It can often improve performance and memory latency and usually only has a minor CPU performance impact as it trades cycles waiting for memory for cycles decompressing memory. It is usually decent even on low power embedded devices.

There are a few edge cases where ZRam is not great. If your data is already compressed or encrypted copying it around in memory is much more expensive. It’s also harder to tell exactly how much data can be loaded into the “free” memory. It’s also a bit slower for serialized memory access in large data sets if the compression ratio is low.

Right image, but under those each one below would also be wearing large pants covering each side of the subtree.

Not necessarily without concern. Some containers have startup scripts that chown or chmod all files in some locations. It can mess up access for other containers if shared.

First will be shared files between the container in a single named volume. The others will create 2 named volumes pointing at different files with example1 from the 3rd not being on NFS.

It’s not well explained for sure but judging by the names of the cookies I bet those store the consent (opt in/out) values for the other tracking options. Another way of putting it would be those are functional cookies related to the cookie consent form itself so that you don’t have to re-select consent options every time you visit the site.

From what I’ve read is not authentication bypass, it’s a RCE using certificates to deliver the payload. If a specific signature is found it runs the code that was sent in place of the signing public key. It also means that only someone who has the ability to generate that specific key signature could use the RCE.

There were some other bits that looked like they could have been placed to enable compromising other build systems in the future when they checked for xz support.

If filesystem UUIDs are IP equivalents. Then device paths are MAC addresses. FS labels are DNS. Device mapper entries are service discovery.

The reasoning is that it is not illegal to fake most student ID cards but it is a federal offense to fake or alter government issued ID documents.

That way if it becomes an issue they can just pass it on to the authorities as their problem.

“Invalid” or “unparseable” are more understandable descriptors in normal language. I don’t think I ever heard of garbage/junk being used for that in language theory but it may be domain specific usage.

There are a lot of edge case characters around visually indistinguishable names. If that is a concern usernames should use a restricted known character sets instead of trying to block specific characters. You likely should also treat lookalike characters as equivalents when checking for username overlap.

As someone who also has produced code that looks like random characters spewed onto a terminal while using fpdf, I feel this one.

It can still have issues with potential attacks that would redirect your client to a system outside of the VPN. It would prevent MitM but not complete replacement.

Likely you needed to include the intermediate cert chain. Let’s encrypt sets that up automatically so it’s quite a bit easier to get right.

There is also SMS passive reading using LEO intercept. Hacked police email accounts are used to gain access to carrier systems where they use “imminent threat” no warrant lookups to pull the SMS in real time.

SMS is a terrible form of 2FA, better than none but not by much.

Your experience may depend on which distro you use and how you install things. If you use a distro with a stable upgrade path such as Debian and stick to system packages there should be almost no issues with upgrades. If you use external installers or install from source you may experience issues depending on how the installer works.

For anything complex these days I’d recommend going with containers that way the application and the OS can be upgraded independently. It also makes producing a working copy of your production system for testing a trivial task.

I’n Windows it is not stored in a keyring but instead in the registry. This has basically the same security threat model as a local key file.

The ssh-agent on Linux will do what you want with effectively the same security. The biggest difference being that it doesn’t run as a system service but instead runs in userspace which can make it easier to dump memory. There are some other agent services out there with additional security options but they don’t change the threat model much.

Initrd contains the systemd binary and enough libraries, services, and kernel modules to get booted this far. The system failed at switch root which is where the real root disk is mounted. Initrd can contain as much or as little as needed to get a working system which can be a lot of you are using a network filesystem as a root for instance.

My memory of the cp command is that attributes such as file times were transferred at the last step. I think this would make rsync safe in most situations where a system crash wasn’t involved.

I think I remember running into that as well but for whatever reason I couldn’t get accelerated-x working with the opengl libraries I was using for school. Likely the issue was just a lack of understanding on my part as I don’t think I had a good grasp of the Linux library loader until well after I graduated.

I’ve had a system in the late 90s with a 3dfx voodoo card. Also had a laptop with a SIS card from the early 2000 era.

The voodoo card was THE card to have it it’s day (mine was an older second hand system though). The SIS card… for some reason they decided that standard VESA mode probing wasn’t a thing they supported and would hardware crash when that API was used. I eventually got it working in Linux after patching xfree86 to not attempt probing when loading the VESA driver.