I’m guessing you got similar drives to the ones I got. I paid ~$72 each (4x 12TB), and now the same HGST DC drives from the same seller are $220. Just glad I got them when I did, even though now I have to continually prune the data so it all fits within this forever limit.

I need to be honest about something: I wrote this post with an AI.

Not just edited by AI. Written with it. I fed it my GDPR data, asked it to analyze the files, and said “help me write a post about this.” The same tool I’m criticizing is the tool I used to criticize it.

And tomorrow, this conversation — me processing my discomfort about data collection — will probably end up in another memories.json somewhere. Another entry in another psychological profile. Another data point about how I think, what I fear, what I’m willing to expose.

I could have written this myself. It would have taken longer. It might have been worse. But I didn’t, because the AI is faster, and I’ve gotten used to outsourcing my thinking.

That’s the trap. It’s not that AI is evil. It’s that it’s genuinely useful. So useful that you keep feeding it, even when you know exactly what it costs.

Even admitting it just seems embarrassing to me, but w/e. Brain-dead gonna brain-dead.

Doubt

Do it already.

I love a lot of the stuff he does, just can’t fathom how he doesn’t connect the few remaining mental dots (for a start, he’s still pro-cop).

https://vimeo.com/218966342

Yeah, it’s Donna’s spaceship keyboard!

100% agree with you re: Proxmox. I’ve recently migrated my gaming PC to a Proxmox setup with a Win10 gaming / VR VM, and a Debian 12 VM solely dedicated to serving, quantizing, and optimizing LLM (with full 3090FE vfio passthrough 😁). The other one I have is a super old mini-ITX tiny box with an i3-4130 in it, and I use it for a Plex LXC b/c my NAS has a CPU that doesn’t support hardware transcoding (even though I’ve literally showed all my clients how to disable transcoding completely so they all get direct streams / direct plays at original quality to their devices), just in case some transcoding needs to be done.

So I decided to set up the Cluster/Node bit a few days ago, and it is SO awesome to have instant access to both servers at one URL and interface to manage all my VMs/LXCs. I’ve only had one problem with Proxmox since I started using it a couple years ago, and I’ve loved everything else about it!

In the spirit of “why not right?”, here’s one of my favorite random services I run: https://github.com/jordan-dalby/ByteStash I love being able to save little snippets that I know in the moment I will hit myself later if I have to look it up again.

I’m currently running 2 Proxmox hosts with 3 LXC containers and 3 VMs between them, and on my NAS - 2 VMs and… 50 docker containers.

I reeeeeally don’t like centralized services. 😂

Well that sucks.

If display is flex.

That’s the exact guide I’m currently working my way through 👍

Yeah, I unfortunately noticed that I didn’t get the CHT high flow nozzle, just the regular brass one. With my tests and high temps, I was only able to get right around 30mm/s³, so I need to order the high flow nozzle, which they advertise being able to get up to 50mm/s³.

I suffered 3 long years endlessly upgrading, tweaking, and maintaining an Ender 3, so I definitely know your pain! 😂

Yep, you’re exactly right! I forgot that I have the volumetric speed capped at 25 for the Turbo PLA (based that off of random reported values other people used, but didn’t think to try and benchmark it yet). So I’m gonna do that right now! The Overture Matte PLA also prints incredibly cleanly at 300mm/s, so I’m going to benchmark it and the Turbo PLA to see how high this hotend can go. I’ll report back when finished.

I just realized that I have that capped, and I’m now running some flow rate tests on my different filaments. I’ll make another post with the highest flowrate I achieve 👍

Heh - I currently have everything set up for the glass bed mod, except I just read that it doesn’t work with the CR touch sensor, so I’ll need to get a BT Eddy sensor or sonar something. I’ll also need to print a new gantry, and I’ve already tried one that didn’t work as well as the stock one, so just waiting to do more research on that, and then I’ll have my glass bed all set up!

Edit: oh yeah! So the way I calibrated it was:

• I removed the extruder from the gantry (3 screws)
• I fed in some filament (cold, obv) until I could see it on the bottom of the extruder
• Flush cut the filament
• Issue some commands to extrude 100mm
• Flush cut that segment off again (try to cut exactly the same as before)
• Use digital calipers to measure that piece
• then just calculate your new value and run the test again to see if it extrudes exactly 100mm
• enjoy amazing prints

Good ✊

Max of 5 device claims per person, so max of $100 per claimant.