I've used Ubuntu and have enjoyed the luxury of the gnome keyring prompting me for my user password instead of my ssh key passphrase.

Then about a year back, I switched to Arch with KDE Plasma, and I enjoyed the same luxury with kwallet.

Now, I've ditched the majority of KDE Plasma (don't ask why the uninstall was so messy...) in favor of sway and a very minimal riced setup with as little GUI as possible. As I mangled my KDE Plasma packages during the KDE removal process, I've done something to make kwalletd6 not start up on boot. PAM something? Don't know, I don't know anything about PAM or D-Bus yet... Everything still works as intended, I just need to launch kwalletd6 before I try sshing anywhere.

But this leaves me at an impasse. I don't "love" the GUI pop-up when I ssh, and I don't use my wallet for anything else. I'd like to get back to a state where I can just use my user password to input the various passphrases for all of my various ssh keys. To get back to a state like that, I could do one of the following:

1. Get kwalletd6 to start when requested once again. This isn't my favorite solution, because I've actually soured on the KDE Plasma environment a bit. I love the idea of completely getting away from KDE. But that doesn't mean I want to run right into the arms of GTK with a seahorse-like solution.
2. Make a custom PAM solution. I see on the Arch wiki page for pam_ssh that you can do some fancy PAM things to add ssh keys to the ssh agent, but that wiki page does have a pretty scary "if you screw this up you'll need to recover your Arch from a backup CD" warning. Although I'd think you could just boot on a USB and fix the file without having to boot into the broken arch OS? Right? Anyway, beyond this scary warning, this also still involves messing with an etc file that would mess things up if I didn't have the pam_ssh module installed anymore. I'd want to couple this custom pam_ssh file to the pam_ssh AUR installation so that if I ever uninstall the pam_ssh package it would also remove my custom pam file.
3. A custom script that works with an encrypted file containing my passphrases. I'm thinking a gpg-encrypted file that encrypts the file with my user password. I could set up the SSH_ASKPASS environment variable to point to a bash script that, upon first invocation, decrypts the passphrases file and loads them into ssh-agent. First of all, I'm not actually sure this route would work. Second of all, if I leave the passphrases on my computer and encrypt those passphrases with a different passphrase, have I actually done anything more than just locking my ssh keys with my user password? I guess I could leave this encrypted file in one of my separate luks encrypted partitions... maybe that means this would be a better route? Lastly, it sounds like if I go with this option, I'm basically just handrolling my own password manager. That could be a bad thing since I'm not a security expert, but it could also be a good thing because it would be a very simple/uncomplicated solution that relies almost entirely just on reliable/ubiquitous programs such as ssh-agent and gpg. I like this option because there would be no added software packages. Just a simple script and an ecrypted file.

I know programs like KeePass exist, but I don't know if I want yet another password manager.

So what do we think? Am I crazy person? Am I missing a really obvious solution?

EDIT (Result): I ended up going with GNOME Keyring. It wasn't as minimalistic of a solution as I wanted and it did make me feel like I was taking one foot out of a KDE ecosystem and putting one foot in GNOME (which I wanted to avoid). But it was so easy; everything worked great out of the box. It was also a pretty minimal installation dependency-wise unlike many KDE components. It does have a GUI prompt rather than a terminal prompt, but maybe that's something I could figure out how to tweak one day. Lastly, while it is more feature rich than I was hoping for... I kinda like the features... Thanks to everyone for the discussion!

Oh, and for anyone stumbling on this in the future, I think I was on the wrong track with pam_ssh. The more I dug into it, it seemed less like a way to unlock keys to be used to ssh into other machines and more for unlocking authentication features if you were sshing into this machine. I could be wrong though! Still happy to discuss in the comments

I'm trying to set up my Advantage 360 (not pro!) and I would like to swap the Fn and Ctrl keys. I'm on linux, so I don't know any good ways to remap the layout besides mounting the "v-drive" and editing the layout files. What I tried was this:

```

[lfn]>[lctr]

[rfn]>[rctr]

[lctr]>[lfn]

[rctr]>[rfn]

```

This successfully remaps the Fn keys to ctrl, but it doesn't make the Ctrl keys Fn. So I have no way to use the F1-F12 keys at the moment...

Any thoughts? Also, is there any good way for me to edit the layout besides mounting the v-drive and editing the layout files?

I want two things (the first more than the second):

1. When I'm on the left-most window of my right monitor, when I keep trying to go left with super+h, I want it to navigate to the right-most window of my left screen. And vice-versa.
2. When I select a workspace with super+1 or super+2, I want it to shift both my monitors to that workspace.

My motivation: trying to keep track of which monitor I want to use is tedious. I'd like to think of them as one continuous monitor.

Possible solutions (maybe? I'm very new to this):

1. Create some sort of virtual display comprised of the combination.
2. Scripts with context like "you're trying to go left, but you're on the far left of the right window. Switch to the right side of the left window"

My starting point is video with AC3 5.1 and TrueHD 5.1. I watch movies on my smart tv with stereo audio and sometimes on a projector with a stereo soundbar. However, one day when I upgrade to a 5.1 setup, I don't want to have lost that original 5.1 data. I have plenty of storage, but I do want to compress as much is as sensible.

My inclination is to output two audio tracks with the video: one AC3 5.1 passthrough, one Dolby Pro Logic II (used for stereo) made from that original AC3 5.1 track.

My questions:

1. Can most stereo speaker setups (smarttv, chromecast into a dumb tv, etc.) use Dolby Pro Logic II for stereo without my server having to transcode live?
2. Same question, but for AC3 5.1 to stereo. Am I wasting my time adding that second "stereo track"?
3. When transcoding that track to Dolby Pro Logic II, what's a good option for DRC? I love having strong compression, but I don't want to introduce any weird artifacts to the track (if that even happens). I've assumed I want to go for 2.5?

I've been pursuing constant HDD writes on my Truenas system. In this chase, I've shut down all services (minus ssh) and I've unset my app pool (no apps are installed or running). A quick look at my setup:

One RAIDZ2 pool conisisting of four 2TB HDDs. One 1TB HDD that has the TrueNAS Scale installed on it.

That 1TB HDD that has the OS installed on it is experiencing constant writes (every 5 seconds). I've found that those writes happen every 120s as opposed to 5s when zfs_txg_timeout is changed from 5 to 120. So it looks like these are txg syncs.

Now for the questions: I have no idea what txg syncs are. I have a loose understanding that zfs works by collecting the proposed changes to a pool, then contains those changes in a "txg", then commits those to the history of the pool. However, and here is where I am confused, I'm making no changes to the pool! Why are there constant txg writes, and why are those being written to my OS HDD as opposed to my pool (the four 2TB HDDs). Not that I want it written to the pool, I'm just confused.

Ideally, if I have this NAS just sitting around not running any apps or services, I'd love to have it just sort of "live most its life in a sleep state" with no HDD activity minus what it needs for syslog type activity.

​

EDIT: I think I'm starting to get it. I forgot that I'm booting from a ZFS pool (my boot pool). So is it probably logs and other things that would otherwise be immediately written to the hard drive, just collected and written to the system hard drive as a txg?

EDIT EDIT: Figured it out. I took a snapshot of my main pool with `sudo zfs snapshot -r mypool@diagnose`, then I could see what was being actively being written to with `watch -d -n 5 sudo zfs list -o name,written`. It was all downstream of "myPool/.system". So it turns out that I had my four 2TB HDDs (myPool) designated as the system dataset partition. I changed that to my boot pool (the one 1TB HDD), and the constant writes to the four HDDs has been fixed! I have also changed my zfs_txg_timeout back to 5s, because I don't want writes to the root partition to only happen every 2 min (sounds scary)

Is there any way I can set up Shinobi so I can give a url to someone on my home network, then they navigate to that url on their mobile browser to view a live stream of an ONVIF camera on my network? Ideally, it would be nice if they could do this without logging in to an account.

I thought I was close with the API keys, but I couldn't get all the way there...

Can I configure Frigate to serve a live stream to clients on the network that they can view in their browser? I'm trying to make an RTSP stream from an ONVIF camera available to guests on my network so they can view this live stream via the browser on their phone (audio is needed, too)

I'm trying to use a vulkan GPU driver on my Raspberry pi to assist with my AgentDVR server. However, when I try to set "Vulkan" as my decoder, I get the following error in my logs:

​

SetupHardwareDecoding: Preferred decoder (AV_HWDEVICE_TYPE_VULKAN) does not support the profile. Using AV_HWDEVICE_TYPE_CUDA

​

Is there anyway I can troubleshoot why AgentDVR does not want to use Vulkan? (and yes, I have installed Vulkan via the guide here: https://qengineering.eu/install-vulkan-on-raspberry-pi.html) It also does the same thing with DRM.