Group: Forum Members
Posts: 40,
Visits: 1.7K

Remember, a directed Graph G is defined to be a set of vertices V, and a set of edges E such that;
 V is a set of unique objects of the same kind.
 E is a set of unique directed edges, where an edge (u, v) denotes a relation R from u to v (u R v).
 An edge (u, v) exists, iff u and v are elements of V.
Furthermore, the rule of transitivity applies to G, i.e. if both u R v and v R w holds, then u R w also holds for each u, v, w element of V, without any exception. G is acyclic, if it has a partialorder.
Tangle realized by the current IRI 1.4.2.1 (IOTA Reference Implementation) is not a directed Graph, because it does not comply with the definition of a directed Graph.
 Transactions as well as bundles can be entities of the tangle of a node.
 Tangle of a node does not have a distinguished set of edges. The edges are hardcoded as attributes of a transaction (branchTransaction, and trunkTransaction). Each transaction contains exactly these two links to other entities. Also, please note that the entities of tangle are immutable.
 The two links (edges) are not necessarily unique.
Again, IRI realizes tangle as a datastructure with directed links (exactly two of them, to be spelled as “has approved/is approved by”), and in doing so it waives on enormous number of outofthebox algorithms, which uncover different aspects of directed acyclic graphs, from their topology to their structure. Even a simple traversal of the reciprocal of tangle, e.g. for tip selection, must be implemented in a very ineffective manner. In addition, no two nodes of the IOTA ecosystem operate upon the one the same tangle. In other words, the tangle of IOTA does not exist. What exists is one tangle per node. Nevertheless, I will dive a little deeper into IOTA, in order to understand SnapShot, and Milestone processing. I am curious how different tangles are handled as time passes by.

Group: Forum Members
Posts: 40,
Visits: 1.7K

No diving was necessary at all (please see Bundles in Tangle).
