The students learned to memorise a 3 dimensional concept from a single angle - when challenged to explain it/apply it from a different angle they floundered. -> to truly learn something you need to learn it 3 dimensionally, to understand its nature not memorise its effects or definitions - the test is whether this learning can be applied in non-rehearsed situations

-> definitions should be given in actionable terms - reminds me of Vernon Blake's belief that a good drawing should give enough 3 dimensional data to carve a statue (such data does not require light and shade or intense modelling - simply a 3 dimensional perception in the mind of the artist translated faithfully onto paper) - simply labelling is a learning trap: give real examples, actionable ones preferrably - if not then give multiple definitions (viz. I. A. Richards) to see around the concept

Here is another annotation...

This is a test annotation from Nowhere Man

test overlap

Core hiring issues

Core hiring issues

The team which built FieldBook

Jason Crawford

distribution issues

distribution issues derives from model

“I can’t solve the growth problem until I know what product I’m even growing; I need to figure that out first. Value hypothesis before growth hypothesis. Once we have a product that people love, we’ll figure out how to market it. If the product/market fit is strong enough, there has to be some way to grow; if it’s not, nothing else matters.”

The distribution model

technical meaning?

freemium model

Brilliant.

We need something like this for Codex. In the case of two discontinuous property nodes, they would probably be related via a third 'empty' node.

This could become unreadable very quickly with machine-generated annotations like PoS.

Impressive.

A namespace in Codex is simply indicated by a slash-hierarchy, such as "style/italic", "tei/core/date", or "codex/agent". This is merely a convention, though, and whatever format you like can be chosen.

An interesting feature!

Comments are of course also texts, in this case ones authored by the editor. They are treated as separate texts in Codex and linked to the original text via annotations.

These are zero-width annotations in Codex.

Thankfully where there is no markup, there is no need to escape anything ...

No formal grammar is required by standoff properties. Indeed, what value does a grammar add to a text whose properties are stored in a database and are addressable by other texts and entities? If grammatical constraints are required, callback functions can be added to the annotation configuration object to enforce this. But, again, for a system whose strengths is its lack of hierarchy, why should the user be confined by a constraint system? In a sense, the constraints are enforced elsewhere in the system by the graph data models for the relevant entities.

Codex will soon support text variant annotations between different text layers.