One could make a community named “Anon Posting” or something, lock it so only a mod can post, and then make the sole mod a bot that would post anything it got via DM (probably after automoding, rate limiting, etc) to said community.

I do think it’s a good idea for the bot to keep a log in case it gets abused for sufficiently evil purposes. One could add some extra functionality to the bot that would give identifying information about the poster to instance admins on demand (via DM), but I think instance admins would have pretty easy access to all DMs made to the bot, along with identifying information anyway. (Feel free to correct me if I’m wrong on that.)

Also, the bot could totally delete its logs and with them the identities of all posters after a while. Maybe a month?

And, of course, this wouldn’t be ironclad anonymity. But it would keep identities secret from anyone but the bot maintainer and instance admins.

Yeah, sounds like a pretty cool concept. Not volunteering to write such a bot (at least any time soon) or anything, but I support it.

So, I’ve been using Arch Linux ARM on Raspberry Pis for some “desktop systems” as well as for a janky-ass NAS solution, but that project is kindof dying. They go many months in a row sometimes without any package updates. It’s wild. And when people ask WTF is going on and offer beg to be allowed to help in some way, the admins lock the thread.

So, I’ve been looking to switch my Raspberry Pi’s to something that doesn’t depend so much on some “project” out there to be able to continue to use.

The main Gentoo project fully supports ARM. And even if it didn’t, it’d be a lot easier to use Gentoo without support than Arch.

Switching my main box (not a Raspberry Pi – it’s an x86_64 system) to Gentoo was basically for the purpose of trying out Gentoo again and evaluating whether I want to take the plunge and switch everything to Gentoo.

Aside from that, there’s SystemD which is yucky. (Yes, I know about Artix, but when last I tried it, it didn’t really feel “ready for prime time”. It depends a lot on the main Arch repos.)

Plus, I do kindof like the idea of “more control over my system(s)”. Configuring/compiling my own kernel (yes, you can do that on Arch, it’s much less “in the spirit of” Arch) to make it as minimal as possible and disable everything I don’t need. And of course USE flags are a plus if you want a light system.

Anyway, those are my main reasons.

Me too!

I used Gentoo almost exlusively from like 2003 to maybe 2012 or 2013. I switched to Arch about then. But quite recently I made the switch back to Gentoo on my primary box and I’m happy I did.

Only thing I still need to do to really make it long-term sustainable for my particular use is to set up a build server on my network. My “primary box” is in the room where I sleep and I need it dark and quiet when I’m sleeping. Can’t have MOBO color-shifting LEDs and fan sounds overnight. And I can’t compile something like Chromium in less than the 15-to-16-ish hours I’m awake in a given day. (And I’d prefer to compile it myself rather than using a binary package.) Hence the need for a build server.

It’s a scam.

Get your blockchain BS out of here.

I upvoted this before I saw what community it was in. Now I wish I could upvote twice.

Disney is no stranger to hypocracy of that sort. Look at them making their billions off of the public domain (Snow White, Cinderella, Aladin, The Little Mermaid, need I go on?) while lobbying heavily for longer copyright terms to keep works they made from being similarly adapted.

DALLE LLaMA

Go:

describing IntelliJ as “good”.

Shots fired back. 😈

I never would have thought to print them at an angle like that, but thinking it through, I bet relative to other obvious-ish options, it a) improved part strength (particularly along the axes where you most need strength), b) saved a bit of material, c) improved bed adhesion. Smart move in general. I’ll have to keep that approach in mind for my own prints.

I haven’t watched the video yet, but just because it’s relevant to the topic…

I used to stream to Twitch with just ffmpeg. No OBS or anything.

I mostly did speedruns, and I needed a timer, so I wrote my own. I had ffmpeg read the current time to display from a file in /tmp/ and had a Go program that would write to that file at the same rate as the framerate at which I was streaming. Worked really well, actually.

I also made some videos (mostly tutorials for pulling off certain glitches in The Legend Of Zelda: Breath Of The Wild) and put them on YouTube. I edited those entirely with ffmpeg and a pretty simple Bash script.

I’m definitely not going to claim they were what you might generally recognize as “well-edited” videos, but they did the job. And I definitely wasn’t really looking to “make it big” on YouTube or anything, so I wasn’t looking to polish my videos.

Here is one of my videos for reference. And here is a clip of a VOD from one of my streams that demonstrates the timer I mentioned.

Yeah, I’m about as rabid a Free Software advocate as you’ll likely be able to find, but it’s 100% time we moved on from Stallman.

I want everything to be owned by myuser, group media

Wait, “everything?” Yeah, that’s probably contraindicated. You don’t want to be changing ownership of stuff in, say, /etc or /bin or whatever to your user. For the most part, stuff in those locations should be owned by root:root. If there are exceptions (things that should be owned by root:<something else>), the package manager will make sure they’re set as they should be.

Yup. Entirely possible. Blocking third party cookies might somewhat reduce sites’ ability to tell that you’re the same you on the same browser between VPN and direct connection, but even that isn’t any guarantee that Linkedin (and/or the ad providers Linkedin uses) and Spotify (and/or their ad providers) don’t know you’re the same user between VPN and direct. And if there’s some amount of collusion and/or purchase of user tracking info going on between those entities, even only first-party cookies are sufficient for them to be able to prove the link between your direct and VPN IP addresses. Even without any cookies, though, there are still browser fingerprinting techniques that are worth looking into if you want to know more about defeating that sort of tracking.

The way I’ve embedded magnets in prints in the past was to:

• Design a magnet-shaped (plus like 0.2mm of clearance) cavity into the print, but leave it completely “closed off” to where it’s “inside” the print.
• But only “closed off” by like 2 or 3 layers (I was printing at 0.2mm layer height for this particular print).
• Use “pause at layer” functionality in my slicer (I used Cura at the time) to pause just before the first layer that would “close off” that cavity.
• Start the print and when it pauses, drop the magnet into the cavity.

Yes, I was a bit nervous about the magnet potentially jumping up and sticking to some ferromagnetic metal that’s part of the print head, but that didn’t happen in my case. YMMV, I guess.

I guess theoretically it could also be the case that the heat from printing could weaken the magnet, but again, that wasn’t an issue in my case.

Just to elaborate on what my project was, I had a freely-spinning part that I wanted to be able to fix in place or unfix. I fashioned a “stop” that when engaged would fix the freely-spinning part in place. The way it works is that the stop can move freely up and down. Putting it in the “down” position fixes the freely-spinning part in place and gravity keeps it engaged. But to disengage it, you slide it straight up. At the top of the “track” in which it slides is where I put the magnet. I used the same technique as described above to embed a little stack of about four staples into the stop itself. So, by sliding the stop to the top of the track, the magnet attracts the staples, keeping the stop disengaged until you pull it back down again to where gravity will keep it engaged until you move it back up.

Then make the “one true frontpage” for Lemmy or whatever (implement ActivityPub, maybe borrowing some code from the Lemmy codebase itself, or kindof making a fork of Lemmy), and if it’s good, it’ll be used. If not, it won’t.

But then, it might well fall victim to this phenomenon:

Lemmy has lots of competing “front pages.” How will one more change anything? A more generic domain name or something?

as it gets better

Bold assumption.

#!/bin/bash

set -e

find /path/to/parent/directory/of/mp3/files -iname '*.mp3' | while read m ; do
  t="$(basename "${m}" '.mp3')"
  [ -f "${t}" ] && mp3splt -A "${t}" "${m}"
done

Something like that?

I’d definitely encourage you to study and understand it so you know basically what it’s doing, but it should be roughly in the direction of what you’re describing.

Whew. My 586 is safe.