Hypermedia Pattern Language
A Hypermedia Pattern Language. Since Human Systems are Complex Systems from Civilization All the Way Down to human brains, our sense-making and collaboration and communication systems should leverage associative/composable media.
- but I'm still biased toward Worse is Better foundations like web/HTTP, markdown, sql + filesystem, etc
Inspirations/references: Claim, Claim Refactoring Service, Digital Garden, Card-deck, Webs Of Thinkers And Thoughts, Assemble-It-Yourself Component-Based Journalism, Scaling Synthesis, Decentralized Discourse Graph....
A 'node' is a storage-object: record or file. A page is a display of 1+nodes; most often, 1 node. I will probably use these words interchangeably.
Every node should be addressable.
Sometimes a node contains a "single thought" (atomic note), sometimes not. Sometimes it contains a single "block" of text (1 sentence, 1 tweet, 1 paragraph, 1 line-with-sub-bullets), sometimes it doesn't. If you want a subset of a node to be addressable (e.g. paragraph-addressable), you should probably split it off into its own node. (Counter: can/should you leverage text fragment model? Combine text fragment with purple numbers?)
A node should have a good name/title: Designing Good Page Names. Sometimes a node is so short it doesn't have a (separate) name/title (e.g. a tweet). We can often treat the node content itself (stripped of "code") as the name - is there a length-limit to that? Technically you could hit 2000 chars, but is there a downside?
A node's URL should "be" its name: Page Name As URL.
- should strip punctuation
- personally, I strip spaces and use Smashed Together Words; other people turn spaces into dashes or underscores.
- issues with non-ascii characters
A space is a namespace within a domain, a collection of nodes
- should access-control be by space? Or manually set by node?
Permeable membranes, discovery, community, mission
A post is a node.
Indie Media Market
Software/content hybrids: opinionated