Kloak

When it comes to worrying about the data that can be used to identify us online, we usually think about the obvious things: tagged photos, hacked accounts revealing personal details, LinkedIn, exposed IP addresses, and other literal variations of our personal details. We don't often think that our keyboard can expose our identities – but it can, thanks to something called "keystroke dynamics." The way you type is as unique as your fingerprint. Only you have that particular nuance where the o key follows the h key like a ricochet, or where you always hit q by mistake when typing ask. We all have traits like these, and when combined and analyzed with enough AI data processing, they can be used to identify who is typing. This allows websites and other applications to track you across the Internet, across sessions, and even across computers.

kloak is a possible solution. It's an offshoot of the wonderful Whonix project, a Debian-based distribution built specifically to allow you to browse the Internet securely and anonymously. kloak attempts to obfuscate your typing style to make identification difficult. It does this by adding a random delay buffer between your key presses and when the key event is sent to whatever input you have activated. The project is easy to compile and run, and its source code is easy to assess if you're worried about its provenance. You can then run the binary as a service or as a single executable, with options to show how much delay is being added, as well as which device is used for the input and the output. With online tests, it works well, although you can feel the variable delay between your key presses and the visual feedback. However, this minor quibble is worth the inconvenience if you restrict its use to online services, which is when it will be most useful.

Project Website

https://github.com/vmonaco/kloak

kloak's default delay is 100ms, which has been shown to be 20-30 percent effective, but you can change this according to risk.

Personal database

Symphytum

There are many things that backup software has in common with databases: They both deal with data, and they both need careful administration and orchestration. And to those uninitiated in their challenges and complexities, they can both appear a little boring or a little too work-like. This is why so many of us forgo backup and hope for the best, and why we maybe prefer a quickly constructed spreadsheet to a fully fledged database. Sometimes, however, if a project goes to the trouble of making an application accessible and unlike work, such as with Apple's Time Machine, it makes what was boring become exciting and useful again. This is what Symphytum does for small, useful, personal databases.

Symphytum feels like an old-school database front end, like the kind you ran on a Commodore Amiga to manage your game collection or to remember to whom you'd lent The Hitchhiker's Guide to the Galaxy. But it's also much more capable than this. Designing data entry forms is a pleasure, for instance. Create a new database (or collection, in Symphytum's nomenclature) with a single click, and you can start adding fields immediately. Quickly add a name and date, click the padlock, and you've got a form where you can start entering data. You can even see any entered data by switching to the more conventional Table view, and there's integrated backup and restore. But its best trick is being able to sync your database to cloud storage, and drivers for Dropbox and MEGA are included. This means you can run the client on different machines and sync to the same database or run multiple clients in read-only mode while connected to the same database.

Project Website

https://github.com/giowck/symphytum

For a personal database, Symphytum has some great features, including two views at the same time and cloud sync.