If i am not wrong, these 3 proofs system represents a way to prevent attacks on the blockchain.
The PoS does represent more than 50% of the blocks, what if someone posses more than 50%... ?
The high proportion of PoS blocks is indeed "not ideal".
But it isn't possible to create a long fork
only based on PoS blocks, because the "chain trust" value of the chain decreases when there are no PoW blocks [with reasonable difficulty] in the last couple of blocks. (For a 51%/double spend attack you must create a fork in secret and then try that others accept the fork as the longest chain. If you don't own significant hashrate that would be very difficult).
Obviously, a miner that is also a "staker" can game the system, but he must be a large whale and a decent miner to achieve a successful attack. Now if he's successful - what would he win? Very likely the value of the currency would tank, and I guess it's impossible that he sells his entire stake (again, he must own a lot to be successful)
before the price drop.
So the profit from the double spend would have to be bigger than the exchange rate loss he will get for his stake!
I want to retake one of the crucial questions for every "decentralized web" incarnation, as mentioned by Graham in one of his last posts (related to IPFS):
Who pays for the hosting of the content?Decentralized web technologies based on Torrent or IPFS technology provide (or should provide) mechanisms for content to be shared easily by many peers. But for a web page being available "all the time" there must be
someone (an individual or a group/swarm) who hosts the content "
permanently", so it doesn't get lost.
To my knowledge, there are the following possibilities:
1. The most simple one:
The publisher hosts the content on a machine that is online 24/7. This machine could be an average PC or notebook, but as these devices are often turned off because they're relatively expensive to run (electricity), most likely will be something like a home server, a Raspberry Pi or a modified router (depending on the technical knowledge) or even an old smartphone or tablet.
2. A variant of 1:
The publisher hosts the content on a server in a datacenter. This could be
- a web space / web server (s)he purchases (can be very cheap or even free)
- a free or low-cost cloud storage service like Dropbox with the ability to share links publicly (Dropbox itself doesn't allow public folders anymore)
- a web-based service giving the possibility to host files, like Github/Github Pages.
- an IPFS pinning service (mostly paid, but relatively cheap)
3. A
blockchain. There are special "data blockchains" like Steem or Datacoin.
4. A
blockchain-based storage service like Sia or Storj.
5. An "
altruists' P2P network" that is able to incentive many people to contribute space, like Freenet or ZeroNet.
Every one of these technologies has advantages and disadvantages.
Option 5 (an "altruist network") looks like something like the "holy grail" for P2P anarchists, but is a bit difficult to achieve if we want to preserve the nice Web2Web feature that readers don't need additional software to access the content. It may be possible using technologies like JavaScript "service workers" that operate in the background of a Web2Web reading app. And one would have to think about long-term scalability.
Regarding option 1 (content on a publisher-controlled computer), Slimweb would solve the problem that home PCs very unlikely have a static IP. It could succeed where Opera Unite failed because it's browser-agnostic. A Raspberry Pi distribution of the Slimcoin client and the Slimweb tools could help to increase the number of "hosts" that seed interesting Slimweb content.
For option 2 (content on a server) you could say - why not use traditional web technologies? Slimweb, however, here could achieve a better distribution of the content on the planet if the content is shared by other peers, and potentially more bandwidth. It would be like a "poor man's CDN". It also provides independence from static IPs.
To be able to "focus" on some of these technologies, the crucial question remains: What do we want to achieve with the decentralized web?
What is the main goal? Possible goals are:
control over your own data,
censorship-resistance, and an
easy-and-cheap publishing tool. For some of these goals, some of the options detailed above are better suited than others.
An attractive vision for the "Slimweb" would be to stay an "
technology-agnostic hub" and work with all those existing technologies. For example, the current web2web approach together with Sia and Namecoin would provide a service very similar to what MaidSafe and Substratum are trying to achieve on one single blockchain (and they've still not released anything). To use several blockchains does also carry scalability advantages.
. I'm open to design improvements.
