coda
- they are just pre-copies using our smart copy+paste logic
- flow for creating a template today, you flag the doc, has to match certain criteria
- world readable
- only contain one page
- editable by Codans
- fails if not satisfiable
- workflow that runs copy over canvas ranges section
- goes over ranges and generates the minimum op log necessary to produce that
- generate 3 versions: 1) dumb plain text, 2) html, 3) Coda op log
- use the 3rd one for templates. big json with what packs are installed, what automations are there, all object properties, and ops for each of those.
- load templates panel, looks at bookkeeping table: “what are all the avilable templates for given schema version?”
- when you start to drag and drop, pre-fetch the data from s3
- when dropped, use the copy info and play all those ops in the user’s doc
- why is schema version important?
- copy logic relies on this bc op logs are versioned
- is this a problem in practice bc we forcibly upgrade?
- no bc impact has been reasonable and able to mitigate
- admin we can generate ops at previous versions