Monero developers have released a new version of the protocol, which will go into effect between October 18th and 19th. In addition to changing the proof-of-work algorithm to deter ASIC mining, the new upgrade includes improvements to Monero’s privacy functions that will allow deeper protection for significantly lower fees. Node operators and miners are urged to upgrade their software, and failure to do so may cause transactions to fail.
Mon of Steel
The biggest additions to Monero v.0.13 are “bulletproofs,” which allow transactions to be verifiably published to the XMR blockchain without revealing their size. “Our current range proofs scale linearly in size with the number of outputs and the number of bits in the range,” Monero’s team wrote in a blog post last December, “meaning they make up the bulk of a transaction’s size.”
In contrast, “Bullet” proofs grow logarithmically in proportion to the range and number of outputs in a message, allowing transactions to provide greater privacy and use less data. Monero developers estimate that enacting bulletproofs will result in 80 percent reductions in transaction sizes, with proportionate improvements on transaction fees as well. “Further,” the blog explains, “initial testing shows that the time to verify a bulletproof is lower than for the existing range proofs, meaning speedier blockchain validation.”
In addition, the ring size—the number of decoy signatures used to camouflage Monero users when they send transactions—will be raised from seven to eleven, enhancing privacy on the network.
Faster than a Speeding ASIC
The new upgrade is the latest sidestep in Monero’s duck-and-feint evasion of Bitmain and the world of industrial-scale mining. Unlike Bitcoin or Litecoin, Monero’s developers have explicitly declared war on ASICs, with regular hard forks that will make specially-designed hardware unusable.
ASICs “cause centralization,” Monero wrote in a blog post last year, “because only those with access to specialized hardware (ASICs) are still able to mine, and these typically come from one or very few manufacturers.” The struggle reached a crisis point last year, when Bitmain openly revealed that it had secretly developed miners for the CryptoNight algorithm.
The surprise revelation—along with speculations that Bitmain had secretly used ASICs to dominate Monero mining for months—caused developers to react with emergency measures:
[W]e strongly believe that it’s beneficial to preserve our ASIC resistance. Therefore, we will perform an emergency hard fork to curb any potential threat from ASICs if needed. Furthermore, in order to maintain its goal of decentralization and to provide a deterrent for ASIC development and to protect against unknown or undetectable ASIC development, the Monero team proposes modifying the CryptoNight PoW hash every scheduled fork, twice a year.
This will be the second hard fork, putting the tenth-largest cryptocurrency one step further away from ASIC dominance.
Herding Cats
However, getting the entire Monero network to change in unison is easier said than done. The first PoW change, last spring, resulted in at least three offshoots clinging to the original chain: Monero Zero, Monero Original, and two different coins called Monero Classic.
Developers aren’t taking any chances this time, and are urging users to keep their software up-to-date. “[I]t is imperative that we, as community, contact economically sensitive nodes (exchanges and services) and inform them about the new version,” Monero developers said in a reddit post, and even mobile and desktop wallets will require new versions.
It remains to be seen how well the community will adapt to the new upgrade, especially considering how rarely some hodlers check their balances–to say nothing about updating their software. Depending on how well developers have publicized the change in direction—and how diligent node operators are about keeping their software up-to-date— tomorrow’s fork may end with more XMR nodes wandering away from the flock.
The author is invested in Bitcoin and Ethereum, which are mentioned in this article.
Comments