absolutely love the visual rule-builder

It looks great! I would love to see the same but done with Vue.js.

cc @Sverrir Thorgeirsson

I don't mean to be a critical but why not just use excel or google sheets and write a script to export to the format of your choice? The UI is much better. Far more people understand spreadsheet UIs than VSCode UIs. You get free collaboration. Maybe I missed it. Only watched half the video and gave up because he wasn't getting to the point of when it became interesting. It was literally 4 and half minutes until he did anything at all. I stopped at 19m

I think the advantage is it gives you a little bit of a visual schema, and you can do things like easily embed images, etc.

you're right that a spreadsheet might be better for most users

one disadvantage of spreadsheets is types and constraints, you could enforce with with scripting of course but you might not end up with an ideal UI

I've worked on a lot of games that have custom (hand written or partially generated) editors for structured data, and you can do variants of this inside popular game engines as well, but excel is also pretty widely used, UE4 can import excel spreadsheets directly, for example (https://docs.unrealengine.com/en-US/Gameplay/DataDriven/index.html)

so I guess the answer is why write any custom tool? I guess it fit this persons needs more than excel did

Good talk so far. I'm not quite through it, but suggestion of "write the meaning first" (effectively a prototype or executable documentation) and then write optimizations that can be turned on and off reminds me a little of John Carmack's blog post on parallel implementations http://www.sevangelatos.com/john-carmack-on-parallel-implementations/

I've tried to do this in a professional capacity almost every time I implement a new system of reasonable complexity but never really get a chance to do it properly due to ridiculous time constraints from product people 🙂

contracts (https://www.eiffel.com/values/design-by-contract/introduction/) and property bases testing go also in that direction

Computing in the future is also something that Carmack talks about a lot. It's pretty reasonable in graphics, if you build with the latest Nvidia super GPU (3090 now) you're 5-10 years in the future of most users

maybe 15-20 depending on what population you're talking about

they did it with say... buying NeXT machines back when they were working on Doom

one thing about this talk is he focuses a lot on programming being engineering or science, and should be a tool for making abstract big ideas more tangible. That makes sense (its Alan Kay!). But I think he's overly negative on what most programmers actually do day to day

most programming jobs are craft, there are no big ideas involved, it's literally making interactive forms for a custom database

tinkering is the wrong word for it I think, maybe plumbing? You could argue that no code\low code tools (or some future imagined tool, or better education, etc.) could get rid of a large class of programmers, but just because you have electricians and plumbers doesn't mean they should be engineers and learn real engineering, and it also doesn't mean that they shouldn't exist as professions and everyone should do their own plumbing and electrical work

On programming as craft, I was reflecting today that my job is 95% tedium, 5% virtuosity, and a good chunk of the virtuosity is in how to turn a thorny problem into a not so fancy custom database.

I can't help but think a system like this is perfect for VR. Similar to how Scott Anderson described his work at Facebook Horizons.

I'm thinking the same thing Ray Imber!

Yeah I haven't had a chance to play with it but I really like the ideas there. It's not radically different from say Dreams, or RecRoom (physical objects plus circuit logic) but it looks fun

Horizon definitely has a lot of similar ideas

I wonder what needs to happen for one of these environments to be useful for "serious work"

might just be that someone has to make a commercial product (probably a game) that is relatively financially successful

i think this other video from the channel is a pretty good elaboration of how it takes the physicalization of logic further than my understanding of RecRoom https://youtu.be/cSPm1xxQ-qY

Did someone try this? Is there any source available?

Ivan Reese can you merge this with the website thread please? Thanks 🙏

No, I can't, sorry. Slack doesn't let me edit messages. My only option is to delete stuff and repost it. Not gonna do that here, since it'd be too disruptive.

Prathyush: this probably isn't it, but this residential programming talk from clojureNorth does have a versioning aspect: https://www.youtube.com/watch?v=Kgw9fblSOx4

This is it!

How did you come across it?

direct manipulation++ wow

i really like this direction but I think direct manipulation like this tends to work so well for manipulating drawings because there's such an immediate feedback loop, you can explore the space of what you're trying to create or to get to where you know you want to get to really fast because you can just see it

with say relational data modelling or implementing a b+ tree, it's harder to directly manipulate i think because you can't see the actual thing you are trying to create because it's abstract

when you see nodes in a node/wire flow or you see UML you can see something you can manipulate, but it's not "direct manipulation" IMO because you're touching the abstract plan for the thing, not the data on disk or an actual binary tree of populated nodes

would love to know if anyone disagrees

in term rewriting systems working from examples it may be more tangible

more generally on systems when you derive/capture behavior working from the data, not from the logic

Oh my wow. Such a well realized prototype. I'll take all of this please. "ACM SIGCHI 2016" — please tell me this line of inquiry didn't just die?!

I would be interested in following this adventure! I have just started exploring generative art and would love to experiment things along this line. I guess you are looking for artists with more experience, that may more or less already use the 15 geometric properties (even subconsciously), but I am really curious about this topic. And it will be a good way for me to enter the Alexander's world.

And to get in touch to more artists, maybe you could try a post on Instagram, I guess they are more present there.

Nicolas Decoster No, I don’t care about experience, I only care about interest in exploring Alexander’s work. And the 15 properties are just an example that looks promising to me; there might very well be other connections to explore.

I’m not aware of anybody having done this before, and there is some thoughtful exploration needed to figure out if and how the 15 properties can map to digital art, but I don’t see any reasons why they shouldn’t.