HN user

Saris

1,532 karma
Posts0
Comments890
View on HN
No posts found.

If I accidentally set up a passkey on my phone (let’s say I use Safari one day instead of my go-to, Brave), can I still log in without that passkey on other devices?

Yes, but you can also add the passkey to your password manager so it's available on all your devices.

Can I add another passkey on another device?

Yes.

How many passkeys can I set up for a particular site/app?

I haven't really seen a specified limit on any sites, but also if you're using a password manager it's only 1 passkey for all your devices anyways.

For now, I’m going to stick with LastPass and use Passwords; because no matter whether I lose my device or not or whether I’m on my own devices or not, I can be sure I’ll be able to get into a site/app.

Your passkeys would be in LastPass as well like your passwords, so arguably the same result regardless of which you use.

Edit: One final consideration, my spouse and I share user/name passwords for some things (notably Pandora and our Amazon Prime account) since they don’t handle things like family logins well; how do both my wife and I use amazon or Pandora with passkeys? Do we each set up passkeys? How do I get her Pass if that’s not an option?

If it was me I'd add a second passkey to my password manager for your wife under a new entry, and share that entry to her lastpass account.

Or if she's not on lastpass, you could just copy the data from the passkey over to whatever she does use.

One thing that frustrates me about MikroTik and OpenWRT too, is a lot of things require multiple steps to set up, like adding a port forward with hairpin isn't just one click, adding a VLAN requires adding stuff in like 5+ different places.

I really liked using Opnsense before I switched to Openwrt and I always want to go back, because it was just so much more streamlined and easy to configure from a 'geek but not a network engineer' perspective.

It's a tough question, I would be fine with ads if they were relevant to the content I'm looking at, minimal (ie; no motion, popups, covering up content), actually checked for safety so all the scam and phishing ads went away, and checked for performance so they didn't slow down a website so it's barely usable.

Maybe if adblockers become more widespread that will start to happen, right now ad networks seem to just allow anything.

Like if I'm reading a cooking article and there's an add to the side for some cooking gadgets or something like that, sure why not!

Are you comfortable sending it to US ones? Especially if installing Claude Code or another tool on your PC and it can collect all the data it wants..

On Openrouter Kimi K3 says it does not retain data or train on it, which is better than what US hosts claim for Claude, ChatGPT, etc.. as they collect and retain data even if you disable training on it.

Opencode or similar open source tool + a zero data retention provider is about the best option aside from running a smaller fully local model on your own PC.

DJI Osmo cameras are good, I still have my original Osmo action and while the quality is a bit behind now, the battery life and general stability and menus are better than GoPro IMO.

I've found DJI cameras also don't discharge their batteries when sitting, my gopro 11 black is somehow always dead when I grab it even after a few weeks, but my osmo action is still at ~70% after a year.

Insta360 also has some neat offerings, but their software/app is absolutely abysmal, it's crammed full of ads and takes up several GB of space. It also requires an account login.

The downside is the 360 editing tools are kind of sluggish and not great to work with, and even at 8k res in-camera, the export for a 'normal' looking FoV is pretty low quality compared to a normal action cam recording in 4k.

I have an insta360 X5, it's neat and there's a lot of flexibility, but it does have downsides.

The app is also a pile of crap, it's crammed full of ads, social media junk I don't want, it's slow as molasses, and the size of the app is massive.

What made you pick proton over something like mailbox.org? You'd get calendar and contacts sync, IMAP and POP access, custom domains, really nice webmail, and so on..

Proton seems really limited for what you get, and the webmail is absolutely abysmal in performance and design.

I don't know about hCaptcha, I tried it out when testing stuff for a websites checkout flow (had card testing spam issues), and they are about the most difficult to solve captchas I've ever done.

I guess there's a low friction mode only available on the $139/mo plan.

Immich 3.0 20 days ago

In proxmox restoring a backup is really easy, I'm surprised you didn't do that!

3D printing isn't stiff enough for a larger drone like this, you'll get flex which will upset the PID loop and cause it to be very unstable if it even flies at all.

You'll notice on your 3D printed drone link that the actual multirotor part is carbon fiber tubes. The motor mounts and other brackets can be 3D printed without issues.

I run it on a 12GB 4070 with 32GB system RAM. 35B A4B means only part of the model is active at a time so it takes a lot less VRAM than a dense 35B model would.

The main thing in LM studio (or whatever software you use, assuming it has fairly up to date stuff and exposes the toggles) is to offload MoE layers to the CPU, and use K/V cache quantization at Q8_0 or Q4_0.

Since you have more VRAM than I do, you could probably get away with MoE offload of like 15-20 so some remains on the GPU.

Just make sure GPU offload is turned all the way up. And I use 64k context size, although with 16GB VRAM you can probably do more.

You can find the best performance spot by playing with MoE offload until you find the number that gives the highest tok/s on your hardware.

Even with deepseek v4 flash I burned though $5 in credits in a day just playing around with Hermes, and qwen 3.6 35B is significantly more expensive.

I can run qwen 3.6 35B on my gaming PC at around 50 tok/s and other than power cost of a tiny bit extra per month, it's hardware I already owned from years ago.

I'm not really sure why qwen 3.6 35B is so expensive on openrouter, it seems abnormally high for what hardware it takes to run it.

Plus I feel like it has completely ruined typical login flows, normally a PW manager would auto fill the username + password fields, but thanks to oauth we often get only a username field, or have to click 'login with password' or some other silly step first.

GLM 5.2 Is Out 1 month ago

Qwen3.6-35b-a3b at 64k context runs quite well on my 12GB VRAM GPU with MoE partially offloaded to CPU. It does use a good chunk of system RAM too, but I get about 40-50 tok/s.

Yeah that's why I have so many ESP32-S3's around, no project I've done actually needs Linux and the boot times and SD card problems that come with using a Pi.