Discussion: (Article) Scaling a Blockchain vs Scaling a Tangle


Author
Message
Winston
Winston
Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)
Group: Administrators
Posts: 3.6K, Visits: 6.8K
Scaling a Blockchain vs Scaling a Tangle
Pointing out possible weaknesses in the DAG scalability
https://medium.com/@lyaffe/scaling-a-blockchain-vs-scaling-a-tangle-8b7182eda980
Winston
Winston
Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)Forum Admin (33K reputation)
Group: Administrators
Posts: 3.6K, Visits: 6.8K
"Professional Network engineer here. Not a crypto-expert, but I know a thing or two about networks and distributed systems, and since this is more of a networking issue than a cryptography issue I feel like I may have some insight.
The issue that is being discussed here is actually the exact same thing which CFB and David are constantly saying is an advantage to the system’s security, and that Vitalik Buterin doesn’t seem to know exists. The system is ‘Network Bound’, or the speed at which transactions can be performed is bound by the topology of the Network.

Being network bound, means that Iota does not rely entirely on PoW to prevent malicious transactions, it can rely on the fact that one cannot simultaneously effect the entire network by flooding it with transactions. One can flood a single node with transactions, potentially disrupting that node, but other nodes will only get secondary effects from the spam and the spam itself will take longer and longer to confirm.
As far as legitimate transactions being effected by the system being network bound, that’s why DAG system currently needs the coordinators. The nature of network communication is that the more nodes exist, the faster nodes can acquire (or direct to) relevant data from their peers. It’s kind of like the x-degrees of separation from Kevin Bacon game, the more connections there are in the system the more likely there is to be a small number of hops required to reach a goal.

The internet would be extremely inefficient if there weren’t so many distributed servers working to coordinate the traffic. If you were trying to build the internet from scratch again, you would need centralized servers to coordinate the traffic, until a critical mass of servers was reached that could handle it on it’s own. Iota is in the same state as the very early internet, before we got DNS and the endless array of backbone routers in place that we have today.

The slowdown caused by over-exhuberant spamming of the tangle is part of the growing pains of this kind of network. Currently the number of users spamming the network with light-nodes is disproportional to the number of full-nodes in existence.
Long story short, if you want to help the tangle then you should spam it. If you REALLY want to help the tangle, though, run a full-node and let people spam you!"

https://www.reddit.com/r/Iota/comments/7eix4a/any_iota_guru_that_can_explain_what_this_guy_is/dq5ijrm/
Edited 2 Years Ago by Winston
GO

Merge Selected

Merge into selected topic...



Merge into merge target...



Merge into a specific topic ID...




Reading This Topic

Login

Explore
Messages
Mentions
Search