While true, this is after the army had already made its way into the area by opening fire on the protestors. The protesters, after being confined to the square, didn’t see much point in resisting further, as their leadership fell apart and there had been plenty of casualties already. Right above the section you quoted:

At about 10:30 p.m., still being pummeled by rocks thrown by protesters, the 38th Army troops opened fire with live ammunition.[176] The crowds were stunned that the army was using live ammunition and fell back towards Muxidi Bridge.[176][179] The troops used expanding bullets,[11] prohibited by international law[180] for use in warfare between countries but not for other uses.[181]

The advance of the army was again halted by another blockade at Muxidi, about 5 km west of the square.[182] After protesters repelled an attempt by an anti-riot brigade to storm the bridge,[175] regular troops advanced on the crowd and turned their weapons on them. Soldiers alternated between shooting into the air and firing directly at protesters.[183][173][182] As the army advanced, fatalities were recorded along Chang’an Avenue. By far, the largest number occurred in the two-mile stretch of road running from Muxidi to Xidan, where “65 PLA trucks and 47 APCs … were totally destroyed, and 485 other military vehicles were damaged.”[37] Although troops advanced into Beijing from all directions, the majority of deaths during the night of 3 June occurred around the Muxidi area.[184][185][186][173][187][188]

Throughout the street fighting, demonstrators attacked troops with poles, rocks, and molotov cocktails; Jeff Widener reported witnessing rioters setting fire to military vehicles and beating the soldiers inside them to death.[189] On one avenue in western Beijing, anti-government protestors torched a military convoy of more than 100 trucks and armored vehicles.[190] They also hijacked an armored personnel carrier, taking it on a joy ride. These scenes were captured on camera and broadcast by Chinese state television.[191]

In the evening, a firefight broke out between soldiers and demonstrators at Shuangjing.[192]

Obviously this was far from a peaceful protest, with protesters attacking and killing soldiers after said soldiers were ordered to make their way to the square, and some of the protest leaders explicitly calling for bloodshed.

I don’t know how I would’ve handled it personally, as I lack the cultural background needed to properly understand the cause of the protest, but it feels disingenuous to call the dispersal “peaceful”. The fact that the government hides the official death toll also doesn’t help their reputation.

Yeah that’s true, in the case of just running a premade compose file sqlite is the better choice for sure

Oh lol I see, it happens

Performance isn’t the only advantage to a full postgres deployment. I have a central database for all of my self hosted apps which makes it really easy to back it all up.

I’ve had a lot of problems in the past from software crashes that left sqlite files in a corrupt state, backups where the sqlite file wasn’t properly closed leaving it in a weird unlockable state, transactions not completing when swap is used, etc. Besides that sqlite really doesn’t play nice with NFS, which is the basis for quite a few cloud storage providers.

“Best option” really depends on what self hosting looks like in your specific setup.

Why would you need this for your scenario? If you’re not downloading you can simply check the listenbrainz recommendations in the playlists that it creates for you

Something that works as a recursive dns server. Unbound, Blocky and Technitium are some examples.

If you’re hosting nextcloud anyway you might also want to check out the notes app. It creates a Notes folder in your account from which you can edit the markdown files, has a nice markdown editor in the nextcloud web UI, and comes with Android and iOS clients so you don’t have to setup file access and find a separate markdown editor.

I initially generated the talos config with the template script, but using talhelper directly offers a bit more flexibility. The first four nodes in this list are the cm4’s

https://github.com/SquaredPotato/home-ops/blob/main/talos%2Ftalconfig.yaml

I also wrote down my setup steps, check here:

https://wiki.abchost.nl/s/cm4-talos-setup

Depending on what you’re going to do with the cluster, 4gb of RAM per node feels rather limiting.

Anyways, as far as storage goes I’m using 4 compute blades loaded with 4 8gb RAM versions of the CM4, each with a 500gb Samsung PM9A1 running Talos to save a bit on that precious RAM.

Got Talos up and running with some help from Onedr0p’s cluster template which saved me a lot of time on the learning curve.