well.. was looking at csv dbs as shown
seems simple, where the metadata can be made to be various and seemingly flexible elements..
In these examples these include data re carL make, year, class, engine size, etc..
However, it seems these elements can be easily modified to be more interstiong, like adding, time-links, speed of productions, cultural history of colours, etc..
For me though, it seems the constant elephatilla in the room is that such db imagines the subject an object that the db talks about – not an integral part of the said db.. ie, ofcourse an actual Car is not indexed – but perhaps detail/image that is Being imagined as an object to be represented in, for example time dimensions.. (eg all cars from 1995.. etccc..)
However, the image is data as well, of its own character, independent of the subject.. The bits in the image can be imagined as db in itself.. a db that can be indexed and have a life of its own.. eg.. car might be ’95, but bumpers might be from ’90, and wind screen linked to ones from ’76.
More over, the bits themselves have time, size and links, no?
all db entries from mornings, for example..
all db entries with 10char arrays..?
all db entries that have letters from deadsea?
perhaps one and another one with y’s for dry?
maybe all that are Wet (if when A is wet) and animalSound (eg + animalSound, B), an If A + B.. This can then be collided with another sequence precisely because its a data string, no??)