w data storage and capabilities ?
Have you read Seth's articles?
Yah but I’m continuing to read
I think the storage question is partly addressed in the collaboration Constellation has with Kinnami, no? Have you read that technical paper yet?
"Additionally, Constellation Network’s approach to tracking the history of value transactions, such as an update to an account balance, is entirely novel. Rather than utilizing UTXO (Bitcoin), eUTXO (Cardano), or keeping track of the global state of an Account/Balance (Ethereum), it instead treats cryptocurrency transactions as a data type that undergoes state-transitions which are reflected within a mathematical object referred to as a cell complex. As wallet address values update so do the hashes of those local state-transitions which are bundled and stored as ‘blocks’ within the global state. This is only possible because of the statically typed topology of the network which enables formal verification ‘at the type level’ of the ordering, sequencing, and arrangement of the data types within asynchronous data streams. The multi-tiered hierarchy of validator nodes enables scale-free storage of the data dependency graph by allowing them to store subsets of history which can then be resolved by executing a map reduce function, eliminating the need for all nodes to know about all data. This is referred to as “Cellular Reserve Banking” and solves the problems of “blockchain ledger bloat” because it allows for the compacting of all transaction history contained within the Hylochain of hashes to be stored into self-similar fractal space which makes it extremely efficient to store and recursively reference. This approach is notably different than the single-tiered, linear, and unproven sharding solutions proposed by other networks to address ledger bloat." From @SethVi https://medium.com/@lee.papa/constellation-network-a-scale-free-ecosystem-of-generativity-59608efd8089
👆👆 That's how it's said 😅 🤜🤛
Is this the same thing as in one of Wyatt's presentation talking about how hypergraph's data being modeled similar to HDFS and Hadoop? And these kinda mapreduce just isnt applicable to linear sturcture of a blockchain?
Sounds like a good question for @SethVi 😃
I probably need to dig deeper into this part, because I don't understand if there is no data storage, then how is the transaction history stored? Surely that data needs to be present (and in sync with the other nodes) somewhere?
Movies != transactional data, the latter is of course stored. We just do NOT store arbitrary data like movies, mp3's etc.
Would a YouTube state channel be possible or a Twitter state channel? Or is that not realistic ? What is this movie = tx data ??
A "social media" channel perhaps but ANY data can be transacted on HGTP. Storage is a whole different animal.
Yes. YouTube could use the Hypergraph for when people click to view a video. Or something similar.
So just to compare - right now the HDD usage of a validation node ist around 16 GB HDD.
Обсуждают сегодня