Talk:Project Xanadu

Needs references: Lots of!

In the current state the Xanadu article is really messy. The history section is vague, full of weasle words and lacks references. For example, a sentence like "...completed a version of the software, ..., though the software did not work the way they wanted" makes no sense -- software is only complete once it works.

Could someone knowledgable on the subject help? -- Syzygy (talk) 08:45, 24 November 2011 (UTC)

reference

Here's a reference --- "hypertext (sic)" -- (word processing and computer storage)

Hyper text did not exist in the 60s. The word was used to mean "really electronic".

There is a cult of people who think the word is an invention. It's hyper-bole. The invention of hypertext (proper) was the advent of a new punctuation TO text, rather than just "really electronic" display and editing.

The advent of typesetting software was important because it helped to streamline publishing and computer use. But it was not hyper text. It was just editing and file storage. — Preceding unsigned comment added by 173.69.141.250 (talk) 14:35, 29 July 2013 (UTC)

Does anyone know if Wikipedia uses bots to verify the accuracy of its inline quotes?

When I think of Xanadu I mainly think of it like a way to solve a lot of the problems Wikipedia has, like how do you know when you are reading a quote is is not misquoted (without doing a lot of manual labor.) Sure an edit that changes only a part of a quote jumps out to patrollers, but in general it would be much cleaner if quoted documents, especially ones in the commons website, were directly attributed via something like transclusion. Ted Nelson has a lot of other ideas that he casually mixes up with Xanadu for reasons I don't quite understand why; editing documents, demos with 3D elements traces everywhere, yet one where text should be navigable through many different pathways would really benefit Wikipedia because right now its basically who ever is the biggest psychopath makes the call on what goes on the front facing documents. That's not sustainable--184.63.132.236 (talk) 18:51, 9 May 2015 (UTC)

On Editing

Editing is fundamentally important. It's not a casual mixup. It's an integral part of a hypertext architecture. This one of the reasons why I do not see the WWW as a true hypertext architecture (the one we desperately need) and neither does Ted Nelson, as we know. A few thoughts:

• Would you want to edit xanalogical structures in linear text? That would be pain, especially when transcluding content. Writing is as important as reading, one cannot exist without the other, so why separate them? What would be a better suitable environment for editing documents if not Xanadu itself?

• When researching (or carefully reading) a topic, I have my own structure. I want the documents I read to follow that structure, otherwise I would get lost in my research. Usually, only a few paragraphs from other documents are relevant for my research, and being able to rearrange documents, when viewing them, in order to create summaries, would be really helpful. And maybe others would like to see my summaries. With the current version of wikipedia, I simply can't have my own versions and share them.

• It's fun to make connections. Connections between similar discoveries in distant fields are more than often the source for great insight. It should be able to form connections between text in an instant, when I see them. This should require no effort (really, why should it?) and god forbid editing textually represented graph structures. I believe, just as Bush did, enabling everyone to making these connections (editing any document) will create a new kind of profession, the trail blazer.

• In short, if you don't include editing concepts in your hypertext architecture, you will have a fragmented (eco)system with barriers, impeding everyone's creativity. I am sorry, if I can't express it any clearer. — Preceding unsigned comment added by 84.115.10.81 (talk) 00:21, 20 November 2015 (UTC)

Suggestion: move "Original 17 rules" to the top?

I read the whole "History" part, before I saw the actual concept described in section "Original 17 rules". Maybe it's a good idea to move this section to the top. What do you think? BWR (talk) 08:40, 22 February 2016 (UTC)

Interview with Ted Nelson

• Interview with Ted Nelson on Aug 18th 2014

- Bevo (talk) 04:13, 23 June 2019 (UTC)

Proposed merge of Tumbler (Project Xanadu) into Project Xanadu

Topic better handled within the context of the larger topic. UtherSRG (talk) 14:30, 9 October 2023 (UTC)

Agreed. There have been no objections in several weeks, if none come forward soon, I think it is fine to boldly merge. Boleyn (talk) 23:54, 13 November 2023 (UTC)

Now merged. Boleyn (talk) 08:35, 25 December 2023 (UTC)

LLM and xanalogical systems like project xanadu

Project Xanadu provided open source access to their green and gold implementations some years ago. Green man runnable, but Gold, the more sophisticated implementation used 12 or so complex mechanisms to implement high performance, scalability, bidirectional links, and a few more complex mechanisms. Much of this is impenetrable for anyone who was not on the Gold development project. As a result independent implementations were not forthcoming.

but there has been a phase change this year, LLMs can understand the Gold cobebase, and are now able to build comparable code bases.

this year a A number of gold implementations have started and been built with similar performance to the Gold implementation.

System	Author	Since	Status	Lineage	Stack
Udanax Green	1988 team	1988	open-sourced 1999	origin	C
Udanax Gold	1992 team	1992	open-sourced 1999	origin	C
udanax-test-harness	Shane Isbell	?	?	Green (actual code)	?
xanadu-spec	Shane Isbell	?	?	formal spec	?
OulipoMachine	Shane Isbell	2017	archived ~2021	Green-inspired	Java
Alph	Adam Moore	2016	active	independent, web-native	JS
xu60	rfinz	?	?	Gold	?
Xudanu	David Jones	2026	active	Gold model, original code	Rust + React
Seed Hypermedia	Beaumont & Anderson	?	?	independent P2P	?

~2026-52000-50 (talk) 16:56, 28 September 2026 (UTC)