https://github.com/LuckyTurtleDev/docker-images/tree/main/dockerfiles/anki

start with env var sync_user1=username:password or something like that

change server url in anki desktop or ankidroid to what you started in docker

done

oh my god what a stupid post

UPDATE: Turned out that the culprit of the downtime was my switch - the D-Link DGS-1210-10P rev. B1.

The way the management web interface of the switch works is pretty unintuitive. Namely, if you change some settings in the web interface and hit save in one of the sections, the settings are saved in the volatile memory of the switch. This basically means that the settings are only saved in RAM, which is cleared on power loss. To save the settings into non-volatile memory which persists on reboots, you need to find the “Save” section at the top of the UI. This is described here: https://networkengineering.stackexchange.com/questions/20158/dlink-switch-loses-configuration-on-power-off

So basically, my problem was that the settings weren’t commied to nonvolatile memory and on a short 1 minute power loss the switch restarted.

I got an UPS anyway now, SMT750RMI2U

As for backup, you can also buy a e.g. Lenovo M920q minipc, buy a pci-e riser, buy a dual port ethernet card, set up Proxmox, set up an pfSense (or OpenWRT, or OPNsense) VM inside, pass-through the ethernet card directly to the VM. The VM is very backupable, since you just copy the VM state and save it somewhere. This would only work for the router though, since the AP’s that’d be running OpenWRT wouldn’t be VMs. This is at the cost of having to deal with an additional layer for the VMs.

I guess the problem you’re asking about in regards in regards to cross-device portability of a backed up config is valid. If you had a four ETH port router, backed up the config, and then uploaded it on a two ETH port router, you’d run into trouble, but I have no experience here.

You can also install OpenWrt on some switches these days (PoE also reportedly works with realtek-poe module):

• https://forum.openwrt.org/t/support-for-rtl838x-based-managed-switches/57875/
• https://openwrt.org/toh/views/toh_standard_all?dataflt[Device+Type*~]=Switch

That way you’d have a fully open OpenWRT-only network lab, so you’d always be working with the same system.

Orgzly Revived on Android, if you’re into emacs org mode. Can sync over git, webdav and dropbox. Probably syncthing too. Samba over Android file protocol with Material Design sadly doesn’t work.

On PC just edit the same org mode notes as in Orgzly, but with Emacs.

https://github.com/orgzly-revived/orgzly-android-revived

You can also create an account and just leave a note on the map (a message that other contributors to OSM see) mentioning your observations. It’s a good option if editing the map is too overwhelming with all the different tags and how routing algorihms interpret them.

(I’m on New Game++++ now)

At this rate I’ll be having a Lemmy user superiority complex in no time

The first quote block refers to what is mentioned in the OP article, and the 2nd is an exaggerated summary of the parent comment.

My issue is that the parent comment is taking imo a lenient stance towards something vile happening

company makes remote workers ineligible for promotion

hey guys yeah it really depends on the job, sometimes you just gotta be in the office heh

did a realtor write this?

a decision was made in early 2024 to become a hard fork

https://forgejo.org/2024-02-forking-forward/

Complementary https://lemmy.world/post/13073457

I know people don’t like sending patches to mailing lists and prefer the zoomer PR UI but god damn if you look only at the protocol openness perspective nothing comes close. sr.ht is great in that regard