Yea, that’s a challenging part for sure, one that I still deal with.

It can be done, but you need to configure the sync jobs “just so”. Send only from the phone, don’t sync deletions, etc.

In the past I’ve setup a master folder on my phone and used subfolders for “upload only” Syncthing shares. Then wherever I put a file on my phone it gets synced to the appropriate folder on my server.

Its really to compensate for the lack of framers.

Five years ago the average age of a framer was 55.

This is what happens when you don’t have a new generation of people trained to do something - constructors have no choice but to use automation.

I’m not blaming anyone - its just an observation of pressures. Framing’s a tough job.

There will be massive outlays for the systems, they’ll probably be leased or you’ll have companies that specialise in managing the system, and as a GC you’ll contract them to implement the design.

Does it have to be SFTP?

I use both Syncthing and Resiliosync - both are better on battery than conventional file transfer mechanisms, and they don’t require babysitting. They simply sync files as you define.

Alternatively if you must have FTP/SFTP, I have used FolderSync (Play store or Apk Mirror) since about 2010 - I’ve even bought it twice because it’s such a great app. Though I haven’t upgraded in a while because I don’t like the new UI.

The free version does pretty much everything the paid version does, and the devs have always been great.

But I’d use Syncthing, you can even set it to only sync on your wifi and while charging, though I’ve found even syncing a few gigs doesn’t impact my battery (I probably sync 10GB a day while on battery).

Even without a VPS Tailscale will work fine after the router resets.

The more open ports, the larger the attack surface.

That’s all.

And today with the script kiddies out there, port scans happen all the time.

I’ve had a consumer router become almost useless from all the attempted connections on an open port someone found that I had up for a week.

Months later I’d still get hits on that port though it had been closed.

Yea, Tailscale would work even if the router was fully reset.

Without a secondary internet connection this isn’t possible.

The router is the connection - its the gateway (a term we don’t hear much these days).

You could setup an independent connection via a cell modem - becoming a secondary connection. This is common for remote locations or even small businesses that need a failover just for management.

You could even have it on a single machine and have a vpn there. Then you could RDP/VNC to that one machine and manage things from there. I’ve done the VPN this way with Tailscale. One machine has it (I’ve even done it with a Raspberry Pi), then you can RDP/VNC to other machines from there.

But there’s not much I could see you doing if the gateway is down anyway.

Its not “targeted at old school”, it’s an open, extensible protocol.

If devs focused on extending the protocol instead of building an app to handle things like this, it could do it, everywhere.

There are currently over 100 extensions.

Still staggering to me that XMPP isn’t the default, since it was used in many chat apps in the late 90’s.

I can’t remember the name of an excel spreadsheet I created years ago, which has continually matured with lots of changes. I often have to search for it of the many I have for different purposes.

Trusting your memory is a naive, amateur approach.

Its spec’d for that

I dunno, my newest cable modem chews up a boatload of power for no good reason - really pisses me off.

It uses more power (50w) than my 2019 SFF desktop I run as a server, with an 8TB drive (20w).

I know because they both sit on their own smart switches.

Better yet, LemmyShemmy should ask his mom - no reason to do his work for him

*California

Why would I?

I use a mesh vpn to manage my servers otherwise, what’s the difference?

Hell I RDP to them all the time (or SSH) - if anything it’s providing another layer of encryption.

I’ve re-read the docs around VirtualBox bridging, and the only thing that I think could possibly cause this (though it shouldn’t since ping is a direct address to a specific IP/mac) is to enable Promiscuous mode (as cappucino mentioned).

This shouldn’t fix it in my opinion, but it may, since ICMP could be considered undirected traffic (I don’t, but I didn’t code VirtualBox).

I’ve sometimes found using trace instead of ping can sometimes work where pings would be blocked, plus trace shows exactly where it dies along the path.

NOW you decide to leave the shitstain that discord has been for years?

Like you were OK with all the other crap it was doing?

Right? Reading that was like watching an action movie!

Hahahahaha

I’ll be over here crying in the corner.

Next you’re gonna mention ICQ

I’ve used it a couple times, what a dumpster fire of an app.

Walked away years ago and refused to use it since.