A key benefit of a blockchain is that it must remain immutable / non changeable once set. So to discard earlier blocks would essentially break the history records of the blockchain. The blockchain then would then also no longer be immutable, and therefore tampered with.

    If anyone can tamper with the blockchain, without getting noticed, then the central benefit, advantage, and trust of the blockchain is broken.

    Apart from the first block (genesis) and the most recent confirmed block, every block in the (normal) blockchain contains a pointer to the block that came before and the block that follows. That’s how the chain of blocks is created.

    The problem of Ethereum network taking too long to sync is a design programming problem, which can be solved by an upgrade that does not involve breaking the immutability rule.

    One answer suggests that the problem will be solved by using a process called sharding – which means partitioning the database, so that each individual node does not need to keep the complete blockchain. I believe sharding upgrade may be in the planning.

    However, at the same time, computers are getting faster and more powerful too. So more powerful key technology upgrades may also provide a solution.

    Great question (and solid answer from @itips9999)! The main reason we can’t discard older blocks is that every new block builds on data from the previous block (hence the term “chain”). The system is designed so that everyone needs to agree that all the data is correct. It is important that when someone new joins the blockchain, they can verify that all the data is correct (which requires checking the math from the very first block till the newest one). If you were to start verifying data from a block in the middle, you would be implicitly trusting that the data before it is correct.
    @itips9999: Thank you for the response. Sharding does make sense as an alternative, feels like burden distribution. A follow up to that is then why does Ethereum wants to go to a Proof-of-Stake? It seems counter-intuitive to me that they would be simultaneously be working out a way to both decentralize further and upgrade to a conses protocol that ultimately encourages centralization.
    It seems that PoS is going to undermine the entire purpose of the blockchain revolution, which is simply decentralization and true democracy. Which is why, in my opinion, Bitcoin is doomed to ultimate failure and if Ethereum switches to PoS I believe it will fail as well. Again, am I being naive?

    Thanks Gentlemen. I appreciate the responses and the openness of this community. It’s amazing how willing people are to sharing information for the better of something greater than themselves. This is evidenced by how miners share solutions to making higher profit

    As the blockchain(s) grow, Proof-of-Work (PoW) becomes (a) too slow, and (b) too greedy in energy usage – not helpful for the climate change debate 🙂

    For a blockchain to compete with the speed in which credit card transactions are completed today, PoW will never offer speeds anywhere near what is needed.

    So Proof-of-Stake (PoS) could, with additional upgrades, could address the two key hurdles above. Of course, another alternative may emerge in the future too.

    Bitcoin seems to also be in a constant state of change, with upgrades and smart contract technologies being explored.

    To-date, Bitcoin seems to have survived 172 “death threats”: https://99bitcoins.com/bitcoinobituaries/. So who knows, it may be around for many more years.

    The cryptocurrency space is incredibly fluid. None of us can tell the future. However, cryptos that (a) solve a real problem, (b) solve more than one important real problem, and (c) keep evolving, improving, to match the real-world environments, stand the best chance of continuing to exist.

    That’s why I think Ethereum, and of course, especially blockchain-based technologies offer an exciting future: so much is possible, and much of that hasn’t even been thought of yet.

    Edited on October 09, 2017
    as far as I know, sharding actually does not discard or forgets old blocks. Sharding only distributes the processing of states both from the storage and consensus perspective in a way that only a fraction of the whole data is processed and stored. So it will probably have the same storage problems only in a little bit longer run.

    There is a theoretical possibility however to snapshot or archive the blockchain, like storing actively the blockchain only from a certain blocknumber and storing the rest passively as an archive. There are initiatives for that for example at the Hyperledger project. The problem is however with archiving and with public blockchains, that it is difficult to implement in a decentralized, un-trusted way.

    (Hyperledger is used only for consortium use-cases, so it is not so problematic if a couple of mechanism are semi-trusted and not un-trusted).