Stellar X: Ethereum Blockchain Considered For The Wrong Reasons?
A recent post published by Stellar X revealed that companies and blockchain startups might be retorting to the Ethereum blockchain for the wrong reasons. The conclusions made supposedly reflect actual tests that have been conducted and an “at-scale app” was initiated to test the Ethereum network for nearly half a day.
The Stellar X team made the argument that Ethereum’s true nature is to support “autonomous organizations and ownerless apps.” Simply put, if a firm believes in decentralization and the elimination of a central figure, then the Ethereum blockchain is appropriate. An example of Augur was given as being an appropriate use case, as “they need a fully automated, ownerless way to make-truth decisions about real world events.”
So, as per their findings, firms that consider Ethereum for either issuing a token, initiating trade among users or do so to reduce costs, are doing it all wrong.
Upon conducting the tests, which were supposed based on Kik’s third party load test was used to assess Ethereum performance, two problems came about.
Inefficient Methods To Accept Transactions
According to the claims made, miners do not prioritize transactions by wait time but base it by each account. Random assignment is the method in which miners’ function, which ultimately results in an increased line up and ridiculous wait times.
To back up this argument made, the team looked at two numbers: transaction nonce and account nonce. The transaction nonce is responsible for putting the transactions in order and the account nonce determines when it is mined. The entire decision-making process is made as follows:
- If transaction nonce < account nonce, ignore transaction
- If transaction nonce > account nonce, delayed transaction
- If transaction nonce = account nonce, transaction moves to block
The findings above sprung into existence, as the team tried to use Kik’s load spec of “480 accounts each submitting 1 txn/minute on average for 3 hours.” To further assess the time it takes the Ethereum blockchain to convert the transactions into blocks, the Stellar X team used ETH Gas Station’s standard estimates for gas, resulting in over 13 hours of wait time, with majority of the transactions still pending.
A line chart dubbed, “Txns Submitted by a Single Account”, showed that as the number of transactions by order submitted increased so did the time to settle them. This long wait time can also hinder user experience on its respective platform which is an important one to consider.
Increased Users Equal Increased Costs
As per Stellar X’s finding, “Ethereum app’s per-user costs go up quickly as it adds users,” concluding it to be the sole reason why price spikes are present using the “network across many accounts.”
This result was an accident, but definitely worth looking at. While the previous problem was caught due to a small number of accounts submitting a lot of transactions, this second problem was noticed when a lot of accounts (a total of 28,800) submitted one transaction. Turns out the median confirmation time was 23 minutes, which was much more than the previous test, noting that early transactions were not confirmed until the very end of the list.
Long line ups were not deemed the problem, but rather the increase in miner fees that forced transactions with pre-test “standard” pricing to be priced out, with priority given to the remainder. One would assume that as the number of users increase, the cost would have a falling effect, this seems to be the contrary with Ethereum.
A chart dubbed, “Cost of Transaction by Order Submitted”, showed that for each transaction submitted, the cost kept soaring. It supposedly cost the team $1,445 for an hour to conduct the test, which only supported 8 transactions per second (TPS).
Ethereum’s TPS was compared to that of PayPal’s, where the latter is able to confirm 240 TPS. Given that PayPal was built on top of Ethereum, they would supposedly see fees of $380 million, as per the findings!
Clearly, it is evident that the Ethereum’s gradually creeping transaction fees and long wait times are a result of unrealistic use cases of the platform. Co-Founder of Ethereum, Vitalik Buterin was also quoted by Stellar X, and it goes as follows:
“If you want to build a decentralized Uber and Lyft on top of an unscalable Ethereum, you are screwed. Full Stop.”
This only shows that most of the problems experienced by Ethereum is nothing more than inappropriate guidance provided to blockchain startups who retort to the platform to begin with. Ultimately, using it for “ownerless and decentralization” purposes is ideal, while all other purposes will only create more problems for the blockchain and its users.
What are your thoughts on the findings presented above? Are you convinced as to what’s slowing down the Ethereum blockchain? Share below.