Mikelius

Daily on my Gentoo server, through a Cronjob every morning. It’s a custom script though, so there’s more than just doing an emerge update. It’ll send me ntfy notifications for the update results, if there are new news items, and if there are any time config merge updates to make. A few other things as well but that’s the main stuff.

Other servers, typically weekly or only manually when I ssh into them (for the ones I don’t really feel the need to update frequently).

I’m shocked how far I had to scroll down before it was mentioned, I was getting scared I would have to say it lol

As a professional, my reasoning for NOT using AI is as follows:

1. I don’t want to lose the muscle memory of what I do. Sure AI might be able to do annoying things like test templates… But that’s not a skill I want to forget or lose, as self written unit tests have actually helped me catch mistakes that “would have worked” in prod (i.e. Code functions, but has undesired outcomes). AI can’t usually spot that.
2. As a person who digs deep in cyber security and monitors heavily the malicious realm, I’m paranoid of malicious or weak code being spit into my repos.
3. I’m a privacy nut, too. Most “good” AI solutions are anti privacy.
4. If anyone here has done a proper code review of AI generated code from coworkers, they should know it adds a ton of extra time because of errors, inconsistencies with repo practices, etc and actually wastes the time of the developer and reviewers.

Am I saying “NEVER AI?”? nah. But it’s far from ready for me personally to even consider for programming purposes. I’m also well aware this isn’t what many others think or feel; I don’t scream at people for using it if it’s what they feel helps them.

I removed the dash at the end. Guess it was a typo

If you go for btrfs, be careful going backwards on kernel versions.

I had upgraded my kernel on Gentoo, which also happen to include a btrfs update. Booted up and found the latest kernel didn’t like something about my full disk luks encryption with RAID mirror setup (for the root partition, and unrelated to btrfs), so I decided to go back to the previous kernel. Big mistake.

My entire root partition got corrupted to hell. It mounted read only at first so I decided to try to go through regular repair steps. It got worse. Got to an eventual step that someone said could take a few weeks to restore (forgot the commands). This isn’t an option for my server. So with snapshots broken, unable to use the old and now new kernel due to corruption from attempting to go back to a previous kernel, I had to restore with a full partition clone backup I had created prior to the kernel upgrade… Also went back to ext4 again afterwards.

Btrfs treated me really well for a few years, and snapshots and performance are great, but once it hits a hiccup, you might in a world of trouble. Don’t think I’ve ever run into such a thing with ext4 over the years, which is why I reverted to it - not saying it’s immune to such things, but this is just me.

Not sure if zfs would have such a dramatic situation, but maybe something to consider about btrfs if you ever decide you’ll need the ability to go back a kernel version due to whatever reason.

Dang, talk about flashbacks I didn’t expect. Nicely done!

Another screensaver I miss and would actually put on my machine if I could… Johnny Castaway.

I’ve been using “passwords” on nextcloud for a few years now. Minimal issues with the app, moving apps, and browser extensions. Not perfect, but hey it’s self hosted and reliable.

Hmm that’s actually something I hadn’t thought about, but yeah it probably wouldn’t work for an offline mode… If the app is already open, maybe it would work, but I’m not 100% sure. In general, it does require internet to connect to your home assistant (my phone is setup with wireguard to VPN into mind so it’s not internet facing). So yeah unless it can be used during those random connection drops, maybe not a perfect option :(

My family uses a custom Todo list on home assistant. I’ve got a separate dashboard of items we can just click to add to the grocery list, and then when shopping, checking them will move them out of the way. Very convenient, quick, and shared. So if you use HA, it might be good option, especially since you’re wanting something simpler (I was in the same boat)

Interesting, I didn’t have this experience a couple of years ago. I wonder if they’ve just upped it to try and “automate” things more with the crazy amount of tourism they’re suddenly getting. Also I’d be curious on which airport you went to, Haneda or Narita?

If the scans and such were in the states, I’ve requested opting out and no one really cared, they just said okay. Funny enough, it actually made me go through quicker than it was taking everyone who did the face scans, contradicting the sign claiming it’s quicker.

I left like a decade ago when they asked me in a chat to verify my identity by answering a question asking what my first car purchase was. I’ve never given then my SSN or that kind of financial details, so the fact they had these questions and details about me terrified me at the time and I immediately requested to delete and close everything with them. Haven’t used PayPal again since then.

I’d say anyone wanting to go this deep into a home monitoring setup will likely go with what works best for them instead of reading and following the entirety of this guide… I’m one of those people…

Wrote my own log parsing software to put into a database, display and alert through grafana, which is alerting through a homemade webhook that sends a notification to ntfy based on severity… And I also use uptime Kuma like mentioned, but my notifications channel is ntfy. No cloudflare for my internal services, only wireguard to connect home and use everything. And definitely no telegram.

Plenty of other stuff setup, but my security alerts and monitoring rely heavily on the syslog/grafana server which helps me monitor everything.

Not an opinion, I have an actual situation with my eyes where they twitch uncontrollably when presented with bright lights for a long period of time. I have tried minimum screen brightness, lowered contrast/colors, auto brightness based on the environment, various software solutions to removing blue light 24/7 from the screen - none of it worked. Went permanently dark theme on everything, magically eyes haven’t twitched in years.

Light theme vs dark theme is not just a preference, it’s an actual accessibility need for some of us.

I converted my gaming machine into a server as well. I actually took the graphics card out as I couldn’t find a major use for it, but kept the 12 core Ryzen and upped it to 128gb memory. It now self host way too many things, including a few game servers my friends and I play… But even with all this, CPU carries along nicely and not even at half memory consumption (yet).

But as others have asked, what’s your goal? Don’t overkill it if you’re only hosting one service or something. If you’re doing a lot like I do, then up the RAM. And seriously consider whether the GPU is even useful or needed if you’re not using a desktop environment.

Cyberpunk worked out of the box for me, but senua 2 absolutely refuses to start no matter what kind of voodoo I try (“fatal error”). I seem to always be on the opposite spectrum of protondb mint users I swear.

I’ve had this issue many times as well. I’ve found changing the MTU would help since it seems some filter specific ranges. Doesn’t always work but I’ve had more success than failure doing so

As a Gentoo user, I can confirm I started from sticks and rocks. I’m now in the space age though because of the customizability and performance boosts, so image is a little dated.

Glad I looked at this thread. The fact they’re cheap and have what sound like reliable PoE hats… Tempted to replace a few old Pis lol. Maybe. But can at least say no future devices will be Pis at this point.

Note: only using them for simple things. Wireguard VPN (no I don’t have a fast internet so I don’t need more than the 1gb connection speed), pi hole, and a touch panel I installed that connects to home assistant on the wall.

This thread has provided genius ideas I somehow never thought of, and I’m totally stealing them for my sites lol.

Friends and I are in the upper 30s and 40s range so not young not old I guess lol. For the family side, I tend to look for all my closer relatives which range in all ages. While there were many many lines that matched our last names, the entries that were a match didn’t have the right phone numbers or addresses (so couldn’t really validate if they were us or others with the same name). Or it could always be that they were addresses so old that I don’t have a record of them to compare to… Considering a large chunk of the data is apparently old, it’s possible that could be a reason I didn’t see everyone, too? I’ll probably go back and dig a little deeper on the family side since I haven’t deleted the data yet.