Dynamicland could be done better with Augmented Reality specs. All that frayable, water- and fire-damageable stuff. No version control? No backups? Those little test tube doodads will fall apart after a few months. Plus with AR you can actually use that amazing innovation they call the Inter Network. Argue away, humiliate me for my lack of Bret Victor worship!

Yep very similar in principle 🙂

As someone who will never strap a computing device onto my body, for philosophical and physical accessibility reasons, I find the immersive computing future to be dystopian.

Less stridently, but I'm also very skeptical of AR as a primary computer interface. At minimum it must remain possible to fully utilize computers without it.

If you think of AR as implemented by Apple's Vision Pro (to cite just the latest incarnation), I agee with @Nick Main and @Andrew F that it's not going to happen for mainstream computing, though perhaps for specific use cases where the inconvenience of the devices is tolerable. But AR doesn't have to be close-to-the-body devices. Using transparent screens, small projectors, etc., digital information can be added to the physical world in a more natural way, i.e. closer to what our senses handle best. Dynamicland is a research project. As such, it starts by exploring extremes. Once you understand what you gain and loose from those extremes, you can explore compromises. Which in the case of Dynamicland includes re-integrating tradtional computing technology.

As someone who will never strap a computing device onto my body, for philosophical and physical accessibility reasons, I find the immersive computing future to be dystopian.

I get this completely - but it's a whole nuther thread really. For the sake of this discussion, imagine a non-intrusive AR tech that's fully under your control, etc.

I'm having trouble reconciling:

Dynamicland could be done better with Augmented Reality specs.

…with:

All that frayable, water- and fire-damageable stuff. No version control? No backups? Those little test tube doodads will fall apart after a few months.

You're suggesting we use AR specs (or whatever) instead of projectors — that covers the output side of the system. So what do we use instead of physical materials for input side?

So what do we use instead of physical materials for input side?

I think with good hand recognition you could simulate physical objects digitally that would be needed for input (e.g. pieces of paper, buttons, knobs). The main thing missing would be tactile feedback.

I think the way that physical objects benefit from, you know, physics, and that you get this benefit for free, is just as big or bigger than the tactile feedback. That's why I'm curious about Duncan's contention that Dynamicland could be done better with AR specs. DL seems to be largely about embracing the real world, and the human beings in it. It defines itself in contrast to traditional computers. So I want to know if by "better" Duncan means (to be a bit glib) "get rid of all that stuff that makes DL what it is and just go back to a more traditional computer approach, but with AR specs, because traditional computers are better", or if he means "here's a way to satisfy the goals of DL without as many downsides"

AR is both output /and/ input as @irvin said, and yes, we're still not fully there with tactile feedback, but there is already work in this area. It would basically work the same as in DL: the camera picks up what's going on (both with hands and with physical objects on the table). AR "embraces the real world" by definition, but can also allow someone remote to be present. I just think the "stuff that makes DL what it is" isn't the blobby paper, which I think is quite a distraction. If you allow AR instead of all the blobby paper then the real points can come across.

It's important for what they are researching that it is a shared physical reality. They wanted the lowest possible barrier for participation (other than physically being present), so I don't think it was an acceptable solution to require headsets, although I think a sufficiently cheap headset (think paper red/green 3d glasses) would have been an ok solution. They wanted it to be the kind of thing you could invite a whole class of 7 year olds to and they could jump right in, with instruction/teaching following after.

@Chris Knott OK, that's a good test: throwing 7 year olds in and they get cracking right away. It's both an AR glasses tech challenge to make the headwear robust and light and unintrusive, plus an HX/UX challenge, to make the virtual world overlay be instantly interactive and intuitive, by following the conventions of familiar physics.

as @Stefan pointed out here , its about the culture.. https://futureofcoding.slack.com/archives/C5U3SEW6A/p1687780288915369?thread_ts=1685895424.155509&cid=C5U3SEW6A

@Duncan Cragg I think people get caught up in the current implementation aspects and lose sight of the big idea - what if all our real world objects could compute? The projectors and cameras and papers are just trying to simulate that reality.

@Naveen Michaud-Agrawal yes, that's a logical endpoint, and you wouldn't need to talk about AR or VR, just the New Reality of enriched matter. Houses that could reassemble themselves when you got bored of their layout. So I do think the paper and projectors /are/ a distraction from that deeper vision. And we could still have the exact same conversation in IoT, AR, VR and (dare I say it) The Metaverse!

Yes “property” is something which can be discussed about a lot ;)

Property works as long as "the people" own it collectively, not Big Global Corps, or even Big Unaccountable Governments, come to that. But having said that, I'm a property investor so I'm also one of the bad guys... I think it's about scale, rather than property itself. Of course, IP is a tool/scam that benefits the Big Global Corps.

I think participating in something as dynamicland will show people after some time that many property constructs we use today are not fitting to that way of workers and living together. So there is a natural transformation going on…

Little, Local and Live not Big, Global and Canned, is my preference. Dynamicland does fit that model.

@Duncan Cragg I had similar thoughts when watching the dynamicland video though I was thinking more along the lines of an MS surface studio pro machine, or even larger multitouch screens. This would solve the lag and glitchiness issues seen in the DL demos and allow the app to be run on popular existing hardware and (if popular) help create a market for larger touch screen devices.

Oh right, yes, good points! 👍🏼

DL is a primarily researching a new medium and is only incidentally a computer. I think this point is often missed. I’m not suggesting this will end up being as powerful as writing, but for the sake of argument consider the very early stages of writing. “All that clunky stone and chisel, taking hours to put an idea down, could probably be done better by inventing new spoken words, no?” Until writing was easy and pervasive it didn’t really become a powerful medium that shapes how we think, understand and share ideas. DL is seeking a tangible and dynamic medium that might also change our patterns of thought and understanding. Using the present day flat screen medium might actually be detrimental by closing off certain avenues of possibilities.

Hmmm, sorry but I'm gonna be all contrary to that! I don't believe in it not being a computer or doing computer stuff. I just don't get that impression. It's not comparable to writing. (Also, while I'm being contrary, early writing was pretty pervasive due to being an easy art of pressing a stylus into soft clay!) And my last one: AR, etc aren't any more "flat screen" than the (flat!) artifacts of DL!

I believe in Bret intending for it to not be a computer*, and I think that's the salient difference here. (*as, say, Apple or Microsoft would define "computer"; not as Engelbart or Bush would. We need to be careful with our use of terms that have an overwhelmingly consistent definition in the world at large but which have a different, much more nuanced definition within this community. See also: object, programming, etc.)

It it similar to writing in the sense of being a “medium” - a concept I feel is under discussed even in FoC. Attaching an excerpt from Kay’s _Personal Dynamic Media_ as a starting context. A medium provides and external representation of thought. More than that it creates an internal substrate for thinking and understanding itself (not just communicating pre-existing ideas). Writing wasn’t super powerful just because I could write “Apple” instead of just saying the word or pointing. Writing evolved into more powerful mediums of maths/algebra etc. How do we “understand” any phenomena? If we use formulae such as

F = ma

or line graphs 📈 our understanding is based on the internalized medium of writing or graphs. Computers screens enabled dynamic interactive media (however we still often use it to copy the old static media). If you read the above essay or listen to Victor this idea of “medium first” keeps coming up way before the “computer” term is used: Personal Dynamic Media, The Humane Representation of Thought. Victor says “we think with our whole body”. So I believe what DL is seeking is medium that is both physically tangible as well as dynamic. Note I am not arguing for or against - mainly reporting what I believe is a key perspective in this case. Yes, computers are involved, but the idea is to get to a medium that enables new forms of thought and understanding.

Personally, I don’t find paper to be a more “humane” medium. Carrying heavy books to school, doing text editing on paper, trying to organize or search using paper are all painful compared to digital. We can have 2D interactive surfaces without all of those drawbacks of paper now.

While working on Crosscut, we feverishly debated the merits of physical paper vs digital paper. We filled notion pages and collaborative documents and poster paper and chat channels with as many different angles on this question as we could. Cost, weight, latency, variety, size, feel, smell, space, history… all of it. They're wildly different and only narrowly comparable. They're both wonderful, and digital is no closer to obviating paper than the inverse.

Editing and even writing is much easier in digital. However the “medium of thinking” is still writing/reading. We are still writing, just interactively. The possibilities of the interactive nature of the medium are exposed when we interact to think. For example, in an essay with interactive diagrams like https://ciechanow.ski/internal-combustion-engine/. Or when we fiddle with the sliders or code in observablehq. That was not possible at all before screens. Non linear writing (hypertext) is also a new medium that still retains the writing medium but adds another layer of links, transclusions and such.

I'll add one more — I didn't really understand much about magnetic fields until I made a bunch of simulations of them. They're part of a video I'm working on, for which I did a lot of (research) reading and (script) writing — but those activities on their own didn't do nearly as much to grow my understanding as building a living system in which I could experience the thing intuitively. And while I was readily capable of making such simulations using web tech and text code, I had to attend to all sorts of incidental details that inhibited my learning process. Dynamicland seems to be especially suited for this sort of "learn a thing by building simulations of it", in the same way that computers seem to be especially suited for editing text.

I quit the page when my GPU started to burn a hole in my legs! 😄 Really looking forward to the optimised version - this can all be done in a shader, is that how you did it?

Just passing by to point out that AR specs can be damaged by fire and water. If you damage a piece of paper and didn't keep another copy, you can just print out a new one though 😉