Codex OS Profile picture
A developer on a journey to build a free, open source, local-first, knowledge workspace using standoff annotation & hypergraphs. πŸ¦‹ https://t.co/XRknP9ZrTD
Jun 29, 2021 β€’ 4 tweets β€’ 1 min read
What are the pages being referenced here?

"Leonardo's genius epitomized the [[Renaissance [[humanist]] ideal]]..."

They may not be what you think ... And let's say I added one pair of square brackets and you added the other ... How could another person tell who marked up what?
May 10, 2021 β€’ 5 tweets β€’ 1 min read
Implication: to reduce the complexity and breakability of your logic, produce better-defined data-structures which reduce the need for complex and error-prone operations. "Data-structure" here does not simply mean a map of typed values - it implies 𝘡𝘩𝘦 𝘳𝘢𝘭𝘦𝘴 𝘸𝘩π˜ͺ𝘀𝘩 π˜₯𝘦𝘧π˜ͺ𝘯𝘦 𝘩𝘰𝘸 𝘡𝘩𝘰𝘴𝘦 𝘷𝘒𝘭𝘢𝘦𝘴 𝘀𝘒𝘯 𝘣𝘦 𝘭𝘦𝘨𝘒𝘭𝘭𝘺 𝘰𝘱𝘦𝘳𝘒𝘡𝘦π˜₯ 𝘰𝘯.
Apr 6, 2021 β€’ 5 tweets β€’ 2 min read
Effortless multilingual aliasing in Codex.

See how the English word "paper" is linked to the German word "Beitrag" & how the linked refs show all matches, in English or German.

Note also how hovering over "paper" on the right highlights all aliases in the main text. Note also that there isn't a single bracket or piece of Markdown, yet the text is formatted & styled like an article.

This is because all markup is stored 𝘰𝘢𝘡𝘴π˜ͺπ˜₯𝘦 the text in Codex: because what's important is the 𝘡𝘦𝘹𝘡 π˜ͺ𝘡𝘴𝘦𝘭𝘧.

And it should stay pure.
Sep 19, 2020 β€’ 5 tweets β€’ 1 min read
Crazy Codex Idea #127:

Graph-as-infrastructure
versus
Graph-as-diagram (diagraph?) What if part of the problem is that we conflate graph structure (a -[r]-> b) with the visualisation of the graph? Sometimes these are do match but other times we need distance from the structure in order to model diagrammatically. We could mix graph + SVG shapes + text.
Sep 10, 2020 β€’ 6 tweets β€’ 3 min read
Seeing is greater than knowledge.

Seeing is the faculty of knowing, knowing in action, the ability to perceive relation, it precedes and guides action.

Artists have "an eye". Musicians have "an ear".

We see before we know. We see TO know.

But how can we see knowledge? We can see without tools, but the right tools can help us see better.

What is the essence of sight? Spatial relation.

If you can't relate text in space it becomes much harder to see connections ... to make adjustments ... to share your insight.

Seeing is believing. Image
Sep 7, 2020 β€’ 38 tweets β€’ 9 min read
Sep 6, 2020 β€’ 17 tweets β€’ 4 min read
An exploration of some of the things you can do with standoff markup in Codex that you can't do in certain ... other systems.

THREAD
πŸ‘‡ - no [[brackets]] ((of)) {{any}} kind needed!

- no need to pick one name for your concept: just highlight the text to create/link to it

- no need to rename embeds later

- no need to define aliases

- find a concept not just by name but by *any term ever used to refer to it*
Aug 23, 2020 β€’ 5 tweets β€’ 1 min read
Wouldn't it be interesting to see a knowledge graph version of Da Vinci's notebooks where ideas, trains of thought, work sketches, etc., can be traced through his entire corpus?

E.g., show on a timeline all drawing and writing on the Sforza monument + related material. This would be a great Codex Commons project...
Aug 18, 2020 β€’ 6 tweets β€’ 1 min read
Crazy Codex Idea #114:

A panel in two parts where you can see events both on a map and in a timeline. They are synched.

You can add event streams based on people, places, actions, etc.

You can filter streams by date ranges, geographical areas, entities, etc. Of course, the content of this panel can be piped (UNIX-style) to any other panel.

In fact, this will be one of the primitives of the system.