You are browsing as a guest. Sign up (or log in) to start making projects!

MentalDecoherence

@MentalDecoherence

Joined June 10th, 2026

  • 2Devlogs
  • 2Projects
  • 1Ships
  • 15Votes
Engineering Embedded & RF Systems for Spacecraft | Space Communications, GNC, & Astrodynamics.
Ship

made a terminal typing test. It's really simple, the only dependency that might appear is the curses library if you're on windows (windows-curses), but the installation takes literally 10 seconds. The README on GitHub is pretty short, since there's not much to describe. Yep, that's pretty much it. Had no struggles while building it, made it in the morning as soon as i woke up.

  • 1 devlog
  • 2h
  • 7.57x multiplier
  • 12 Stardust
Try project → See source code →
Open comments for this post

1h 32m 25s logged

Glyph

A typing test that never left the terminal.

Repo

https://github.com/MentalDecoherence/Glyph

What is this

Every typing test lives in a browser tab nowadays. Glyph doesn’t.
It’s a fully functional typing test that runs entirely in your terminal - same ghost text you type over, same live stats - minus the tab, the ads, and the internet connection. Open a terminal, run one command, start typing.
No account. No telemetry. No JS. No TS. No BS. Just curses and a word list.

Why it feels right

Most terminal typing tests either regenerate a whole new screen per test which is annoying, don’t show you a real prompt to type against which feels like a guessing game, or make you dig through a settings menu every single run (annoying again.) Glyph fixes all three.

How it’s built

Glyph is one dependency-free Python file built on curses . A few things worth knowing if you’re reading the source:

  • Layout engine - a small word-wrap pass maps every character index in the target passage to a (row, col) on screen before rendering, so the ghost text, the typed overlay, and the real cursor position always agree with each other, even after a terminal resize.
  • Time mode auto-extends - if you’re typing fast enough to reach the end of the buffered passage before your timer runs out, Glyph silently generates more text behind the scenes. You’ll never hit a wall mid-sprint.
  • WPM math - standard (correct characters ÷ 5) ÷ minutes elapsed, plus a separate raw WPM that includes mistakes, so you can see both your real throughput and your typing accuracy trade-off.

Contributing

Issues and PRs welcome - especially new themes and word lists. Keep it dependency-free, that’s the whole point.

License

MIT. Do whatever you want with it. Seriously.

0
0
15
Open comments for this post

3h 29m 13s logged

Why deep space doesn’t have GPS? (and how i will solve it)

Started working on ADSNS today.


Here is the problem nobody thinks about until they hit it - GPS only works because Earth is wrapped in satellites broadcasting atomic clock precise time signals. The second a spacecraft leaves that bubble, that whole system is gone. NASA’s missions fall back on the Deep Space Network, giant ground antennas that triangulate your position from Earth. It works, but it’s slow, light-time delay alone can be hours round trip by the time you’re past Mars - and it needs Earth to be free, pointed at you, and listening (also DSN antenna time is very expensive).


The good part is the universe already broadcasts its own precision clocks. Millisecond pulsars rotate with stability that rivals atomic clocks. If you can measure when their pulses arrive at your spacecraft and compare that against a predicted model, you can solve for your own position, the same basic math GPS uses, just with pulsars standing in for satellites.


That’s the whole bet behind ADSNS: an Earth-independent GNSS, built from natural beacons. Spent today scaffolding the repo, brainstorming how it will even work, and writing out the data packet that will be used for transmission later which required.. a lot of time (in fact its 400+ lines of code and several hours of structure xD).


I think i will start working on signal processing tomorrow, since to calculate the position, the very first thing we need to do, is to locate where the signal came from, and convert it into a somewhat favorable energy plot for further comparison with our onboard SOM (Space Object Map).


It’s available on GitHub

0
0
10

Followers

Loading…