Conversation

Replying to and
This conveniently captures my thinking, and why I’m more worried about new data structures and affordances for interacting with them than teaching people use cases for what we’ve put forward so far Roam is still wrong and incomplete in fundamental ways
Quote Tweet
some musings on software as data structure, not a collection of use cases: buttondown.email/geoffreylitt/a
5
38
Replying to and
“ Roam is still wrong and incomplete in fundamental ways” and yet you charge customers $15/month, all while ignoring their pleas for help & being openly antagonistic to even committed supporters. Last year at least it felt like excitement & promise. Not so much this year…
3
5
It’s not either-or. You’re doing R&D on the human mind as much as you’re doing R&D on blockrefs and transclusion s and graph dbs. “Bicycle for the mind” is suggestive. Human anatomy matters as much for refining the design of a bicycle as mechanical engineering.
1
12
To call it “onboarding” is to misread half the challenge. It’s like thinking the human factor in bicycles is limited to making saddles comfy. Nope. The human skeleton as a 206-bone/650-muscle mechanism is a very different boundary condition than the human butt that needs a seat.
The power move here is to unify the concerns into a single “why not both” research frontier, not create an unnecessary either-or tradeoff that will only create a long-term weakness in the product and needless superstitious biases.
1
15
Nice that a genius like Patrick finds it worth his time to hack away at Roam, but beware the Straussian trap of underestimating the power of a million or billion ordinary minds, and the leverage latent in empowering them even a little.
2
19
Show replies