Topic: [INFO] Νέες τεχνολογίες γύρω από το Bitcoin - page 15. (Read 30738 times)

Εξαιρετικά νέα!
Ερώτηση: αν το υποστηρίξουν τελικά οι Miners θα είναι ισοδύναμη πρακτικά η αύξηση του block μέσω segwit patch σε σχέση με την αντίστοιχη αύξηση που θα έφερνε το hard fork?

Στο tesnet που έχει μπει σε λειτουργία το segwit patch δημιούργησαν block με μέγεθος 2.6mb
...
από τι είπαν στο slack κανάλι τα block από το βλέπουν στα test μπορούν να πάνε ακόμα παραπάνω και να φτάσουν και τα 3mb

A major remaining issue is that although you can enable pruning to significantly reduce the required disk space, the network can't currently function if everyone did this, you still need to download all of this data to become a full node, and you can't do any sort of rescan with pruning enabled. So for those who want to or must remain archival nodes, 4 MB blocks would be up to 210 GB of data per year, which is awfully uncomfortable. AFAICT, the solutions to this are:
Distribute the storage of old blocks so that each node stores a few GB of a random range of old blocks. When doing the initial sync, you'll have to search out nodes that have the blocks you need.
Do syncing backward so that you become an "almost-full node" in a short while, and then slowly work on downloading the entire history in the background over a long period of time.
Allow rescans to work reasonably on just the unpruned section of the block chain.
Create some sort of standard privacy-protecting protocol for offloading rescan work to block chain archive services, and/or create the capability for Bitcoin Core to be a filterblock client for network-assisted rescans.
Compact blocks does significantly improve the biggest issue preventing the safe increase of the max block size, which is bandwidth. Once it is widespread and working well, a max block size of 4-8 MB looks pretty possible.

Summary of findings
• We show that selfish mining is not always a rational strategy.
To capture rational adversaries, we therefore quantify the
double-spending resilience of PoW blockchains and objectively
compare the security of different PoW blockchains with
respect to the required number of transaction confirmations.
By doing so, we provide merchants with the knowledge to
decide on the required number of confirmations for a given
transaction value to ensure security against double-spending.
• Our results show that, due to the smaller block rewards and
the higher stale block rate of Ethereum2
compared to Bitcoin
(from 0.41% to 6.8% due to the faster confirmation time),
Ethereum (block interval between 10 and 20 seconds) needs
at least 37 confirmations to match Bitcoin’s security (block
interval of 10 minutes on average) with 6 block confirmations
against an adversary with 30% of the total mining power.
Similarly, Litecoin would require 28, and Dogecoin 47 block
confirmations respectively to match the security of Bitcoin.
• We show that the higher the block reward of a blockchain (in
e.g., USD) the more resilient it is against double-spending.
• Finally, we analyze the impact of changing the block size
and/or the block interval on selfish mining and double-spending.
Our results surprisingly show that setting the block size to
an average 1 MB, and decreasing the block interval time to
1 minute do not considerably penalize security. Our results
therefore suggest that PoW blockchains can attain an effective
throughput above 60 transactions per second (tps) (which
implies that the current throughput of Bitcoin of 7 tps can be
substantially incre

95% of the last 2016 blocks since the last difficulty adjustment must signal CSV.
As a maximum, only 100 of the 2016 blocks can be non-CSV.
If the above criteria are met within the current 2016 block cycle, activation of CSV can happen in less than 13 days.
The first block for this cycle is #415296.
As at block 415484 only 17 blocks were non-CSV.

Sidechains allow Bitcoin to be fully programmable. Unlike the ‘smart contracts’ of a general environment like Rootstock/Ethereum, each individual sidechain is completely “opt-in”: by default, users won’t be affected at all by any new programming.