Since last week (and our first issue), HyperTalk evolved a lot! We created more channels, the server is now close to 200 members and the quality of the exchanges is really astounding! We even did a small get together yesterday in Brooklyn, the first one of our « LocalTalk » events.
This week we reflected more on what HyperTalk should be. There are already wonderful communities like Future Of Coding, Future Of Text and Tools For Thoughts Rocks. So why create a new one and if we do what should it be?
As we’ve already discussed, HyperTalk started as a directory of people gravitating around tools for thinking and computer history/future. Forresto suggested that HyperTalk could be a metalabel (link to his original message). I think this is an interesting concept! If you do not have the time to read the full article here is a short summary of the definition of a metalabel:
« A metalabel is a release club where groups of people who share the same interests collaborate to drop and support work together. A metalabel is a lightweight structure that creates economic, emotional, and creative alignment between collaborators. »
I think this suits HyperTalk really well! But what are we releasing at HyperTalk?
When I look at all the messages, conversations and more generally the tone of the server it seems clear that people here are exploring one main line of thinking:
How computing history informs the design of the products we are using and building today.
This is something I’ve believed for a long time: History driven product development exists and you can even witness it in action around you (just asks some of the Betaworks THINKcamp companies how they came to create their projects).
HyperTalk is the metalabel of those who want to create a better computing experience by looking at old computers and the guiding principles behind their design. Any artefact that represents this principle is a good answer to our metalabel “mission”. It can be a program, a podcast, a newsletter, an article, a blog post, a message, a demo, a prototype…
But if some of us are developing products because (or thanks to) computing history, it means that we learned about that history at some point. This is a topic that seems to be very important in the tools for thoughts community (Jeff Martin is at fore front of this pursuit and I want to thank him here again for his time and for sharing his notes about how we could teach computing history).
The main question is how did we learned about computing history in the first place? Where? In which order did we read some of the foundational articles that are at the heart of this dream machine pursuit? This is the core of HyperTalk: to put together a coherent, approachable and crowd sourced corpus of content to learn computing history easily.
To make it accessible to as many people as possible, to prove that beyond the computing industry history lies a greater and immensely important field: Popular Computing History.
Popular Computing History is the history of the computer for the masses, the small Apple //c where you learned how to program in Basic, the Macintosh SE that opened your eyes to a world of live documents with HyperCard, the Amiga that stunned you with a palette of 4096 colors and a dedicated chip to produce such great sounds that Calvin Harris used one in 2007 to compose his first album I Created Disco. This is the history of how and why we are here today. This has been a lifelong dream and I think I may finally have found the place and the people to make it happen!
So, Hello (again), HyperTalk: The community of popular computing history.
Before jumping into what happened on the server this week, here is a short update on HyperTalk IRL event and the other events happening around New-York this week:
LocalTalk, HyperTalk IRL
Thanks to Jared, Justin and Alan we were able to organize a small get together on Friday 12/2 in Brooklyn! I hope this is only the first in a long series of events!
Alan took the time to share on discord the list of books we discussed at the event, books mentioned with brackets are related but were not mentioned yesterday:
Rise and Fall of American Growth: The US standard of living since the American Civil War
Accidental Superpower
The Verge
[Invention (should’ve mentioned this one! Awesome Norbert Wiener book he hated - about the Econ, social, sci conditions needed to foster invention)]
Survival of The Richest
[Thinking Like an Economist]
Radical Markets,
Entrepreneurial State,
Endurance,
Creative Selection,
Notes on the Synthesis of Form (obv),
Art and Craft of Writing
White Collars
Cubed: A Secret History of the Workplace.
Other events in NYC this week:
Still related to Events, Cole Lawrence was kind enough to put together a list of things happening across New-York next week:
Betaworks demo day
Wednesday, December 7, 3:30 PM EST
Companies from Betaworks' THINKCamp accelerator program share their Tools for Thinking and where they're headed next!
What’s your principle?
Thursday, December 8, 6:30 PM - 9:30 PM EST
Let's come together to make friends, and share our projects and guiding “principles” as described in the talk “Inventing on Principle” by Bret Victor.Audience includes Future of Coding community and Tools for Thought Rocks groups in NYC area.
The venue will be at a spacious bar in the DUMBO area.
BrowserTech NYC
Wednesday, December 7, 6:30 PM - 8:30 PM EST
A meetup about modern browser technology and the things people build with it. We’re talking WebAssembly, WebRTC, WebGL, WebGPU, WebSockets, WebCodecs, WebTransport, Web-everything. (Well, except web3, that’s a different thing.)
What happened around HyperTalk:
New channels created:
Open Hypertext Systems: Gabo is working on the introduction of that channel and I honestly can’t wait to read it! Open hypertext systems are certainly a vast and important subject.
Stand-off properties: have a look at that one to learn more about how we could annotate, comment and share text in a very different way. As always, Iian from Codex is leading the way here!
Text and alternative representations of data: Stand off annotations are showing us how text could become layered, explorable and moldable. But beyond this, can we represent text as something else? Can we act on text with our voice, with our environment?
AI and LLM: As you know we are witnessing a cambrian explosion of tools around AI. This channel will probably be more and more active in the coming weeks and months!
Other nice things: A channel to share events, nice things found around the internet and anything that’s not related to computing ⭐️
Companies channels:
Our friend Horacio over at Mintter is looking for beta users and feedback on their amazing product (seriously, go ahead and schedule some time with him here). Mintter is demonstrating how transclusion, publish and re-use can change the way we share our writings. Everything is based on IPFS and completely decentralized!
After months of work, our multiplayer mode is finally working. We also had a number of discussions about how a spatial canvas with an embedded web browser can make new workflows possible (parallel reading and editing, multiplayer web browsing for researchers, etc) and enhance current ones! The discussion starts here.
Over in the Meemoo channel, Forresto shared this great article about why Apps are too complex so maybe features should be ownable and tradable. I think a lot of people over at HyperTalk would agree with this point and Jef Raskin had an entire part of the Humane Interface dedicated to this problem! But if everyone agrees on the necessity of tradable features or small tool palettes that works across several environments, where are they? It seems the software vendors were always extremely resistant to the idea… It is believed that users wouldn’t pay as much for a feature as they would for an app or a package. I think there is some truth to that and it is supposedly one of the things that killed OpenDoc (I’ll try to dig the article that mentions this). I think that Figma is demonstrating that it is possible to create a nice ecosystem of “small” apps (plug-ins) that can generate revenues! For the designers here, have a look at Diagram and their wonderful tools! The Figma ecosystem is so vibrant that they raised $3M to build new tools on top of the platform. One can wonder what could happen if an open-source, data driven environment came to be 🫧
Note: It seems that some subjects have a tendency to spill over a lof of channels. To keep this newsletter as readable as possible I will continue to edit a section per channel but also regroup some topics to make them easier to read as a “long form” article 🙂 I hope you’ll still find the content enjoyable!
Future of computing channel:
In the future of computing channel, John Underkoffler shared some of his work around architectural spatial computing:
I’m pasting the message from John here because it contains a few very important nuggets for the future of computing:
“[…] here's a bit of video documenting a prototype/demo that I and a colleague designed a half-decade ago at oblong. the main points are (a) an architectural-scale spatial ui, in which (b) participants who come within a proximity-based threshold can cause an application on their laptop to 'jump' across to a giant computational pixel wall and begin to execute there; (c) whereupon the laptop becomes a control interface for the new big version of the executing code... and (d) that big ui physically follows the person who originated it as she moves along the wall; while (e) a background protocol that lets individual applications 'advertise' their capabilities, inputs, outputs, etc. allows them to interoperate a little bit.
Computing history
This week the computing history channel got very lively! Gabo was king enough to share a thread by Brad Neuberg about why the Xerox PARC “failed” (or rather, how Xerox seemed incapable of taking those innovations to the market).
Here is the original message in discord:
And also the content shared in the twitter thread! I find article particularly interesting:
Bruce was then kind enough to share an enormous amount of material about Apple SK8. So much that I think it is worth describing what SK8 is and in which context it was developed before sharing the messages themselves:
SK8 (Skate) was supposed to be HyperCard successor (and more). After HyperCard was handed over to Claris, it progressed at a slower pace. The last version of HyperCard is generally considered to be HyperCard 2.1 with HyperCard 2.4.1 (released in 1998) the last true revision. Even those versions contains relatively few changes compared to the original. HyperCard 3.0 unfortunately never saw the light of the day (and it would probably have been quite different as it was based around QuickTime and interactive videos).
But in Apple ATG (Advanced Technology Group) several researchers were still trying to push Apple to release a new, augmented, “authoring environment”. SK8 and Dylan are two such projects. They respectively began in 1988 and 1992 (and were canceled in 1997 for SK8 and 1995 for Dylan). They were conceived and developed in the midst of what I would call “the great object oriented revolution”. Object Oriented Programming and End-User Programming were in the 80’s seen as two faces of the same coin.
Because object oriented programming can be represented in part by graphical objects that the user/programmer can move, organize and link, it is way easier to learn how to program. (I am distorting the argument a bit here, but that’s how it was interpreted at the time as you’ll see later).
This was at the heart of the thinking behind HyperCard, and I would even argue, the NeXTSTEP programming environment. Steve Jobs never said outright that NeXTSTEP would allow anyone to program but he did said that it would democratize programming and make apps better! Because it was so much easier to develop in an object oriented environment, the quality of apps would go up dramatically and the time needed to build them would decrease.
This theory is at the core of a famous video where two programmers are asked to build the same application. One is working on a SUN workstation, the other on a NeXT computer with NeXTSTEP, Project Builder and Interface Builder. You can see the video here (TL:DR, the NeXT computer wins):
NeXT even advertised its developer environment in its marketing material (extracted from the 1992 NeXT computer brochure):
To go back to Dylan and SK8, they were supposed to be the Macintosh way of creating such an environment, with a stronger focus on the end user and its empowerment. The few screen captures that you’ll find online about Dylan show an interface that it remarkably close to the NeXT system conceptually:
You can find more screen caps over on the OpenDylan website (a gold mine to learn more about the project).
Imagine if Dylan or SK8 had been completed and distributed (like HyperCard) for free, with every Macintosh. Where would we be today? And what would have happened if those environments had been capable of talking to each other over a network? The Mac, the Web, and our whole “digital life” could be very different. Maybe we would all be running custom apps with sharable snippets of code or interface on top of an OS that would only act as a glue between them. The OS would be a place, customizable, personal, humane.
SK8 seems even more approachable (I never had the opportunity to test the system but bruce shared the github page of the implementation of SK8script and you can find what seems to be an executable of SK8 on Macintosh Repository here).
But What is SK8? I’m going simply paste the text from The Apple SK8 Multimedia Authoring Environment website:
SK8 (pronounced "skate") is a multimedia authoring environment developed in Apple's Research Laboratories.
Since 1990, SK8 has been a testbed for advanced research into authoring tools and their use, as well as a tool to prototype new ideas and products. The goal of SK8 has been to enable productivity gains for software developers by reducing implementation time, facilitating rapid prototyping, supporting cross platform development and providing output to multiple runtime environments including Java.
SK8 can be used to create rich media tools and titles simply and quickly. It features a fully dynamic prototype-based object system, an English-like scripting language, a general containment- and renderer-based graphic system, and a full-featured development interface. SK8 was developed using Digitool's Macintosh Common Lisp.
That’s quite a description!
SK8 Project Builder interface (screens from Mac Repository) is also based around collections of elements that can be assembled together to create an app or (and this is the fascinating part), SK8 could be used to develop other development environments!
Contrary to Hypercard, SK8 is not limited to stacks and can generate much more complex environments.
The project seems extremely vast and for anyone that has some time this week-end I would encourage you to read the fascinating messages from Bruce (the thread starts here) and explore the links shared above.
Finally, I wanted to share snippets of the zeitgeist of the late 80’s around end-user programming:
As you can see, Visual Programming really was the Next Big Thing!
University of computing
Since the focus of HyperTalk seems to be gravitating more and more towards computing history I think the “university of computing” will become a very important project. We have a few ideas around how this could become a real thing and I hope to share some of them with you next week!
Reading list:
Last week, the reading list channel was home to a great discussion around the following question: What is the shape of a document?
For clarity here is the original debate, you can look at the first message here. The debate began when I posted this article: What is a document
The discussion continued with a message from Jack Rusher about Notecards! If you do not know about Notecards, head over here to learn the principles of this utterly important Hypertext Software.
Notecards was attempting to solve a number of issues:
How do we write in a hypertextual system?
How do we re-write?
How do we follow links?
How do we handle multimedia content?
How do we orient ourselves in a system with no hierarchy or linear path?
Above is a screenshot of Notecards. If you look at the third line in the lower right window you will see an icon before “Missile”. That icon is the link to the Map you see above that shows the missile range. Instead of showing only the range of the Tomahawk missile you were reading about, Notecards shows you other links (small windows icons) that you can open. We’ll try to do a podcast on Notecards soon as it is definitely a foundational system. Bruce even offered his help to try and make the system run in an emulator. You can see the Notecards repo shared by Bruce here.
The discussion around hypertext and the shape of documents spilled everywhere in the server so please excuse me if this part feels a bit messy!
Justin shared Azlen famous experimentation mimicking the Xanadu beams to show « same origin » content in the system. In addition Justin also shared a demo of the open Xanadu system:
I almost feel guilty about writing about such an important project without more context but I want to stick to the spirit of this newsletter and share what happened in Discord, so you’ll have to wait a bit to learn more about this monument! But I promise that we will expand on Xanadu very soon!
Briefly though: Why is Xanadu important? What can we learn from it and why did the web skewed so far away from its principles? Tim Berners Lee original proposal for the web (Information Management: A Proposal) provides a few answers:
"Hypertext" is a term coined in the 1950s by Ted Nelson [...]
[Note from Paul: The term was coined in 1965 in a communication to the ACM titled A File Structure for the Complex, the Changing and the Indeterminate, if you have some time, read it, as it is probably one of the most important document about hypertext and linked documents ever written]
, which has become popular for these systems, although it is used to embrace two different ideas. One idea (which is relevant to this problem) is the concept: "Hypertext": Human-readable information linked together in an unconstrained way.
The other idea, which is independent and largely a question of technology and time, is of multimedia documents which include graphics, speech and video. I will not discuss this latter aspect further here, although I will use the word "Hypermedia" to indicate that one is not bound to text.
It has been difficult to assess the effect of a large hypermedia system on an organisation, often because these systems never had seriously large-scale use. For this reason, we require large amounts of existing information should be accessible using any new information management system.
The first problem: the term refers to two different pursuits and it may be hard to reconcile them. But is is the second quote that I think provides a clear answer (emphasis mine):
Most systems available today use a single database. This is accessed by many users by using a distributed file system. There are few products which take Ted Nelson's idea of a wide "docuverse" literally by allowing links between nodes in different databases. In order to do this, some standardisation would be necessary. However, at the standardisation workshop, the emphasis was on standardisation of the format for exchangeable media, nor for networking. This is prompted by the strong push toward publishing of hypermedia information, for example on optical disk. There seems to be a general consensus about the abstract data model which a hypertext system should use.
Choices had to be made and everything was not possible, Tim Berners Lee had to choose what to prioritize and there are arguable advantages to the approach he took in the context of its time. This is I think something we sometimes forget as “tools for thoughts” enthusiasts, tools are created in a precise cultural, epochal framework with limited resources.
If you want to read the conversation about Xanadu on HyperTalk you can go here.
Xanadu, the web and the mirror worlds
Those messages about Xanadu and infinite docuverses were followed by other long exchanges in the spatial-ui-design channel, I’ll try to summarize some of them below:
Published in 1991, Mirror Worlds has of course aged but provides a fascinating account of how software was already seen as eating the world back in the 1990’s. Because it was published before the democratization of networked computers and the web, it seems even more prescient. Mirror worlds describes what we would now call the metaverse but in a more political way (how we could vote in it, how it could influence democratic life, etc).
What is fascinating in hindsights is how networks were always thought of as places in the 1990’s. The first “network” released by Apple in June 1994 (eWorld) was also a place:
But maybe I’m looking at this the wrong way: the early 1990’s are fascinating because they represent the last decade where we tried to propose an alternative to metaphorical UI design.
Look at a few of the examples below:
Microsoft Bob:
Apple At Ease:
General Magic “Magic Cap”:
Before the advent of the web and the shift towards the browser as our main software, people in the computing industry were actively trying to move away from the desktop metaphor! Beyond the shape of the document (which has a shape because it is bounded by apps), it is the shape of our environments, and probably our operating systems, that we need to re-assess.
Just like Xanadu, this would require another article, or even a book (or two) but those examples are a proof that the desktop metaphor was at some point seen like something that could be replaced by something else (hopefully better). It is not eternal and it is now a limiting the depth of what we can do with our machines! I’ll leave you here with a few images created by Gray Crawford in the channel (you can look at the images in the channel here), just imagine what is possible:
Have a look at more of Gray’s work on his personal website. And to conclude on those images, a message from Gray about why we need to change the grammar of our UIs.
If you’re interested in learning more about the benefits of spatial UIs (≠ from infinite canvases) you can read this this article by John Siracusa about why the classic MacOS finder provided better orientation than the current macOS finder.
Conclusion
Last week I was writing the first issue of this newsletter on a place from Paris to New-York. This week I wrote this one between Betaworks, a delicious phô restaurant, a park, a bar and part of it on an iPad and the other on a Mac. It was a great experience, and a testimony to the triumph of “mobile computing”, but it was also a reminder of why it is so frustrating to be using amazing hardware limited by old metaphors (or siloed mobile operating systems). This makes me want to end this issue with a quote by Jef Raskin:
Right now, computers, which are supposed to be our servant, are oppressing us. Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining.
This is the “sling shot” effect that I was hinting at in the sub title of this newsletter. We all think that computers are fine until we read about computing history and systems like HyperCard, Dylan or SK8. Once you’ve begun to explore this topic, and begin to envision what could have been you want it to be.
We can all have better computers, with nicer user interfaces, with better data management, with less privacy invasion and they could be more fun and humane too. With the advent of LLMs, VR, decentralized storage and spatial interfaces, it seems that this is finally within our grasp. The next question is, will we run that on the web? on macOS? Or on something new?
Have a good week-end 🧡
Paul
Fun thing of the week: Horacio shared a Classic MacOS CSS theme.
Beautiful old computers, shared by Justin.