I just saw your reply to me and was about to say the same thing, but they worded it perfectly. And I did mention metadata as a key point in my original post.
We literally have /var/log/ as a well-known standard though. Almost every piece of linux-standard software dumps to a subfolder by the app name in there. Systemd should at the very least have the capability to mirror there so you can get at the logs in a sane way.
Been using systemd for at least 6 year now, and yes it is indeed quite stable.
But making startup services is hot garbage, and accessing system logs is even worse. journalctl is an unapproachable mess, and I really don't like the idea that systemd is kind of slowly replacing the linux kernel in its entirety.
It doesnt affect my day to day as a normal user, but when I switch to power user mode its... It makes maintaining my system very unenjoyable.
Average journalctl avoider
Yes, it is possible!
I hope you have some solid soldering and hot air workstation skills, but assuming you do then yes its 100% possible.
I do not recommend it.
I think that part of the issue is that every disability is unique to some degree. No two people who are blind have blindness to the same degree. Colorblind people have several variations they could be experiencing. Deafness also follows a similar pattern, whereby some people can't hear well where others can't hear at all; or in some cases have a constant sound in their ear that drowns out everything else. There are neurological disorders that range from not being able to read sentences normally because the words start to become jumbled to being unable to focus on large amounts of text. There are physical disabilities of all sorts that affect the arms and hands or even the ability to sit upright to look at the computer screen.
Because of that, there are two options:
-
build a desktop environment custom-tailored to each individuals needs.
-
build some general purpose accessibility options that can (and must) be adjusted to meet an individuals needs, which may or may not be able to meet them 100%.
Out of the two, the second one is far more feasible, and more possible to improve upon.
EVERYTHING in Signal follows an encryption or tokenization chain. Not like crypto coins but real actual chain-of-custody type encryption workflows. It uses elliptical curve cryptography where the key for each message moves forward along an elliptic curve, which are excessively difficult to guess the factors for once it is selected if you are not the key holder. This means that even if someone cracks the key for a single message you sent, they are going to have to crack the key for every other message still as each one is different. Even the metadata is encrypted by the user's keys.
Signal doesnt have usernames in the traditional sense. It's phone number+6 digit pin hashed into an encrypted signature.
The signal company can't see anything you do besides account create date and last login date, even if they wanted to due to how their platform is set up.
Meanwhile, Matrix literally clones the metadata between servers when a user connects to and starts talking to users on another server, in plaintext (maybe encrypted at rest but not E2EE).
I really really don't. As soon as he does, corporate vultures (such as MBA degree holders) and people who "want to change what 'open source' means" will swoop in. If we replace Linus, I hope its someone very similar to him who isn't afraid to be a hardass where it's needed and will keep the current vision of Linux alive.
There is a process called "learning how to learn."
Essentially its a series of questions or question-templates you ask to gain deeper understanding of something beyond what the source material you are learning from provides.
The basic versions of them include "how" and "why", where the "who", "what", and "when" matter less in most cases (and if they matter more, it will be extremely apparent).
Then, listing out every answer to each of those questions you can possibly think of and testing them against logic and evidence.
"How do apples fall to the ground?"
-
Warping of spacetime?
-
flat disc that we stand on accelerating through space?
-
The pull of the aether?
You would then test each of these ideas by gathering evidence and eliminating others. Eventually in the example through rigorous testing, logic (including mathematics), you'd come to the correct conclusion of the warping of spacetime.
Then, if you are trying to come up with new ideas, you apply this method of thinking to the unknown and reverse it a little. Take facts and evidence that exist, and ask the same questions against the unknown. For example, if you're writing a story, the who and the what become important in the reversal, but the how and the why stay relevant.
"What would the main character do after they realize they were betrayed by the grand duke of blahshire?"
-
Who+Why: what are the main characters beliefs and convictions? What drives them?
-
How: What kinds of actions has the main character taken in the past? Have they grown since then?
-
What is the current sociopolitical climate that the story is taking place in? What kinds of behavior is expected of the main character and those around them? Are things like revenge frowned upon, or is not exacting revenge considered weakness? Maybe both depending on who you ask?
You can also apply this to general ideas, making yourself or a project/concept/etc. the main subject.
This kind of thinking is a muscle you have to exercise until it becomes natural. Eventually you start doing it without realizing.
That only works after you've found out the developer is a scumbag though
The syncing of keys allows for much greater attack surface.
Its being worked on right now but the standard hasn't been finalized yet.