Or how bumblebee did an “rm -rf” on uninstall without a quoted path, which ended up nuking important directories

Server or desktop, and what types of files? I find that a self-hosted version of NextCloud does pretty well for keeping contacts, images, and videos in sync.

(You could run it on a Pi as an intermediary to both if desired)

I used to use stuff like AndFTP in the past for similar functions

I’ve been fairly happy with Alfred locks using ZigBee. There’s still a Bluetooth+App component to those as well for adding choices but you don’t need it to use ZigBee for locking/unlocking or viewing status.

Don’t use the wifi bridge though. I briefly tried that and it connects to some address hosted in AliCloud of all frigging places

In some cases a wipe/reset of the TPM from the BIOS might do it as well, is it’s still functional but scrambled

This is actually one thing I’ve been thinking AI and deepfake tech can potentially do good. Let’s say you have an idea and can code… You have an idea for music but no instrumental talent, so the best you can do is hum it. You can’t afford voice actors or other professionals.

Or maybe you’re artist with an idea who can storyboard but not code. Maybe you can make 2d designs but not 3D models, or aren’t great at animate.

But… there is software that can take what you say and change it to a different voice. It can animate a model to match the words. Similarly, software that could generate instrumental sounds from humming is possible. An AI can generate interactive dialog. It could also provide assistance in the generation of music, debugging of code, and eventually more advanced 3D modeling.

A lot of game design software is much more a GUI to an environment/model and triggers etc than stuff like writing hardcore backend C++ code etc. AI could take that even further.

Then add VR. Drop somebody into a blank-slate where they can create a whole world with a word, a gesture, and a great idea.

One day, that might be a reality.

Huh? Is the previous poster an OpenRGB developer? That’s cool!

Thanks. I’ll check into it but TBH I do really prefer .DEB based distros and that one seems to be Fedora based

That’s actually not what I was referring to.

First of all, RedHat now belongs to IBM, and they’ve never been shy about squeezing customers for a buck.

Second, having dealt with their support, it’s hit or miss to get a somebody helpful or an endless cycle of tickets. Patching and versioning is sometimes a complete mess.This especially sucks as the main reason most organizations go with RH versus others is for patching and support.

There’s also a lot of things where there’s a RH-specific implementation , which is further distancing fun other Linuxes and often ignores standard ways of configuring things.

RedHat actually benefitted from Fedora, CentOS etc as it allowed the community to develop products in a way that could be tested to be reasonably compatible, and to develop our port back fixes etc. It wasn’t just “RedHat made this and others just took it” but in many ways a symbiotic relationship. Yeah some orgs just went with CentOS but often it was those who worked on RH corporately would run CentOS at home in order to have a similar environment.

I used to be “Debian on the server, Ubuntu on the desktop” but recently I’ve spun up a few Debian boxes for desktop and I’m pleasantly surprised.

Kinda wish Valve would go for a full-out supported distro that stays in step with the Deck for Linux gamers (the old desktop SteamOS is kinda abandoned from what I can see), among with making the deck frontend a supported desktop manager. It would make sense for them to do so and rake in the game sales whilst providing a well-supported platform without the shit others are doing.

Increasingly so, and following the path that RedHat was taking prior (and probably worse to come given their new ownership)

Anyone pushing you to do something you don’t understand, or understand poorly. I could see an actual security researcher pushing for a code update to fix a vulnerability.

Heck, even as an occasional contributor I take some pride in seeing my fixes etc make it into the mainline codestream.

But yeah, you definitely need to be wary of somebody you only know from online pushing a change that doesn’t make sense or you don’t understand.

Oh yeah I forgot to mention that. It’s important when using wildcards or recursive permissions!

Tip: you can also use chmod u+rwx,g+rx,o+rx etc to add permissions

With the initial letters corresponding to “user”, “group” and “other”, and ®was, (w)rite, e(x)ecute for the rest.

In the case of directories, x specifies access to files/etc within the directly (read just let’s you see them)

You can also use i.e “o-rw” etc etc to remove existing permissions

I have similarly annoying issues when one of monitors is turned off. It collapses everything back onto the other display and doesn’t reset when I turn it back on.

So maybe an issue with the display going to sleep?

On Linux, you don’t download random stuff from the internet, e.g. a new browser. You get it from a central source, usually package manager, where it is verified and secure

Devs tend to make strong use of packages on GitHub, PyPi etc which have been targeted quite a bit with malware. Malicious snaps and

Linux software is written to need only as many permissions as needed, but not much more.

Hooboi. Depends on who writes the software. There are plenty of dumb devs for either OS, and I’ve had to yell at many for requiring their commercial software (built in Java with an X11/web front-end and exposed listening ports) run as root, usually because they didn’t want to figure out the permissions needed to access a device. There’s a surprisingly narrow intersection of devs who understand OS security and networking.

Linux is usually always updated because of the central update mechanism, so that vulnerabilities are fixed very quick

For OS packages, sure, but are all your Docker containers, snaps, flatpaks, and appimages updated whenever one of the underlying libraries had a significant vulnerability? How about that PPA, or the stuff you compiled from source a year ago?

Because people are increasingly using those for software not available on the base repositories

Linux users often have a false sense of security that leads them towards insecure practices, often for the same reasons as Windows users (I just want it to do X and work). While traditional signature-based antivirus doesn’t help much for either OS, there are plenty of other controls to fill the space that most people/organizations can - but don’t - implement on either OS.

On Linux, that includes strict management/review of software+code sources, SElinux/AppArmor enforcement, remote logging+review, and much more. These often conflict with Linux devs idea of “freedom” and thus area a hard sell.

I see you noted how you’re running gnome and what video card, but not how you’re running Steam, so I’ll ask:

How is Steam installed on your system? I know some people use a flatpak-based install but one of the potential issues with Flatpak (also Snap) is permissions to certain locations or devices can sometimes require extra config.

If you’re currently running from Flatpak, perhaps try the direct install from .DEB instead

It depends on where the encryption data is stored. If the bootloader and bios/efi are locked down and the data to unlock is stored in an encrypted enclave or one is using a TPM (and not an external chip one that can be sniffed with a pi), that’s a reasonable protection for the OS even if somebody gains physical access.

You could also store the password in the EFI, or on a USB stick etc. It doesn’t help you much against longer-term physical access but it can help if somebody just grabs the drive. It’s also useful to protect the drive if it’s being disposed of as the crypto is tied to other hardware.

Even just encrypting the main OS with the keys in the boot/initrd has benefit, as ensuring that part is well-wiped makes asset disposal safe®. Some motherboards have an on-board SDCard or USB slot which your can use for the boot partition. It means I don’t have to take a drill to my drives before I dispose of them

Not too many users, but an ever changing variety of devices and services :-)

Update: Based on some other sources, it sounds like giving another shot at freeIPA might be worth investigating. It’s still got Samba etc and the last time I tried it things weren’t more RedHat exactly friendly to my favored flavor (Debian) but it sounds like it might be better supported now

Update #2

OMFG it’s years after I tried and FreeIPA on Debian is even more of a pain. Docker container issues galore, and it basically won’t start without adding a bunch of options that reduce the container security to a smoldering ruin

I haven’t played with tailscale, and most of my wireguard shenanigans have involved connecting to others’ systems. Wouldn’t those mostly control the network-level access but not the account-level access (centralized account/UID/gid and remote permissions) part?