You are viewing archived messages.
Go here to search the history.

hamish todd 2022-04-04 19:39:46

Hey folks 😁 here's an update on my geometric algebra programming environment project https://www.youtube.com/watch?v=HYETohI5UVE

Kartik Agaram 2022-04-05 02:07:19

"Oh, by the way, I can at any point rotate all the variables around." 🤯

Konrad Hinsen 2022-04-05 07:58:38

Nice! I was wondering a bit what this could be used for, but the obvious first use case is as a learning tool for geometric algebra. Which is probably the most underappreciated corner of mathematics today.

hamish todd 2022-04-05 11:01:26

😅 That is indeed the main part of the goal! It's propaganda

hamish todd 2022-04-05 11:01:43

I also intend for it to be used for live shader coding 🙂 more here https://t.co/ZrmVvIqHM2

William Taysom 2022-04-07 15:26:33

Some retro sci-fi aesthetic choices there, some subconsciously similarity to some Amiga UI?

As for "backsolving," constraint satisfaction is a good way to approach that kind of question. Read line = plane1 ^ plane2 as establishing a constraint. The semantics of how to fix line, plane1, and plane2 depends on an objective function and other constraints. When read as a definition, you have the implicit constraint that plane1 and plane2 are already been fixed. In some contexts, a satisfying way to satisfy constraints is to make them kind of springy.

Ivan Reese 2022-04-06 06:17:42

Last week, Ink & Switch published the Crosscut essay, which describes our (@Szymon Kaliski, Marcel Goethals, and my) work on a prototype tool for drawing dynamic models of your thoughts. It's a fount of FoC-relevant ideas. For instance, combining the place where the programming happens with the place where the program runs, but even further using the same tools for building and working inside the program. Also, Crosscut tries to not feel like programming, and present a minimalist interface that is ruthlessly focused on emulating the best aspects of traditional pen & paper. If you haven't taken the time to read the essay, you're.. probably busy, I shouldn't bother you, sorry. I'll just leave a note on the door.

Before working on Crosscut, I had been slowly chipping away at my own FoC project called Hest (yeah, still just that one blog post). I also made a podcast just to think out loud about the project, so that I could self-reflect and thus hone my ideas and goals, and maybe someday build another prototype. But when the Crosscut project started, it drew all my attention, and the podcast stopped.

Today, with Crosscut done and released to the world, I'm here to share a new Hest podcast episode — 46 minutes long — in which I compare and contrast Crosscut and Hest.

  • How are their prototypes similar and different?
  • What goals do they share and disagree on?
  • How has my work on Crosscut shaped my thoughts on visual programming generally?
  • What effect will that have as I continue pursuing Hest (or any other FoC projects)?

To find out, kindly use this hand-crafted RSS feed, search for Hest or Ivan Reese in your podcast player of choice, or click one of the buttons here: https://pod.link/1559446316

And stay tuned for more episodes. They won't be on a schedule, but they will be about as relevant and entertaining as this fine slack message. So, fifty-fifty, eh? ;)

Kartik Agaram 2022-04-07 04:06:53

Ok, I grok the Varignon parallelogram construction now. This is pretty 🤯 In particular, encapsulation without lamination levels up the debate around whether abstraction is good or bad.

One minor quibble: "encapsulation" is typically considered to turn things into "black boxes", which is the opposite of what's going on here.

Ivan Reese 2022-04-07 04:46:18

I'd love better terminology than "encapsulation without lamination" — suggestions welcome!

Ivan Reese 2022-04-07 04:47:48

Also, that parallelogram proof is borrowed from ThingLab. It's a great "Todo MVC" for dynamic drawing tools :)

Kartik Agaram 2022-04-08 19:15:14

Then again, it is black box w.r.t. meta ink. Just white box w.r.t. concrete points. I don't think there's a good existing term for this so what you chose is as good as any 🙂

Tom Larkworthy 2022-04-10 14:17:34

I loved that podcast and its great to run with someone else's perspective and constraints to see where it leads.

Gregor 🕰️ 2022-04-03 16:04:15

Hey muchaches, I've been quietly consuming the newsletter&podcast these months...errr..years (thankslots Steve, Mariano, Ivan and all of you for the inputs+inspiration!) while slowly working on my side-project: Tofu. When thinking of this community, I always place Tofu as more of an attempt at the near-future-of-programming. It does not aim for revolution as much as many of the cool things here (booo incrementalism boo boo).

Anyway, after years of working on it and using it on my own, I want to slowly start letting people in and this community is prime for me there. It is currently an experimental extension for VSCode for structurally yet fluidly editing JS & TS. If you don't have VSCode installed, you can also use it within your browser here (there should be a pop-up asking if you want to install the Tofu extension). If you do have VSCode, here's a link to the extension in the marketplace.

Feedback of all kinds wanted and welcome (here or over in the GH issues)!

Gregor 2022-04-06 16:14:58

Here is a little, fairly unscripted, primer video on Tofu

Nick Arner 2022-04-06 23:42:04

Wrote a post on Latency in Computer Systems - what causes it, and why it frustrates us:

https://nickarner.com/notes/the-causes-of-latency-in-computing-systems-and-why-exactly-it-frustrates-us-april-6-2022/

Oleksandr Kryvonos 2022-04-10 08:18:13

made an intro video about my Lisperanto project https://youtu.be/QNyuMK8ZrHI