See... I liked it. I understand that many folks here hate AI, but if this was done a few years ago by a human, with the exact same outcome, people would say it's great.
And weirdly deceptive; for example in my comment I noted the conspicuous absence of Kajiya's Rendering Equation / path tracing paper, and yet there it is in the visualisation panel somehow? Hmmmmm...
In 1968 or 1969, I saw an Evans & Sutherland graphics input computer, complete with light pen, at MIT. Along one wall was industrial shelving. Students slept on the shelving, concentration camp style, to get a session on the computers in the room. It was very inspiring; I had already had two Fortran 4 classes at my high school, 1966-67, which had its own IBM 1130 system as well as an Olivetti Programma. But it was not more interesting than Physics.
No mention of The Rendering Equation / path tracing from James T. Kajiya in 1986? For shame!
Also I was surprised not to see mention of Cindy Goral's Radiosity paper, instead it being apparently developed much sooner, before raster output even?
No mention of Lance Williams inventing shadow mapping, absolutely ubiquitous until the recent increase in ray traced lighting and shadowing?
No joke, I might be a better historical record than this page! As a teenager spending all my pocket money at a crummy internet cafe, I found a huge archive of CG papers from SIGGRAPH and other conferences, going back to the 70s, and absolutely studied all those papers to death from printouts (since I was only allowed to use the home computer for 1 hour a day on the weekends until about 1997). I was also punished with detention for several weeks for "wasting" the Christian Brothers College boarding school printer for printing out all these papers when I said I was only going to print a few; I wore Veach's Thesis printout down until I could barely read the pages by the time I graduated high school in 2001, and nobody ever helped or encouraged me, studying completely alone in South Africa. Different times...
Have you written anything public? These days with open access you could almost write a guide that just references the originals and have that be a usable curriculum.
These days I'm sure AI can do that meticulously. In the last 25 years I've been teaching CG to basically anyone who'll listen, mostly 1:1, and have many unofficial "students" who've gone on to have their own careers, e.g. lead graphics programming at Rockstar Games even :)
For my part, I was long stymied in doing anything beyond basic 2D by the tools available and CPU speeds/memory limitations --- OpenSCAD was the first 3D software I was successful with, and since then I've moved on to the Python-enabled soft-fork:
When the first programmable shading was announced with the GeForce 3, I thought "they's no way this is going to work, if you just multiply the number of pixels with the frame rate and the number of instructions you get an impossibly huge number, it's going to crawl!"
Greetz from my RTX 4090 with 88 teraflops of FP32, lol :)
I clicked on this and I was irrationally disappointed that it wasn't more flashy lol I guess AI is truly rotting my brain at this point.
Anyways, I had Claude take this page and give it the treatment I expected: https://claude.ai/artifact/LLCo69bUdNFrxUSy2p4itp
Not sure if it adds anything of real value, but I had to scratch the itch.
See... I liked it. I understand that many folks here hate AI, but if this was done a few years ago by a human, with the exact same outcome, people would say it's great.
It's mostly worse that way.
I disagree. I find it tastefully done.
As a longtime fan of your work, I think we could do better ;)
And weirdly deceptive; for example in my comment I noted the conspicuous absence of Kajiya's Rendering Equation / path tracing paper, and yet there it is in the visualisation panel somehow? Hmmmmm...
amazing, really transforms the reading experience!
given your insinuated standards it's lazy as fuck
In 1968 or 1969, I saw an Evans & Sutherland graphics input computer, complete with light pen, at MIT. Along one wall was industrial shelving. Students slept on the shelving, concentration camp style, to get a session on the computers in the room. It was very inspiring; I had already had two Fortran 4 classes at my high school, 1966-67, which had its own IBM 1130 system as well as an Olivetti Programma. But it was not more interesting than Physics.
Alt title from the site index:
Milestones of CGI (1999)
No mention of The Rendering Equation / path tracing from James T. Kajiya in 1986? For shame!
Also I was surprised not to see mention of Cindy Goral's Radiosity paper, instead it being apparently developed much sooner, before raster output even?
No mention of Lance Williams inventing shadow mapping, absolutely ubiquitous until the recent increase in ray traced lighting and shadowing?
No joke, I might be a better historical record than this page! As a teenager spending all my pocket money at a crummy internet cafe, I found a huge archive of CG papers from SIGGRAPH and other conferences, going back to the 70s, and absolutely studied all those papers to death from printouts (since I was only allowed to use the home computer for 1 hour a day on the weekends until about 1997). I was also punished with detention for several weeks for "wasting" the Christian Brothers College boarding school printer for printing out all these papers when I said I was only going to print a few; I wore Veach's Thesis printout down until I could barely read the pages by the time I graduated high school in 2001, and nobody ever helped or encouraged me, studying completely alone in South Africa. Different times...
Have you written anything public? These days with open access you could almost write a guide that just references the originals and have that be a usable curriculum.
These days I'm sure AI can do that meticulously. In the last 25 years I've been teaching CG to basically anyone who'll listen, mostly 1:1, and have many unofficial "students" who've gone on to have their own careers, e.g. lead graphics programming at Rockstar Games even :)
One day I'll write my book...
The irony of this being a long scroll of centered text without any images or photos is quite a bit of cognitive dissonance.
For folks who are curious how the story continues, there is:
https://www.shapr3d.com/history-of-cad/introduction
(which has some overlap)
For my part, I was long stymied in doing anything beyond basic 2D by the tools available and CPU speeds/memory limitations --- OpenSCAD was the first 3D software I was successful with, and since then I've moved on to the Python-enabled soft-fork:
https://pythonscad.org/
though I've been trying to work out how to nicely create apps for making projects with acceptable wireframe previews:
https://studio.flet.dev/run/bicQWc3Aul/
which I hope to make projects such as:
https://www.reddit.com/r/cncwoodworking/comments/1xto532/bam...
> For my part, I was long stymied in doing anything beyond basic 2D by the tools available and CPU speeds/memory limitations
I remember trying OpenGL in the mid-90s and thinking "well this can't be used for anything serious!"
It was good for screensavers though.
When the first programmable shading was announced with the GeForce 3, I thought "they's no way this is going to work, if you just multiply the number of pixels with the frame rate and the number of instructions you get an impossibly huge number, it's going to crawl!"
Greetz from my RTX 4090 with 88 teraflops of FP32, lol :)