I’ve been self hosting for years, and am familiar with many of the topics here, but it’s still an interesting read for things like talking about breaking out the three part router yourself. I’m really glad he out this together because it means I can see what others do in detail, even if it’s NOT the 100% recommended way (OPNSense, wireguard, etc)

On one hand, I agree that having a small overview with links to make this non monolithic would go a long way to making this functional and less scary.

On the other hand some information is scattered fairly heavily. Take the switch discussion. He mentions a 15 dollar switch, and then the upper end 1000$ switch early on, to emphasize the range. It’s not until a much much later section he talks about the more practical 20$ switch or 400$ switch he’d use here. So it being monolithic aides Ctrl+F to find this segmented info.

He also mentions the capability/value of having a manged switch (the latter switch is managed) specifically with VLAN, and yet doesn’t to my mind ever state why/when I would do something with the switch management to that end. As far as I can tell, many newer switches will pass VLAN tags (even when unmanaged) from the router, which will enable you to offer a WAP with split SSIDs so you could use something like TP-link 8 port 2.5gb unmanged switch (which at 100$ seems like a meaningful bridge between the 15$ 4 port 1 GB switch, and $400 16 port 2.5gb, 8 port poe switch). He talks about PoE & speed merits but IMHO doesn’t really cover the significance of a managed switch other than saying it had features for vlan (even though the cheapie would pass VLAN tags)

What does the managed switch offer me for VLAN? Specifically just the capability to isolate certain ports so specific hard lines are mapped to a certain vlan?

deleted by creator

Edir: i see this was already mentioned.

Not sure if you meant the video, or written guide, but for the written guide -

OPNsense is not even mentioned.

When we build a router using a standard computer, we can install router software like pfSense or OPNsense,

There’s a bit of a debate between pfSense and OPNsense. TL;DR, the developers of pfSense are not the nicest people sometimes. If this bothers you, consider checking out OPNsense. Since I’ve been using pfSense for a decade, I’ve built much of my infrastructure around it. I am well aware of its quirks and don’t feel like setting up my network from scratch, so I am using pfSense for this tutorial. Regardless of the developers, you are infinitely better off using pfSense on your own hardware than standard routers.

But what are you both thinking you’ll archive?

I really like Symfonium. I’ll second giving it a check.

I love using it for Navidrome & offline caching, and it’s nice because it does offer some casting capabilities internally.

Anything you’d recommend over nextcloud for filesharing? I’ve had it setup so I’ve left it.

I just suggested something like it because OP wanted to they and their friends to be able to use the images.

Thanks! I’ll have to check it out myself.

It doesn’t help if folks are iOS, but perhaps ludicolo@lemmy.ml could set up a file sync server (e.g. nextcloud, etc) that syncs the files to their phones, and then an app like this could be used to pick images?

Or maybe there’s a way to set it up to read from a WebDAV mount point either in application or through Android?

I would say an image archiver like Hydrus might be suitable for an individual, but I don’t think it would have anything for your multiuser use-case.

I am curious if any popular keyboards accept custom gify/etc replacement URLs. It’d be pretty slick to set up Heliboard or Futo keyboard with an integrated self-hosted gify

I think jellyfin offers DLNA support? You could enable it localy and see if your TV(s) can access files that way, without the jellyfin TV app as a middleman

With WebOS app, I’ve noticed sometimes when watching TV shows, if I select one audio track, it will

• go to the next episode
• keep the desired audio track selected
• play the original/nase starting track.

To remedy it, I have to switch from the current selected audio track and back ¯_(ツ)_/¯

As far as possible solutions to bad app behavior on your device - get a separate TV device (roku, etc) or some pc or SBC run a jellyfin app on it. You’d want to research first to see if you could use your existing remote somehow (e.g. maybe an IR receiver)

Under the hood its mostly tables and reports, so ultimately not much, if you were dedicated enough to using Excel to rebuild GnuCash’s views. It’s more streamlined than excel would be because you won’t have to worry about implementation, overhead of adding a new account, etc. Some things like auto-recommending accounts during import (and import itself) could be arduous in excel if not supported natively. Split transactions could be a headache (think your paycheck, which might be split into 401k contributions, several taxes, money into your bank, etc).

But fully recreating it in excel when it already exists would be a headache. More than likely you will have a more limited view in Excel if you’re just creating a handful of tables to represent all of your many accounts.

I use GnuCash. I typically update every couple weeks up to a month. Beyond that it can be hard to remember what specific transactions were.

It’s double ledger and I really like that it forces strict accounting. That sounds cumbersome but once you’re set up (it may take some trial and error), for me my workflow is essentially:

• Copy prior paycheck splits & update them to reflect new paychecks.
• export QFX files from credit cards
• import QFX, check / set transaction accounts
• any small manual updates (interest payments in accounts, etc)

It’s not automated but my data always remains local, and I can use the Linux or android application. I don’t bother daily tracking on my phone, else it might be cumbersome. I’ve never used any of the budget features, just tracking where my money comes and goes.

I don’t use Trillium, but I’ve enjoyed Logseq. I use it with nextcloud to back up. It is not browser friendly as far as I know when self hosted. It’s primarily just markdown.

This sounds quite interesting!

Also symfonium

I realize it’s not a satisfactory solution, since you won’t be able to use facilities like artist/album/etc to explore.

But in navidrome I think you can make a smart playlist based on file path, which you could use to include all of YOUR songs in the playlist if it is in a different root path.

Anyone with Navidrome should consider trying this. It provides a means to create & modify the smart playlists graphically & makes it easier to rate songs with stars vs the web UI

Responding to myself…

Datasheet reports 7.05 idle watts (~11w at active random read) so depending on what it considers idle, it’d be 8*7.05|11= 56.4:88W

Server clocks in at ~102W. Halving the drives would reduce the power by 27 : 43%

And in theory other components (motherboard, CPU…) must be using anywhere from (102-88) :(102-56.4)= 14 : 45.6 W.

Thanks! I need to look more into what the power implications of 8 drives is - they never spin down, so I assume they are a non-trivial portion of my power consumption.

That said, I’ve been considering upgrading to something recent and low power anyways. It would be a good opportunity to sneak in some useful features too,

• Maybe the possibility of transcoding a video stream
• USB3 (not a huge deal)
• Non VGA display (useful, for when connection issues arise)
• Audio jack (I could use navidrome jukebox mode!)

Which the old hardware wouldn’t support without adapters, cards, etc.

Proton, some of their paid exit nodes support P2P