I’ve been browsing the forum while trying to wrap my head around working in Tinderbox. I saw this statement and was wondering if someone could help me understand that. My searches here and elsewhere on the Internet haven’t been very helpful.
Great question.
The Internet is the global interconnected network of computers that makes it possible to communicate and share data. Its broad application includes a mix of hardware, software, and standardized protocols.
Hypertext, developed by Ted Nelson and Douglas Engelbart and the community of thought leaders (Ted coined the term in 1965), is a way of organizing/structuring non-linear information through links. In comparison to the Internet, its application is narrow and is made possible through software like Tinderbox.
This is a big question and the answer doesn’t fit in a TikTok length answer.
The Web was a kludge, due to constraints of the time. This was not Tim Berners-Lees’s intent or fault. His original plan was to something more like the hypertext of the day but we only got part of that. The World Wide Web took off for a number of primary reasons (following list is not exhaustive):
- ‘Anyone’† could write a web server and join the World Wide Web (aka, WWW, the Web or just ‘web’).
- Internet [sic] access wasn’t heavily controlled and likely free if you worked at a University or research institute.
- The web failed gracefully. The 404 page might be annoying but nothing broke—you just didn’t get the info you hoped for.
So for those with the skill 9or with access to such people), there was little impediment to getting online and using the web. Lost is all that were hypertext concepts like links in a linkbase (i.e. stored outside articles), two-way links, editable articles (webpages), etc. We also need to view things in the contains of the time and not by today’s tech.
Still, don’t look at this through a binary good/bad lens. Is the web an example of a hypertext? Yes. Is it a fully figured hypertext? No. To understand all that it is worth looking at some early work:
- Doug Engelbart (https://www.dougengelbart.org). Now perhaps better known for the computer mouse (actually no just Doug’s work) and the ‘Mother of All Demos’, read up on NLS his late '60s
hypertext system. It got lost as he stuck with time-sharing computers (offering multi-user use before networked PCs) but the (text) world went a different way (but also see the PLATO system at UoI Urbana). - Ted Nelson, who coined the word ‘Hypertext’ in 1965. Try and see/read Computer Lib/Dream Machines (1974) and Literary Machines (editions. 1984-93). The latter can be seen in whole/part at the Internet Archive and it is useful as it gives an insight into Ted’s Xanadu system.
- Take all look at papers at the early HT/ECHT conferences which are available (free access for pre-2001 papers) at the ACM’s Digital Library.
- With relation to ‘maps’ in the last above pay attention to papers on Spatial Hypertext. In the rush use algorithmically drawn graphs of networks we for get the role of the human. HI (Human Intelligence) is just as clear as AI.
Tinderbox’s design goes back to the turn of the Millennium and draws, in part, on prior work in Storyspace which was being coded initially is the late 1980s [sic]. Not everything powerful is new. People perhaps over-focus on the map view with it becoming a bit like a box and line chart (apps too many to mention). Link types are more than line labels in a diagram. For instance, consider the Hyperbolic view.
Engelbart’s NLS had a notion of ‘viewspecs’. You told the system the manner (spec) in which you wanted to see the information at hand and depending on the viewspecs passed, you saw a different display of the same data. Tinderbox’s views can be considered the same way (although more of a separate evolution of a good idea). Indeed, if you embrace the affordance of multiple views, a tool like Tinderbox offers so much more to the user.
I’ve only just touched the surface of this but hopefully you’ve got some places to dig into the rather broad question. 
†. Actually vanishingly few: it means anyone with the arcane (in those days) tech expertise.
2 Likes
Also, check out the “Mother of All Demos”: https://www.youtube.com/watch?v=yJDv-zdhzMY
2 Likes
Which would put it more in the realm of Google Docs than Xanadu?
In reading the Wikipedia entry for the Web, it seems like his ENQUIRE system leant towards hypertext and the current model, where “anyone” could write a server came about as an affordance to improve maintainability of the overall system. Am I understanding that correctly?
It seemed like Microsoft was attempting something similar with WinFS, but ultimately decided to abandon it.
Thank you for this. I’ll set aside some time to watch it in its entirety. I’m also working my way through your Learning Tinderbox videos. Thank you for such a great resource!
I ran across this just yesterday after reading your reply and wouldn’t have given the link a second glance had you not mentioned it .
I’ll give it another go, but it’s a very dense document and one that I think will benefit greatly from being printed out for reading. I remember watching a video of Ted Nelson demonstrating the system, but couldn’t figure out how exactly it would work with multiple users or even systems.
I’ve started looking through the papers listed on the Tinderbox reference. I’m sure that they will lead to their own questions and rabbit holes.
Scanning over many of the references in this thread, I can believe that wholeheartedly. In fact, I’ve come back to trying Tinderbox by way of many of those applications. Each helped me accomplish one part of why I wanted to do, but almost nothing is as malleable as I find that I seem to want (not sure about need just yet. It will take some experience before I can answer that question).
I certainly have. And I’m also looking forward to the new Thinking in Tinderbox book that might release any day now.
When you refer to “it” I assume you’re referring to the “Interent.” I have trouble equating “Google Docs” to “Xanadu” as the “Internet”. The Internet is the infrastructure that makes it all happen, and then you have applications that ride on this infrastructure. The infrastructure was on and has evolved with various standardized protocols. This work was often based on information theories like set and category theories. The former being more of what we have today went thinking about the Internet and the later is closing to the linkedbased hypertext line of thinking
Neither. When we talk of the Web we implicitly means the Internet too. The Web is just one protocol for using the Internet to communicate. You should look up things like FTP and Gopher that pre-date the Web. If one post-dates the start of the web—or frankly, even if one experienced its arrival—it can be hard to recall the pre-Web experience of network communication.
Google docs is a Web based app (or service if you prefer). As a whole it isn’t a hypertext, though user can interlink docs if they so choose and add web links. Google Docs is, in a way, a WYSWYG webpage editor.
By contrast Xanadu is a thing of its own and a complete system. Never built out to scale, we can only guess what it would have looked like. But essentially, Xanadu would have been a single data environment. Apart from primary sources, articles in the system would be the render of a Edit Decision List (EDL)†
Comparing the Internet with Hypertext is actually to compare two different things though both are arguably concepts. What we call the internet is simplistically, a network of computers but the connections in-between are as important. The internet frew out of ARPA net. I’d recommend reading Hiltzik’s Dealers of Lightning (see) which fills in some detail. The invention of TCP/IP (Vint Cerf‡) arguably pushed the evolution of the fledgling network towards when we think of as ‘the internet’.
Thus the term is both a portmanteau for all that makes up the network and its functioning and thus for its existence as a whole. The Web is only one of the services running atop the Internet, albeit the most well known. The Web also happens to be an example of a hypertext. However, Hypertext as a whole is as much a philosophy about non-linearity as being a specific protocol (‘http’ is just a Web protocol, hypertext can be effected in many different ways).
†. This is a deliberate acknowledgment of the EDL as using in movies and TV. Ted Nelson’s father was an award-winning film and TV Director.
‡. Although Cerf is most associated with TCP/IP he acknowledges many inspirations and contributors.
1 Like
Yes. Sadly the book was self-published and paper copies are hard to find and not cheap ($$$), but you could likely print a scanned e-copy onto paper (don’t let the content scale as the page size is unusual). In reading Ted’s writing, one has to remember the tech of the time he was writing. His original Xanadu proposal was written before PCs, indeed network PCs, and screen were small and rare—you printed to fan-fold tractor feed paper or teletype.
[quote=“isheyelens, post:5, topic:8134”]I remember watching a video of Ted Nelson demonstrating the system, but couldn’t figure out how exactly it would work with multiple users or even systems.
[/quote]
Although there were several demos, Xanadu was never fully implemented. The demos mainly show the interconnecting links—Ted’s “with VISIBLE links”. If nothing else they show transclusion, (part of) another doc’s content being rendered in this document. an echo of that transclusion is seen in Tinderbox’s ^include()^. Sadly, Ted’s meaning of ‘transclusion’† has been lost to a meaning more convenient to modern programming, i.e. use of libraries/modules. As can be seen from making an EDL, parts of the doc are literally a character range from another Xanadu document. Xanadu came out before styled text on screens to there is a lot missing from what today’s eye might intuit being there.
Sadly there is no good functional demo of the Xanadu concept right now.
Sorry for this dive down the rabbit hole. Hypertext is my area of interest/research.
†. Ted also invented this word, and a fair few more.
1 Like