This is the manual for the Elliott 803 library subroutine D2. It was implementing this sorting algorithm in 803 machine code that lead to Tony Hoare suggesting to his boss that he had a better algorithm, which turned out to be quicksort.
Sad to hear of the passing of Tony Hoare. I only met him once at a CCS meeting at the Science Museum celebrating Algol-60. Tony and his team wrote the Algol-60 compiler for the Elliott 803, which we still use today at TNMOC. The "team" kindly signed my copy of the 803 Algol manual.
✅
That should ensure some dry weather for a while 🤦♂️
Some experiments with finding intersections between triangles.
Tools to help with thinking about how to draw 3D triangles :-)
Triangle fill ported from the driver for my RC2014 VDU.
Now to sort out the drawing order.
Yes, it's a bit of a step back in time. A better CPU than a 1980 home computer, 320x240 16bpp, but still not as good as a GPU, so still need something better than a per pixel sledge hammer using a z-buffer.
Z sorting is def. part of the solution, with additional check only done between triangles A and B when they overlap and zmin(A) > zmax(B). But detecting overlaping is not trivial !
You are right, but since I only want to draw low-poly objects it feels wrong to be doing anything per pixel rather than per vertex, per edge or per triangle.
It's the "needs to be convex" constraint that I want to avoid !
That will do "back face culling", but what about overlapping triangles ? Do you use some sort of Z buffer as well ?
@breakintoprogram.co.uk What techniques did you use for hidden surface removal in your Next 3D library ?
I finally remembered how to do rotations properly :-)
I have mastered it several times in the past, but it's one of those things you instantly forget about once the code is working 🤦♂️
"I don't know what the programming language of the future for supercomputing will look like - but I know it will still be called FORTRAN!"
Not for the first time, rotations in 3D space are "doing my head in"
Still a youngster then :-)
A bit of 3D fun with github.com/recp/cglm
The rose seeds that have been in my fridge all winter have started to germinate :-)
@pimoroni.com I've been working on a C API for the Tufty2350 display, and the basics are now working :-)
Maybe not very exciting, but it's working :-)
Nearly there :-)
I might have to get one of these ! shop.pimoroni.com/products/ada...
I'm in unfamiliar territory today. I'm developing a C wrapper for the Pimoroni C++ libraries for the Tufty2350. So far I have some "proof of concept" code running on a PICO2 which calls a C++ class constructor from a C programme.
Today's Start Trek Academy, Wow, just wow.... :-)
Spring is in the air here, so I've been out doing some gardening, well mostly pruning, trimming, dead heading and weeding.
Yes that's how I started :-) I would suggest trying a simple output port next. But you will need to use some clever PIO code to get the access time down when acting as ROM, but I can share some code for that when you are ready.
It's only a bit more work to make the PICO act as a peripheral, or as a block of memory, then you can make the programme loading process much easier :-0)
