Did a talk last week on the rise of block-based interfaces (Notion, Coda, et al.), the historical problems with structured data, and some of the bigger picture goals of the
Here's the talky version:
youtube.com/watch?v=AHblHP
Thanks for hosting the conf!
Conversation
And here's the longer form text and images version: maggieappleton.com/block-data
(still only half transcribed π¬ β will be done by tomorrow)
2
8
46
what is bothering everyone is the semantic desktop. files are now forced into blocks instead of folders to make it searchable like how it is on the web. the web has AI but on the desktop how sustainable is it to rebuild everything as blocks/ .md files? π€
1
Not sure I follow. Can you ELI5?
Are you imagining a hypothetical world where we apply the idea of blocks in a desktop context?
All the block-editors I mentioned and the BP itself are all designed for the web. No desktops involved (unless applications keep data stored locally)
1
yes "a hypothetical world where we apply the idea of blocks in a desktop context". right now they are making us put our ideas into .md files which we manually link together to assemble this semantic knowledge base right? aka semantic desktop/ laptop/ personal cloud
1
i literally have to sit and tag and link all the files together to get a semantic desktop, if you get the trauma. and they are all multiple file formats
1
Replying to
Ah, I get you now π
Yeah, feels clear markdown is not the right format for semantic structures. I appreciated 's thoughts on it in here smashingmagazine.com/2022/02/though
Feels like JSON has become the defacto interop format for now, but still no standardised semantic structures.

