Faceman🇦🇺

for manga I’m still running FMD2 via wine in a docker container. it works, but it’s clunky and outdated. that feeds into Komga, which in turn is pulled by Mihon/tachiyomi on my devices.

for books, I just manually search usenet and public trackers. books i;m managing manually at the moment. used to use ubooquity but I dont read a lot so I move files to my boox tablet when I need them.

when they watch an entire movie in 2mbit 720p from a 4K HDR source and you know they have a proper TV and a shield pro.

there are only a small handful of devices that can fully decode FEL, regardless of the software you run on it.

That’s not going to change untill we manage to completely crack it and decode it in software or figure out how to bake it in via some kind of transcoder.

if all you need is a tiny NPU for things like image object identification or OCR the IGPU you already have will be better.

if you need slightly more than that, but not as much power as a large dedicated gpu, then you would have to look at a newer mini-pc with a more modern chip.

I do immich face/object detection and Birdnet audio analysis on the igpu of an intel 14600 and it’s fine, tiny amount of usage and plenty fast. my other node has a core ultra 235 and the igpu is faster than the on-chip npu in every regard, the npu is also a bit harder to actually get working and put to use.

I’ve been running the RC’s, smooth sailing in all regards.

Using Wholfin as my main client on all compatible devices, works great.

I run two synchonised Adguard instances on two different devices as DNS servers for the house (and for my work PC via vpn). on top of that I run local ad blockers on top of that. I dont care that my DNS ping times are a few ms longer. I dont want to be pestered, tracked, solicited or interrupted.

I cant stand the barrage of ads, especially on mobile websites, like its at the point where you want to read an article and you get two lines of text and the entire rest of the screen is ads you have to navigate.

I’ve always kept my apps and databases on solid state media, putting them on spinning rust, even in a well tuned ZFS pool with lots of disks is way too slow for their (admittedly inefficient) database.

It absolutely does but you have to do it yourself, they dont accept any liability for doing that and nor should they, so if you dont know how to do it with at least the minimal level of safety and security and accept that risk yourself then you really shouldn’t try. Ideally you would only do it via a private VPN to trusted users.

Plex has some features built in to make just opening the port reasonably safe, though I’d still recommend against that. They also have a reasonably secure account verification system, and the software itself has proven to be somewhat safe.

Jellyfin has very little security built in and is not intended to be public facing, it also has the double edged sword of being open source, so you could read through all the code and pick it apart for flaws and exploits if you wanted to and we all just trust that when exploits or flaws are found they are reported properly.

JF is great but you are on your own if you want to make it available via WAN.

honestly Wholfin being so nice to use was my last hurdle to getting the whole house over the JF from Plex, just a few remote users with uncommon players that still need Plex and once they upgrade I can be done with it for good after starting out with OSXBMC way back in the day and being a lifetime licence holder from the start of paid plans.

Most people just use Sonarr/Radarr for this, though it can be tricky with anime specifically because you are pretty much locked into TMDB metadata, which might not be what you want for some series (alternate orders, absolute or series numbering, season 2/3 being different shows entirely, side series being specials not a season etc. etc. etc).

I’ve used TinyMediaManager before, not open source (other than a very old version that is still out there last time I checked), but the free version is reasonably powerful, especially if you want to customise the details of your naming/folder structures or do custom metadata or ordering. has some CLI functions but is mostly intended as a gui app.

apart from the usual suspect like, Jellyfin+Plex (still have some straggler users that need to upgrade their players for JF to work properly), Immich, nextcloud etc, I really like running Birdnet-Go. I give it my CCTV camera feeds and it uses a NN to identify birds it hears.

I get statistics about species heard around the property, charts to correlate detections to seasons and weather, and some interesting info about rarer species that I didn’t know were in my area. Every morning I check it to see if there was anything new recorded.

it doesn’t really have a purpose, though the data can be submitted to larger databases, but it’s just a nice thing to have.

EDIT: Also Vaultwarden, forgot about that, super happy to be hosting my own password manager instead of trusting (and paying for) companies like Dashlane to host it like I used to.

there are a couple of projects I’ve seen, and surely people will fork off one of the open robo-vac projects to spin knives.

Robot mowers are in this grey area where if you have a big enough area to mow that you cant walk it in 15-30 minutes with a push mower, you could get a small riding mower for the same price, and now you also have a handy tractor too.

I like them, but that’s because i’m a lazy tech head, but just like high end robot vacuums they seem like one of those things that you spend good money on, they work reasonably well for a while, then break and are unfixable and the company behind them doesn’t help.

that said… there is a decent crossover in the region of “lazy tech head” and “enough disposable income” that would use a robot mower on a small lawn.

for a beginner with just a few remote clients, tailscale all the way.

though I still like doing it the old way with a custom nginx setup, fail2ban and a domain name, but its more work to make it secure and even then it’s still somewhat of a liability.

yea, very easy to fall back into the bad habit of blindly updating without giving the release notes a skim.

immich hasnt had any major breaking changes for a while in my experience using it for a few years, anything that does break is probably some third party integration or plugin that would have to wait for it’s own maintainer to fix, that’s pretty normal in the open source world.

Yea I changed over to v3 without issue recently too, I read the release notes and some of the issue conversations on github, so I knew what to expect.

That’s something a lot of people dont do, if you are self hosting an open source project it’s on you to stay on top of those things, stay updated as much as possible but dont do major jumps blindly without backups.

In my case, I went from a regular immich in a container to one with openVino built in for gpu accelerated image analysis and face detection, and I had to change my database container image. Then it fired straight up and everything seems to work 100% so far.

I have a stack of PCBs and some parts from an early revision but never got around to it.

that’s what the U3 is supposed to be, but I haven’t played with that newer one to know if its any good. I think its still not a full activity thing with power and input state tracking, but just a more advanced macro based system.

Would work for 90% of people though,

The U2, which is the one that is portable, I.e. no base station, doesn’t have activities, but you can fudge activity behaviour with macros. Its handy to just throw some codes onto it from the app and go.

The x1s and X do have activities but they are more expensive and need a base station.

The sofabatons are imperfect, but they’re as close to harmony as you will get