You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Internal nodes are not so important because we can reconstruct them if they are missing (and as long as all leaves are present).
Some notes:
When adding nodes to a storage, we can keep only the leaves if space is constrained (add an option to save, and check if a node is internal before saving)
Add functionality to allow internal nodes to rebuild themselves (find all the children leaves, rebuild other internal nodes along the way)
Split the nodes dict into internal and leaves? need a bit more of bookkeeping for this, but it is not strictly necessary.
For the IPFS storage it might be interesting to pin leaves when adding, but not internal nodes. This is also good because signatures are potentially shared more easily than internal nodes of the tree (think the signatures from RefSeq or SRA, for example).
Another usage is for 'increasing resolution': We use somewhat small Bloom Filters for internal nodes, you might want to use bigger ones to avoid false positives. Not a priority, but it kind of comes for free with the other stuff.
The text was updated successfully, but these errors were encountered:
The most important things in a SBT are:
Some notes:
save
, and check if a node is internal before saving)nodes
dict into internal and leaves? need a bit more of bookkeeping for this, but it is not strictly necessary.IPFS
storage it might be interesting to pin leaves when adding, but not internal nodes. This is also good because signatures are potentially shared more easily than internal nodes of the tree (think the signatures from RefSeq or SRA, for example).The text was updated successfully, but these errors were encountered: