How did you mount it outside the cluster? Did you have a look at the mtab and used the exact same options in the compose file?

There’s no difference between using a volume in Compose to mount a share or your server’s fstab file. Both do the same kind of mount.

I’d suggest /opt/docker/_compose/ for all the compose files. Or, if you keep all the config files for your containers on your NAS, maybe create a share there and put all yml files in it, then mount it on the host. This way everything is on your NAS and nothing is lost if the host freaks out.

And I’d add the NFS mounts to the compose files as well. When specifying volumes, you can use anything the host OS has a mount.xxx command for. Docker will take care of mounting everything.

Put that mount point into the compose file(s). You can define volumes with type nfs and basically have Docker-Compose manage the mounts.

That probably doesn’t work unless you power-cycle the picture frame after changing the photos.

I had this with some offline Samsung picture frame and a Transcend WiFi SD card. The SD card runs a small Linux and can be unlocked to add own scripts. I had a script that would rsync files from my storage to the SD. However, while the new files were written to the SD just fine, the picture frame never re-read the list of files from the SD. And after power-cycling, my specific model needed to be turned on manually again. So, that wasn’t a satisfactory solution.

Following a profile logged-out is impossible now

What do you mean? I can just open an Incognito tab and go to x.com/<username> and see all posts (without replies, though).

because they don’t run alts

I think you underestimate the dedication of some of those trolls. Also, most apps allow to easily switch between profiles with like 2 taps.

Of course, not. But closing and locking the door doesn’t prevent the person on the other side to still listen in on your conversations…

And it’s exactly like this now, if I understand the change correctly. They only removed the “you can’t see this post because the owner limits who can see it” thing. Blocked people still can’t reply.

Yep, that’s why I don’t get all this panicking about the Twitter change…

In closed systems like messengers, where you don’t see any content unless logged in, yes. There, it works brilliantly. But on Twitter, this is like cutting out something from a newspaper when there’s a news stand right next door.

Having a public (i.e. not locked) Twit𝕏 account and believing you can block single people is a bit stupid to begin with.

When screaming on a market square, you can’t demand for single people to “please not listen” to what you’re screaming.

Rather people have no idea how blocking on 𝕏 worked/works. You were ALWAYS able to see tweets from people that blocked you by simply logging out or using an alt account.

I don’t understand all this fuss about this simple change. He only removes a useless feature that was never more than a minor inconvenience for those that got blocked.

If you don’t want people to see your tweets, lock your profile. This worked before and this still works just fine.

There was a discussion about this topic on Hacker News a few months ago:

https://news.ycombinator.com/item?id=40133976

One ex-Googler pointed out that due to the machine learning stuff and every new employee trying out the latest “AI” stuff on top of it, no human can understand and thus debug the search engine properly anymore.

I’m using UberSpace for 5€/month for a few small web projects and for emails. Unlimited mailboxes, unlimited aliases. However, you have to configure it using console commands via SSH. But it’s all explained in their documentation.

If it’s the system with the (locked) KeePass database on it, you should be fine. The encryption can be tweaked so that unlocking the database takes a second even on modern systems. Doesn’t affect you much, but someone trying to brute-force the password will have a hard time. It also supports keyfiles for even more security.

If somebody infiltrates your end user device, no password tool will be safe once you unlock it.

After trying them all, I’m back at having a local KeePass database that is synced to all my devices via iCloud and SyncThing. There are various apps to work with KeePass databases and e.g. Strongbox on macOS and iOS integrates deeply into Apple’s autofill API so that it feels and behaves natively instead of needing some browser extension. KeePass DX is available for all other platforms, and there are lots of libraries for various programming languages so that you can even script stuff yourself if you want.

And I have the encrypted database in multiple places should one go tits up.

Yeah, but I didn’t want to fiddle with some custom settings. The same official postgres container works great with other apps.

I didn’t notice any big drops in network or CPU performance. Usually, because other network traffic had priority. But my server’s HDD constantly rattling along got me thinking that it wasn’t worth it. There are several other containers running on that box and I don’t have that much HDD activity with them.

I did this for a while. However, after subscribing to several groups, there was constant disk activity and it ate network bandwidth. After two months I’ve stopped my server and went back to using a public instance.

You might want to read the recent blog post (linked at top) and discussion on Hacker News first.