On Linux, it completely captures your physical keyboard (meaning it doesn't work for other applications anymore) and creates a virtual keyboard using the uinput kernel module [1]. The configuration is for how the keys on the physical keyboard maps to the virtual keyboard.
It should be able to do whatever you suggest. But it can be a lot of configuration. I use 3 different layouts in 2 languages and that will be too much to configure manually. Instead, I just use 2 layers that are mostly just pass-through (they just pass the keystrokes to the native xkb layout). Layer1 is entirely pass-through except for a single tap-dance key to switch to layer 2. Layer2 has some additional features like home-row-mods that are not possible with plain xkb. The keyboard layouts are managed by xkb. Thus I'm able to use the features of both xkb and kanata.
When I began studying physics I felt like a clueless novice. All those equations looked like chicken-scratch on paper. And, "What's the difference between kinetic and potential energy?" And there were so many concepts to understand - forces, masses, fields, charges, and particles. And differential equations seemed incomprehensible. Now, after a few decades I've mastered most all of that. Yet, when I try to grasp the real basics, like "What is space-time?", "How does superposition lead to a single macroscopic universe?", "What is the Higgs field?", "How did this universe 'start'?", "What exactly is mass?", then I realize I'm really back at novice again. I know almost nothing.
I wonder how hard it would be to write a program to automatically refactor code like this into something reasonably clean. My intuition thinks it should be difficult but possible... and extremely useful.
There are linters which do a bit of this, but they only cover the lints they have rules for. I'm pretty sure no one has a rule for "all loop indices declared as class members" because no one writing code of this quality has yet used a linter. Many issues in this code-base are not tractable at all for automated repair: how are you going to rework thousands of line long methods into well-factored classes and methods?
ReSharper (and presumably IntelliJ) supports expression-based search and replace.
Here are some commits where I used it to remove boilerplate from some test cases. The code was written long before MSTest added support for Assert.ThrowsException.
Organizing customers against companies that try to exploit their customers is an awesome concept. I suspect it will be a battle, but I wish you success, and hope you earn enough to make the effort worthwhile.
What about expanding this?! Some companies frustrate me with their customer abuses, and you could organize customers against them.
- Cell phone companies send bills that are confusing gibberish so we won't read them. So, provide a service that scans my bill and sends me a readable summary with red-lines on anything surprising or changed.
- Banks and credit card companies send 'privacy statements' that are rambling non-sense. Provide me a service that scans my statements and gives me a nice summary of anything surprising.
- Companies send endless offers for loans or useless services (paper 'spam') mixed in with bills. Filter this out for me.
- Alert me to hidden fees.
- Read the legalese and summarize it.
Basically provide me a service that I will try to accept my bills and statements from monopolistic companies and give me a short, readable, reasonable summary. When something is unreasonable, help me protest or fight it, or find an alternative.
Something like a consumers union, I suppose.
I'd gladly pay well for this.
I use a mix of setxkbmap to remap the control key and AutoKey but each has its issues. Can this replace those?