News for hackers. Stuff that matters.

starfallg

This is a really good point that goes the other way as well. Some use cases are great for smaller and cheaper models, but it's not obviously when that's the case so I can bank the savings.
Re:Artificial Analysis Intelligence Index v5 on Wednesday October 28, 2026 @02:47AM (#50269988)
It makes sense for Muse Wonder though as that model does not trade blows with the best frontier models by a longshot.
Yeah, I too was confused for a second.
Yeah, thinking that after the last supply chain attack, people would be smarter than ship things with LiteLLM.
Does anybody know how Wayland performs on these?
Re:Codex is EATING my Mac's disk space on Saturday October 17, 2026 @04:52PM (#50135035)
I'm working on a side project that requires a lot of fine tuning of 3D worlds. Codex ended up creating thousands of artifacts, basically loads of screenshots, in its worktrees, and doesn't clean up after itself.

On a work project, we're moving financial processes from Excel to code, and Codex ended up creating and storing thousands of validation tests unnecessarily. That ended up eating 90GB after we let it loose on a Windows VM that we connected with a custom agent between codex and the VM.

This is definitely a Codex thing and I've never seen this happen in Claude Code or Antigravity.

Re:Desire to Move Permanently to U.S. at New Low on Friday October 16, 2026 @01:44PM (#50121724)
The issue is that people's idea of illegal immigrants are hardened criminals living a lift of crime, whereas in reality the majority of people that are being deported are normal people that overstayed their visa, ie. the low hanging fruit. So people can be both supportive of the policy but dismayed at the implementation, which is what we are broadly seeing here.
Re:Gemini 4 Argon on Wednesday September 30, 2026 @08:22PM (#49916193)
I found 3.8 Flash in Agy to be generally better than GPT 6, and only behind from Opus 5.5.
Yes, but 1. Cowork burns tokens, lots of them, and tokens are rationed on the Enterprise plans, and 2. You need to know how to prompt it to make the changes. It's often a slow and painful process if all you want to change is a number here and there, or the colour of a box.
Thanks. To be honest, I don't know because I have a personal Max plan I use for this stuff and never set any telemetry up on it to get accurate metrics.
Thanks for the pointer on this, will check this out.
Notes was suggested in another comment as the next app in the series, so yeah I think it's a very good idea to do something like a workspace.
Thanks. Kinda jealous that you don't have to do slides to be honest.
Thanks for bringing this up, it's a valid point and also something I haven't really focused on specifically. I'll add at least the beginnings of these accessibility features to the next release.
That's a really good suggestion, thanks. I've been thinking about what to do next, and notes is probably the natural progression, especially if you want to share between different devices with no cloud.
That's how Claude works with pptx files, but the format is terrible to work with harnesses. Maybe that's going to improve but the it's currently the user experince is not good.
Bento has auto-update baked in. It checks for signed updates in a manifest, downloads it, rewrite the file structure with the new and reloads with the new file.
Technically yes, and the order is just dependent on the chain of the keypress event triggering slide changes. Will investigate the best way to do this.
Thanks for the kind words. I'll put some contact details in my Github profile.
Let me reproduce it on iOS and push out a fix shortly. Thanks for the heads up.
Good spot, it does look like a bug when you are using 1 axis. I'll roll a fix in the next release today. Thanks for the feedback.
For layouts that need to be static with a fixed page size like PowerPoint slides it's better to have specific placement, otherwise you will be constantly fighting with the layout engine. It's just a tradeoff I guess.
We tried so many AI presentation makers but larger companies have brand standards to adhere to and these are terrible at handling them. That's the reason we started to hand roll web apps as presentations.
Thanks, I hope this approach catches on!
In my day job, we have several teams doing this replacing mainly SaaS vendors. It's a easy sell and very low risk replacing basically CRUD web interfaces over databases.
Yes you can! Use the agent skill, instructions here - https://bento.page/#get-skill
I really like your system. It is very similar to Bento in some respects, but I haven't seen any editing capability though. Thanks!
This doc goes into more detail -

https://github.com/nyblnet/bento/blob/main/docs/security.md

But basically when your bento HTML first gets saved, it generates a number of keys used for managing the permissions to join the Cloudflare DO. You can then send out an invite deck with a key that allow other people to join (with their own generated keys). With this architecture, you can revoke people's access, as well as giving them only read-access to the sharing session for example.

Thanks for the kind words. To collab, it's simple, go into share and click on invite, then it will save a HTML file with an invite key. Just send this to the person you want to share the session with, and they will automatically connect and join the session when they first open it in their browser.
I'm still having trouble tracking this down the Bento app. There's nothing in the source on cloudflareinsights.com and nothing is connecting to that either when I inspect from Chrome. Could it be from the website? (The website is a GH pages fronted by Cloudflare)

Can you share data or show where this is so I can quickly check?

Older »