Future of Coding • Episode 57 Alan Kay & Adele...
# share-your-work
Future of Coding • Episode 57 Alan Kay & Adele Goldberg • Personal Dynamic Media In stark contrast to the dismal, overbearing weight of last episode's Augmenting Human Intellect, this episode features a paper that is playful, joyful, a thing of charm and wonder. I'm no longer lamenting the future we could have had but didn't (mostly), and Jimmy does all the heavy lifting in terms of bringing sharp analysis and thoughtful reflection despite my best efforts to derail him. As usual, I'll throw a little extra flavour in the thread. Enjoy! https://futureofcoding.org/episodes/057
🙂 2
🌶️ 4
❤️ 2
The song at the end of the episode (which I chopped up to make the sounds for the Bear Who Got Lost segment) is a really, really old song of mine.. 2008-ish, I think, back when I was first learning how to record music. It's called "Grenades", because the piano part involves taking your right hand and lifting it off the high keys on the far right, lifting it way up high in the air and then forcefully PLONKing a really low note at the far left end of the keyboard — your hand makes this big ark motion through the air, like a grenade. You have to do that. It's against the rules to play the song without doing that.
🤣 2
❤️ 2
This one's grabbing me from the start in a way the previous one didn't. 000 300: @Jimmy Miller wanting visual programming for low level programming in particular where you want it for everything leads me to wonder if there's a difference in short term memory capacity. My short term memory is shorter than both of yours, and I have done a lot of low-level programming, and it seems to take an extreme reliance on debug by print. My usual debug process is: • Spend 30 minutes to 30 days getting the log of prints just right • Problem becomes obvious by glancing at a screenful of logs, fix it in 30 minutes (most of them spent writing out a commit message) So maybe my visual tools are textual logs?
🤣 1
300 430: Regarding the difference between what a human considers a "move" instruction and the 40 kinds of such variations the processor itself recognizes, I'm reminded of an exchange with @Chris Knott at http://akkartik.name/archives/foc/thinking-together/1588881606.231300.html. (Chris elaborated more on this in private, and the idea is to express even low-level constraints uniformly in the syntax of the high-level language. Most interesting, and I never did anything with it in Mu.)
I enjoyed this episode much more than the last two. I think this is largely because in this episode you mostly sounded like two students doing your best to understand a master, while in the last two episodes you sounded more like two teachers discussing the shortcomings of a student. The cute nicknames still come across as disrespectful and detract from the atmosphere. A few specific thoughts: "Communications system" means system for communicating between people. Games are dynamic, because they respond to the person playing them, but they usually aren't personal because they are written by a team of specialists and the player cannot usually encode new ideas into the game. The game communicates ideas from the creator to the player. (Yes, there are exceptions like Minecraft.) The point of the meta-medium is that any idea that can be encoded and presented in another medium can be encoded and presented in the computer, though not necessarily in a personal or dynamic manner. For example, streaming a Hollywood movie online uses the computer to emulate film, but it is neither personal nor dynamic. The "real thing" that children get to do with Smalltalk is communicating by expressing their ideas in a responsive form in a personal dynamic medium, and then see other people interact with those ideas. Children were the target audience because they can still learn. As Seymour Papert said, all adults are learning disabled. When you're trying to come up with a whole new way of communicating, it helps to work with people who are still able to learn new ways to communicate. At the time people thought MIT and PARC were kind of crazy for letting children "play" with computers. The fact that some children were able to create programs with Smalltalk as well as some adults using more normal systems of the day was an impressive achievement. The Alto had a maximum of 512K RAM. That's less than a maxed out 8088 PC, so I'm not sure it's fair to criticize the primitive animation techniques. How much capability was realized in such a small computer is a testament to geniuses like Dan Ingalls who made the system so much more expressive per line of code then almost all other systems of the time. You probably know this, but the team didn't stop in the 1970s. You should watch Squeakers http://www.squeakland.org/resources/audioVisual/ for some insight into later work on pedagogy. Describing the music editor as "typical" doesn't make sense when you're talking about the first program of its kind. Adele Goldberg wrote Smalltalk-80 The Interactive Programming Environment, and her views and philosophy appear here and there throughout that book, especially at the end.
❤️ 2
@Personal Dynamic Media Glad you liked it. But don't you think you are a little biased ;)
🍰 1
The Alto had a maximum of 512K RAM. That's less than a maxed out 8088 PC, so I'm not sure it's fair to criticize the primitive animation techniques.
Yeah, I didn't do a particularly good job of articulating this complaint. I was tempted to remove it from the episode, but left it in because the entertainment value of the rant outweighed the expected downside of listeners issuing counter-complaints.
Describing the music editor as "typical" doesn't make sense when you're talking about the first program of its kind.
I have no idea if this is the first time someone made a computerized version of this musical interface, but if so, just mentally adjust my wording to "prototypical" and see if that allows sense to be made.
For the curious, a good entry point into the "what is computer animation missing?" question is this short video from Polygon about smear frames —


. Historically, smear frames in 3d animation haven't been practical because of the limitations of 3d character tools: bones, skinning, IK, vertex weighting, etc., don't allow you to "bend" the character between bones, or duplicate a limb for a single frame, or wildly change the structure of a character. The geometry is too reified. The idea of "make a relatively static thing and then move it around, unchanging" is too deeply rooted. There was a great article (can't find it right now) about the difference between animation in Crash Bandicoot (which was vertex-based) and in Super Mario 64 (which was object-based), and how the former allowed for all sorts of wild expressivity that let them achieve a classic WB cartoon feel, while the latter (a much better game, notwithstanding) looks really stiff, like a puppet. Same era, same(ish) hardware constraints, different implementation, wildly different results. There are examples like this going all the way back to the origins of computer animation. Tradeoffs were made, and those early tradeoffs became institutionalized, and the culture of computer animation is narrower as a result. You don't see a lot of folks using 3d animation tools to make stuff

like this▾

. No technical reason you couldn't make a polygon-based modelling, animation, and rendering tool that could excel at that. You do see this sort of wild expressivity using computer-based tools to produce traditional, frame-by-frame 2d animation. That's great. The computer is helping. But this is an example of the computer just being a better version of a light table and onion paper, as opposed to unlocking entirely new approaches (a la 3d animation).
And for the music rant, in short, I would have loved early computers to borrow more from the Mellotron, and less from the player piano. Focus on the manipulation of recorded sound, without requiring that sound to be digitized first. What You Hear Is What You Get.
Been loving these bookclub episodes. Thoughts... • Can (and have) people visualized a Game Boy's by "simply" mapping its internal state entirely to the screen? And by "simply" I mean "not at all simply." Then from there it would be fun to trace, for example, what goes into lighting up a given pixel or, going the other way, how a button press influences the game. • Multimedia vs metamedia? • Non-affine animation


. • The 😮 of learning that the Circle of Fifths is a hack https://en.wikipedia.org/wiki/Pythagorean_comma.
Multimedia vs metamedia
Non-affine animation


I'm going to have to be much more careful about how I deliver my off-the-cuff rants in the future. This video is actually a fantastic example of what bothers me about modern 3d animation. Here's

a similarly short video▾

showing some of the fun silly things you can do with pencil and paper. Where are the 3d animation tools designed to help you be that freeform? Where are the animators pushing hard against the design of typical 3d animation tools in an attempt to make that sort of thing? The difficulty lies in the bias typical 3d tools have toward reality. Pencil and paper have no such bias. It's a tradeoff — rendering photoreal animation is hard with pencil and paper, to put it mildly, so the bias is justified. But I'd love (love!) to see computer animation tools designed with a different bias. Flash, Toonboom, et al had a different bias, and did some wonderful things. I want more!
The 😮 of learning that the Circle of Fifths is a hack https://en.wikipedia.org/wiki/Pythagorean_comma.
Yeah… sigh. It's like the incompleteness theorem of music. Heartbreaking — I can't have properly tuned octaves and fifths. I can't have properly tuned fifths and major thirds. I do lay awake, staring at the ceiling, lamenting this. I do think about writing music software that helps me write/record music where the tuning system changes on a chord-by-chord basis throughout a piece — sort of like what

Jacob Collier does here▾

, but to an even greater extent (because, unlike Jacob, I have a terrible sense of tuning but the ability to write software to help me overcome it, hah)
For full nerd credit, here's a very quick photo of my office. Mac SE, frequency-to-note-to-midi chart on the wall, Animator's Survival Kit in the bottom right.
No hard drive? Must be a challenge to run System 7.
You can barely see a little UltraDrive 45 (20 MB) that the SE is sitting atop :) And I can't quite recall, but I think this has System 6 installed. Have been meaning to get set up to transfer some items from Macintosh Garden over, but haven't done that yet. Too many things to do, hah. Music first, retrocomputing can wait.
Nice. 😁
So the animation frustration @Ivan Reese is that Pixar is kicking hard against its own tools to get effects that come naturally on paper? Inside Out stands out in my mind because when it came out (in 2015 mind you) the squishy bodies and fuzzy surfaces of the emotions were remarkable.
An animation frustration is that 3d animation tools expect the topological structure of your model to remain relatively constant from one frame to the next. Creating a single frame where you have an extra limb, or morphing/transforming from one kind of object (eg: a rigged character) into a different kind of object (eg: a landscape) is hard. The tools that normally assist your animation (eg: character rigs) now work against you. An animation frustration is that what traditional cartoons achieved with smear frames, 3d animations achieve with motion blur. This is so pervasive that if you look around for examples of smear frames in 3d cartoons, you only see frames with motion blur (and maybe a little squash / stretch, which isn't too unusual or tricky since the mid 00s, [caveat, caveat, etc]). Smear frames are beautiful and wildly expressive and are created lovingly, individually. Motion blur is pretty much an automatic "How much do you want?" algorithm. The motion blur in Pixar looks the same as the motion blur in Dreamworks looks the same as the motion blur in fucking Earwig and the Witch. Not true of smear frames from classic Disney vs WB vs (etc). An animation frustration is that 2d animation tools let you work frame-by-frame, or use tweens. Tweens are ugly by default, and fine at best. Frame-by-frame is the only way to make transcendently beautiful motion. (There are exceptions, of course, but they prove the rule). But, frame-by-frame is only moderately better on a computer than it was when done on paper. There are probably tremendous gains to be had by improving how computers help you do frame-by-frame animation (Looom comes to mind), and by improving tweens. For all of the above, it's worth noting that I'm talking very much about the way things move and change over time. I'm not talking about what those things are, or what they look like when stationary. That's a whole separate area — another rant for another podcast, perhaps.
🤔 1
Smear frame or motion blur?


Motion blur :(
😿 1
I just finished the whole thing. I think I might be starting to get this funky new "podcast" thing.. I love the time loop at the end, where you record the sample that you then use at the start. Favorite quote:
We both have high brows, they just go in different directions to make a pretty repulsive face.
--- I liked the two possible meanings of meta-medium: • One that can simulate any medium • One that encompasses other media, as Ivan said. (though the example of film also brings up all the ways people woke up to the ways film was a whole new medium even though it encompassed plays and music and opera.) But I'd like to suggest a third possible meaning: • A medium that can be used to create other mediums. I think that fits computers the best in a world where computers are working their ways into all sorts of devices and form factors. For example, consider the original Kindle device as a medium. It literally imposed new kinds of restrictions (books that are easy to read, books that are hard, places that have connectivity, places that don't) that were different from other computers. Devices like iPhones also become new mediums in the things they forbid (such as programmability) and the activities they encourage. --- Here's a link to the New Media Reader. I just ordered myself a copy. https://www.amazon.com/New-Media-Reader-MIT-Press/dp/0262232278 --- Another great quote I wish to hear more elaboration on:
Beat poets can't escape their medium.
--- The goal Ivan articulates for FoC: • be useful today • while pointing unambiguously at all the possibilities for the future