💾 Archived View for gemlog.blue › users › maleza › 1607749145.gmi captured on 2024-02-05 at 12:07:35. Gemini links have been rewritten to link to archived content
⬅️ Previous capture (2021-12-04)
-=-=-=-=-=-=-
Spoiler: Neither. I will gyo!
Tired of the bloatware that always emerge from serious problems and user expectations about how a professional solution to their serious problems should look like I decided to commit all my computing attention to a silly made-up one.
From now on (until I get tired of it too) the only problem I want so solve is 9/01 and all it's possible incarnations.
It is a problem simple enough to let me mess in different languages, architectures, protocols, etc without drowning in frustration for not having the time and concentration needed to solve it.
So I started with a GUI plan9 c implementation that is n't published yet because it was made before I totally defined my silly made-up problem. So I have to fix it first.
Then came a cli with a simple companion example to generate it's input with formulas ala bytebeat.
A gopher CGI that is running in my hole at the republic, with sister implementations for Gemini and HTTP in the works.
Gopher cgi (just modify the query string to create new drawings)
And, of course, the wooden hardware implementation that started it all.
In process prototype image (web)
The roadmap grows by the minute. It already includes:
The ubiquitous fileserver (web)
I'v been feeling more in the low level mood than the GUI one, so NES and Forth suddenly popped up to the top of the list.
Since the problem is so simple and involves single bits, even c seems bloated for me this days. It's not that I'm a low level wizard (not even close), is the silliness of the problem that makes abstractions look like to much indirection.
So it's time to solve the dilemma.
I'v been reading a little bit about Forth and almost nothing about NES but for some mysterious reason assembler sounds creepy and tempting at the same time. Let's go to @neauoire's wiki to read about NES. Wait. What's gyo? Imaginary! 4 bits! Small instruction set! Everything sounds so 9/01. Let's try it!
:o You can even "build" this thing over a table with some nibble dices and paper! (note to self: this one goes to the roadmap. Woo!)
A professional, well informed choice as you can tell.
When your only problem is a nail, you can hammer it with anything that calls your attention.
A basic 9/01 cli that only translates the hex code into a drawing seems pretty doable. The rotation instructions look promising to pick the correct bites.
Once that is done, I would like to attempt implementing a bytebeat like thing to create some drawings based on simple math. That involves codding some missing arithmetic instructions, some kind of parser and the simple logic to translate the code into drawings.
I'm barely starting to wonder how but already excited and sure it will be an enjoyable adventure.
Wish me luck. I'll post updates of the process.