid close which is of course when you can't see it fail and my laptop battery dies unexpectedly. You'd say use hybrid sleep but that reliably always fails with the nvidia driver too. Both work flawlessly with Nouveau.

Since I don't need the extra perf on this laptop I just use Nouveau to drive the the dGPU + the AMD iGPU most of the time which is powerful enough for my non-desk needs.


Agree on both counts. I use debian unstable and is usually 50/50 on whether the machine will reboot on a working display after a kernel upgrade. Very easy to fix if you have a bit of knowledge, but certainly not ready for the general public.

I don't have a laptop with an nvidia card, but I often suspend the linux gaming machine on my living room, and sometimes it doesn't come back from sleep, while my steam deck never failed to.


dude, the whole Linus sticking his finger up at nvidia meme? Its still real in 2026. The opensource ABI whatever the fuck they call it is trash. I'm absolutely ready to purchase an AMD card next GPU I buy. I don't want to give nvidia anymore money, I'm done. It'll be AMD GPUs from now on no matter the performance diff, purely because they've got a better attitude to supporting non MS deployments.

There's too much TPM/SecureBoot/Enroll key hoops you have to jump through that a lot of distros just haven't bothered with.

If I'm being completely real, I'd be running FreeBSD 15. I just could not get a working nvidia driver going in 15 and get a working X installation. Supposedly 15.1 fixes it, we'll see in June. I've always preferred the BSD design, fs layout, etc, and I would love to have a FreeBSD desktop with a wine 11 install that actually plays games.. the dream!


Nah, nvidia drivers on Linux 2026 is hands down just as easy as AMD. I’ve had no more issues than running an AMD card and everything works flawlessly. They’re 100% right at the absolutely generational improvement in nvidia drivers since they’ve released the open drivers. Linus himself straight up said anyone trying to say this stuff in current year is being super disingenuous twisting his words from ages ago and that he considers nvidia a fantastic partner nowadays. And frankly, anyone unironically trying to use X in 2026 deserves the pain, it’s been officially deprecated for a while now and they maintainers were ultra clear that the only reason to use it now is for compatibility reasons and that you should expect issues if you do. Wayland is so far ahead of X now that anyone still trying to use X is being purposefully obstinate.


Is it thanks to ML needs or is it unrelated?


Which honestly doesn't sound that bad given how many users one server is able to serve.


People that were tracking this were already aware but glad to have confirmation.

This blog post I was reading yesterday had some good knowledge compilation about the model.

https://blog.devgenius.io/z-ais-glm-5-leaked-through-github-...


https://status.claude.com/

what do you mean it's right there. Judging by the Github issues it only took them 10 minutes to add the issue message.


It appeared there like 5 minutes ago; it was down for at least 20 before that.

That's 20 minutes of millions of people visiting the status page, seeing green, and then spending that time resetting their context, looking at their system and network configs, etc.

It's not a huge deal, but for $200/month it'd be nice if, after the first two-thousand 500s went out (which I imagine is less than 10 seconds), the status page automatically went orange.


It was longer than 10 minutes, I'd say 15-20 minutes this time. They should be much quicker, I would expect <5 minutes.


It took them about 15 minutes to update that page


https://github.com/opencode-ai/opencode there was also the competing opencode which because crush done by charm.sh people.



Surprised no one has shout out zsh4humans, perhaps because it's basically in maintenance mode but it's not like I need any new features. Love the ssh teleport feature.

https://github.com/romkatv/zsh4humans


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: