Levels of abstraction in a modern computer that's running a web application. (The simplicity of the table belies the galaxy of information needed to understand every layer.)

Very cool… I have 2 questions:

- why does this list not start from zero?

- What about the “dependency layer”*shivers*… the web frameworks, libraries, and apis people use or build on top of—used to standardize code to be built in a specific way. 
Would you consider this a layer? what defines a “layer”?

I wonder if it would be ideal to one day pesk back some of these layers to rebuild them, and which ones those would be.

hmmm, my understanding is quite rough at the details for each layer, but i’ll try anyways since this post provoked a thought in me.

blockchains, are a method of immutable data storage that enables sharing of RWX privilege in a permission-less way. a global database.
The problem existing today, that we must duplicate data TOO much, and have not found a way to scale the network cheaply/efficiently. i.e how much replication must occur to classify sufficient decentralization? a quantify problem

on a similar note, there is the IP layer responsible for routing people to their desired destination, and they have solved the problem of smaller machines joining to contribute to a much larger network. i.e data sometimes travels through routers that didn’t request it, and it is performed in a trustless/verifiable manner.

Therefore, it seems that the ideal state for blockchain would be to integrate with the network layer, adding data storage s well as the currently used IP routing - routers=validators

An argument can be made for distributed systems (e.g., blockchains, consensus protocols, state machines) being the highest layer. Beyond that, the only major change I’d make to the list is adding a delineation between user/kernel/hardware applications and processes. Minor change: adjusting for OS specifics.