Algorand CTO responds to criticisms of centralization and permissioned nature of blockchain

John Woods, the CTO of Algorand Foundation, has responded to criticism from Justin Bons of Cyber Capital that Algorand is “centralized and permissioned.” Woods accepted some “critical views” but argued that the chain is still decentralized or permissionless.

Whilst I obviously disagree that Algorand is centralised and permissioned – there are a number of fair criticisms in this thread. It’s vital to be able to consider and accept critical views!
Thanks for the thoughts Justin.
Opening up the Relay program is part of my 2023 strats. https://t.co/RaoxAxR7A5

— John Woods (@JohnAlanWoods) October 21, 2022

Centralization and censorship on Algorand

Bons took to Twitter on October 21 to praise Algorand for its “cutting edge” technology but argued that “permissioned gatekeepers can censor any TX on a whim.” The 25-tweet thread went into detail to say that Algorand is a centralized blockchain controlled by the Algorand Foundation.

“Even though consensus is run through “participation nodes,” which are permissionless The “relay nodes” are hand-picked by the foundation, making them permission!”

According to Bons, the relays can act as “gatekeepers to the entire system” as they are “solely responsible for block propagation.”

In one tweet, Bons did accept that “full decentralization is only a few relatively minor code changes away” for Algorand. However, he believes that the current state of the blockchain, as he postulates that “relay nodes” could be compelled to comply with the OFAC sanctions list” just as we have seen on Ethereum.

Bons also commented on claims that Algorand had solved the trilemma issue of blockchains. The trilemma is an issue whereby achieving adequate levels of security, decentralization, and scalability are a limitation of a blockchain’s makeup. His assertion is that Algorand’s approach to the trilemma may be “good enough” but has “some major limitations.”

It solves it with some major limitations

This means it is still within the trilemma but could very well be good enough

That is if the relay nodes are ever fully decentralized, that is the big question mark here

— Justin Bons (@Justin_Bons) October 21, 2022

John Woods’ Response

While Woods denied Bons’ central thesis, the Algorand Foundation CTO affirmed that “It’s vital to be able to consider and accept critical views!” to accept some of the current failings of the chain.

Woods also retweeted a reply from Patrick Bennett, CEO of TxnLabs, who argued that “to censor, ALL relays would have to censor… the 4 random relays my node happens to be talking to at *that* point would have to censor.” Bennett contended that “a majority of relay nodes” is not a simple task to achieve, given the randomness built into the relay system.

Further, Woods confirmed that opening the relay program will be a “priority for 2023” in a small admission of the limitations of the current closed system. Woods also stated a need to “make it more open,” agreeing that the issue is “non-trivial.”

Relays are equiv to network switches, but nodes gossip to 4 random relays. To censor, ALL relays would have to censor or at least for me, the 4 random relays my node happens to be talking to at *that* point would have to censor. So ‘majority of relay nodes’ is key here.

— patrick.algo (@patrickbennett) October 21, 2022

Overall, Bons’ criticisms are based on an appreciation of the Algoran blockchain as he stated that features such as “ALGOs on-chain governance… is excellent.” The conversation between Bons and Woods within the thread also showcased the ability for genuine dialogue and progression to occur on Crypto Twitter from time to time.

While the conversation was highly technical at times, it is fascinating to see such high-level discussions about technology improvement with a market cap of over $2 billion being discussed so openly in public. Further, the public interactions with the subject matter illustrate the technical competence of the web3 community overall.

»

Source