IPLD with entanglement

[Wondertan]

No description provided.

[govi218]

Here's my idea for alpha entanglements. This is how recovery occurs:

a = 1
b = 2^a
c = 3^b
...
h = 8^g

To recover a data block, XOR the previous and next redundant blocks; for the 1st you can just use the next as it is a duplicate:

recover 1: use a
recover 2: use a^b
recover 3: use b^c
...
recover 8: use h^g

[govi218]

I have questions about choosing indices for nodes; the paper specifies an algorithm for this, but how should we approach it in the context of the DAG? We also need actual links between consecutive redundancies (the dotted lines in the figure are virtual). For e.g., to recover 2, a would need a direct link to b. There is still the issue of how links between data blocks can be recovered, same as #5.