The Bitcoin whitepaper [1] (which is very short, concise, and definitely worth the read) describes how the Merkle trees might be pruned (transactions could be pruned when all outputs are spent) and in fact Bitcoin 0.11 added pruning (and 0.12 added further compatibility for pruned nodes).
Vitalik Buterin (Ethereum lead) also posted a description of how Ethereum will implement State Tree Pruning [2].
These are short/medium-term scaling solutions.
Many people are pushing sidechain [3] / child-chains [4] as longer-term ("Visa" level) solutions.
* Lightning Network [5][6] is what Blockstream and others are pushing for Bitcoin scaling.
* Raiden Network [7] is the Ethereum version, and may come out before Lightning, funnily enough.
For those interested in the topic matter, Vitalik wrote a pretty dense paper on blockchain scaling last year [8].
While sidechains will be an option, Ethereum is also pursuing on-chain scaling as well.
Sharding is scheduled for the upcoming Serenity release [9] (which also switches to Proof of Stake (Casper), another important scaling-related change). Writeup on Serenity PoC2 [10].
In general for scaling atm, processing/network latency/bandwidth for initial sync is more of a bottleneck than raw storage. By that measure, I'm actually a lot more interested in the on-chain tx-processing that Casper/PoS might bring [11].
Of course there's no requirement for a single chain to handle everything. Especially over the past few months, we've seen a few relatively smooth migrations to multiple chains as Bitcoin tx processing has "maxed out" (or more recently as fungibility/privacy issues have loomed). At the same time, Bitcoin volatility has stabilized, so it might be just the natural life cycle for blockchains to "scale" until they can't. It's still early days, so who knows?
Vitalik Buterin (Ethereum lead) also posted a description of how Ethereum will implement State Tree Pruning [2].
These are short/medium-term scaling solutions.
Many people are pushing sidechain [3] / child-chains [4] as longer-term ("Visa" level) solutions.
* Lightning Network [5][6] is what Blockstream and others are pushing for Bitcoin scaling.
* Raiden Network [7] is the Ethereum version, and may come out before Lightning, funnily enough.
For those interested in the topic matter, Vitalik wrote a pretty dense paper on blockchain scaling last year [8].
While sidechains will be an option, Ethereum is also pursuing on-chain scaling as well.
Sharding is scheduled for the upcoming Serenity release [9] (which also switches to Proof of Stake (Casper), another important scaling-related change). Writeup on Serenity PoC2 [10].
In general for scaling atm, processing/network latency/bandwidth for initial sync is more of a bottleneck than raw storage. By that measure, I'm actually a lot more interested in the on-chain tx-processing that Casper/PoS might bring [11].
Of course there's no requirement for a single chain to handle everything. Especially over the past few months, we've seen a few relatively smooth migrations to multiple chains as Bitcoin tx processing has "maxed out" (or more recently as fungibility/privacy issues have loomed). At the same time, Bitcoin volatility has stabilized, so it might be just the natural life cycle for blockchains to "scale" until they can't. It's still early days, so who knows?
[1] https://bitcoin.org/en/bitcoin-paper
[2] https://blog.ethereum.org/2015/06/26/state-tree-pruning/
[3] https://gendal.me/2014/10/26/a-simple-explanation-of-bitcoin...
[4] https://www.ardorplatform.org/
[5] https://lightning.network/
[6] https://lightning.network/lightning-network-paper.pdf
[7] http://raiden.network
[8] https://github.com/vbuterin/scalability_paper/blob/master/sc...
[9] https://github.com/ethereum/EIPs/issues/53
[10] https://blog.ethereum.org/2016/03/05/serenity-poc2/) from March
[11] https://twitter.com/VitalikButerin/status/760185856057638913
(phew, maybe one day HN will have markdown support, maybe one day...)