Blockchains Must 'Guarantee Immutability' to Remain Competitive, Ethereum Classic Developer Says

  • Ethereum Classic (ETC) developers believe strongly in the "principles of decentralization and immutability."
  • ETC developer, Igor Artamonov, recently noted that centrally managed blockchains are extremely expensive to maintain.

Igor Artamonov, the founder and technology lead at ETCDEV, a group of software engineers focusing on Ethereum Classic (ETC) core development, recently explained that ETC’s technology is compatible with Ethereum (ETH) on the API and Ethereum Virtual Machine (EVM) level.

However, features developed on the Ethereum network that could potentially compromise security and make it prone to centralization are not supported by ETC, Artamonov noted. He added that smart contracts and decentralized applications (DApps) built on the Ethereum blockchain will also function on the Ethereum Classic network.

The computer science and management graduate clarified that developer tools created for Ethereum will work with Ethereum Classic as well. The developer tools would also be compatible with most Ethereum forks and private Ethereum-based networks, Artamonov explained.

ECIP-1017, ProgPoW

He continued: “So, technically we have multiple public networks [that are] compatible with each other in 99% [of] cases. Notice that it’s compatibility for 3rd party developers, which means that there could be differences in mining [protocols] (if ETH will move to ProgPoW or PoS).”

As most ETC developers claim, Artamonov wrote that Ethereum Classic’s philosophy is similar to that of Bitcoin (BTC) and “the ETC community cares more about security and decentralization.” He further noted that the ETC community “decided … literally community [and not just] developers [came to an agreement, which] took a year of debating to implement a predictable and objective monetary policy (ECIP-1017).”

Notably, Artamonov mentioned that even though technologies for Ethereum and Ethereum Classic are currently compatible, their respective ongoing development and roadmaps can be expected to “take different paths.”

The experienced Java developer explained that “ETC doesn’t plan to move” to any proof-of-stake (PoS) based protocol “at this moment, because PoS is much less secure and brings centralization around big capital (and early investors, which got 70% of ETH supply).”

1,000,000x More Expensive To Maintain Centralized/Managed Blockchains

Artamonov added that the ETC community “believes in the principles of immutability and decentralization” as they are the best, or ideal, use case for a blockchain. He also thinks:

Any blockchain without the guarantee of immutability would be competing with Amazon AWS (web services/cloud computing platform), which is already much more user friendly and a thousand times cheaper.

Igor Artamonov

According to Artamonov, developing a permissioned (private) and centralized blockchain is “a waste of resources” as it does not provide an effective or efficient solution to any problem, or provide better data management (compared to a database).

As CryptoGlobe reported, startups funding platform Republic CEO, Kendrick Nguyen, has also said that “many enterprise companies intend to use permissioned blockchains for uses that are better suited for a database.”

Moreover, centralized/managed blockchains may be “1,000,000x more expensive on every aspect of operating” - Artamonov pointed out.

Bitcoin Mining Pool Tries to Help Tone Vays Win $10K Bet Against Roger Ver

Cryptocurrency mining pool SlushPool has recently manually added a BTC transaction into a block to help Tone Vays, a derivatives trader and analyst, win a wager mage against BCH proponent Roger Ver.

The wager was made at the 2019 Malta AI & Blockchain Summit, during a debate between bitcoin (BTC) proponent Tone Vays and Bitcoin.com CEO and BCH advocate Roger Ver. In it, Ver argued BTC transactions are too expensive for business use due to the cryptocurrency’s small block size.

Vays, on the other hand, argued segregated witness (SegWit) and second-layer scaling solutions like the Lightning Network allow users to make small transactions without paying high fees, and that he has been using BTC on-chain without paying too much for transactions.

The debate ended up seeing Vays send Roger Ver $5 worth of BTC with a one satoshi per byte transactions fee – equal to the fees paid on the Bitcoin Cash chain – to see if it would confirm the same day. If it did, the CEO of Bitcoin.com claimed he would donate $10,000 to a charity of Vays’ choice.

During the debate, both parties noted the transaction was “priority 23,836 out of 24,355 transactions,” meaning that most transactions on the Bitcoin blockchain had to clear before miners picked that one up, at least according to fees paid for transactions.

As Vays soon noted on social media the transaction cleared after 10 hours. Some, however, found it strange. Cobra Bitcoin, the pseudonymous co-owner of Bitcoin.org and Bitcointalk, pointed out on social media that SlushPool – the mining pool that found the block the transaction was included in – manually added it to help Vays win the bet.

Cobra Bitcoin figured it wasn’t mined “naturally” as it was the second transaction included in the block – right after the coinbase transaction – despite the fee being less than 1% of that of all other transactions included in the block.

On Reddit, users pointed this out and accused SlushPool of manually adding the transaction. The mining pool, according to some users, is known for supporting BTC and being against Bitcoin Cash.

Should Roger Ver Pay?

The wager quickly became a controversial topic that seems to bring back memories of the scaling debate that was going on before Bitcoin Cash forked off of the Bitcoin network back in August of 2017.

Some argue that Roger Ver’s point stands as the transaction wasn’t “naturally” confirmed, but manually included in a block. Moreover the CEO of Bitcoin.com claimed he’d donate the money if it confirmed that day, and when the transaction did confirm it was past midnight in Malta.

On the other hand, some claim the transaction did go through anyway, and as such Roger Ver should donate the funds to a charity of Tone Vays’ choice. Moreover, Vays himself argued BCH supporters could have spammed the BTC mempool with two satoshis per byte transactions to stop his from clearing on time.

On Twitter, Vays created a poll that was retweeted by SlushPool and admitted the mining pool did prioritize the transaction. It currently shows most users believe Roger Ver should donate the funds. As one commenter pointed out, however, the results may change if Ver and other BCH supporters retweet the poll.