What Makes a good network

WHAT MAKES A SUCCESSFUL PROTOCOL/BLOCKCHAIN

For scaling a computer system

1. identify the ‘bottleneck’. Anything else and you’re optimizing the wrong things.
2. The bottleneck in our case are the physical internet ‘pipes’ the connect us. Internet pipes are like straws - it doesn’t matter how you hold a straw or what container you use - the same amount of liquid can flow through it.
3. Even if every blockchain had a supercomputer at it’s disposal, the network link will still only be pipes and straws.
4. It’s different that the current client/server model - those involve high-speed ‘direct’ connections to centralized data centers.

Therefore.

It’s apparent and highly likely to me that protocols attempting to minimize on-chain data instead of trying to scale on-chain will end up winning.

Something is useful if:

1. We can use it ourselves
2. We can compare it and contrast it to something else
3. It can combine/interact with something else

As far as 'tokens' go - or you could call them 'appcoins', any successful token is most likely going to have the property that it doesn't rely or depend on a single 'parent' network to survive. Any ERC20 token (Ethereum) will figure out a way (and they already have to a small extent because ETH has been forked) to become interdependent of a specific chain or network. Many possible way this could happen or other 'actors' could facilitate this to happen, but it will happen.

Ethereum quite literally is like a toy at this point - a single production-level application would cause a massive clog of transactions that would slow down the entire network. The ICO transaction clogs were a thing that slowed down the entire Ethereum network (http://www.zerohedge.com/news/2017-06-21/ethereum-flash-crashes-after-status-ico-clogs-network). Even reasonably sized production-ready app would essentially cause the network not to function properly. Ethereum in concept is amazing - but the internal economics are questionable, and the size of the current network is not good for it. Ethereum Classic is a better project and as a long-term investment makes far more sense (as of now).

A blockchain itself has no concept of ‘users’ - only transactions. The user does not care if the transaction happens on-chain, side-chain, or off-chain so long as it’s verifiably and reliably recorded and settled. So - targeting projects that are aware of this ideal are likely to be profitable long-term. Rsk is one.

Other successful networks will likely include 'cockroach' protocols - start from the ground up, strong communities and valuable use-cases. I think these resilient protocols by nature are more likely to succeed.

RESOURCES FOR LEARNING CS and Blockchain

CS is all algorithms and data structures. So if you want to get it on the ground floor you have to start there (as well as programming)

Algorithms https://brilliant.org/wiki/sorting-algorithms/ (Sorting Algorithms are best way to start imo) http://www.mi.fu-berlin.de/wiki/pub/ABI/DiscretMathWS10/runtime.pdf http://gonitsora.com/algorithm-types-and-classification/

Data Structures http://beastie.cs.ua.edu/DS-intro/ (This one is really really good)

Why is a VPN important https://www.vpnsecure.me/blog/articles/the-importance-of-a-vpn

How bitcoin works (and cryptography) https://en.bitcoin.it/wiki/How_bitcoin_works

Critique of "Proof of Stake" https://medium.com/@tuurdemeester/critique-of-buterins-a-proof-of-stake-design-philosophy-49fc9ebb36c6

DECENTRALIZED INTERNET

Urbit - Something Like this will be "Real Estate" of the new internet - Blockchains will interact with Urbit - Important to conceptualize things like this to see how blockchain will be used in the future https://www.urbit.org/ - Decentralized internet on top of the existing internet Galaxies (Network of Decentralized Stars) Stars (Networks of Planets) Planets (People) Moons (Computer, Phone, Laptop, iPad) Commets (bots)