ON–027: Layer 1s (Part 2)
Jun 26, 2020
Contributor: Richard Red, Decred Researcher
- Decred is a UTXO-based blockchain like Bitcoin, but consensus also has a PoS component whereby Decred holders time-lock their credits in exchange for tickets which confer voting rights. A majority of the transactions on the Decred network relate to PoS in various ways. Using data for block 458,101 (Jun 13 2020), 41.5% of transactions are regular, 29.5% are ticket buys, and 28.3% are votes - with 0.06% of transactions related to revoking tickets that failed to vote (missed or expired).
- Each transaction type has a different on-chain footprint, and within the set of regular transactions there are a number of sub-types that can be easily identified. 1.4 million transactions (43% of all the regular transactions) are engaged in setting up ticket buying transactions. Within the ticket-buying transactions we can differentiate between these which have a single voting address and those with more than 1 (as required for ticket buying with Voting Service Providers that maintain online voting wallets for a small fee). Ticket transactions are relatively small in size, but have accounted for most of the fees paid to PoW miners. This is due largely to a period early in Decred's history when ticket price volatility drove competition to have ticket-buying transactions included in blocks with cheap tickets.
- Looking at when the 11.6 million DCR in circulation last moved, again voting plays a big part in accounting for what we see. The ~50% of DCR that participates in PoS is on the move, buying tickets. 8.4 million DCR has moved in 2020, that's 72% of the circulating supply. The other point of interest here is the premine of 1.68 million DCR, of which 762k remains unspent. 465K of the Founders pre-mine remains untouched, along with ~290k of airdrop DCR, or 1,024 airdrops. Between the genesis and recent past, every month some number of decentralized credits stop moving, presumably finding their way into the wallet of a passive holder.
- Locking DCR for tickets is an action that tells us something about the holder who performs it. This action represents some degree of long-term commitment and, for those who use their tickets to participate in voting on consensus and Politeia proposals, an interest in the how the network is run. The final two graphs are constructed by following the DCR paid to 1) contractors paid by the Treasury, 2) miners paid by block rewards - recording the proportion that ends up in different kinds of "outcome". See here for more detailed methodology, here for code. After following the DCR paid to contractors for 5 hops, 31% had been sent to exchange-associated addresses, 29.4% had been used to buy tickets, 29% was unspent, 7% has unknown outcome (moved more than 5 hops without triggering one of the outcomes), and 3.5% was mixed into the Coinshuffle++ mixing service which has been online and available to Command Line Interface wallet users since mid-2019.
- PoW miners have sold a greater proportion of the DCR they received, particularly in the early stages of the project when the great majority of mined DCR was moving to exchange addresses. Once ASICs were developed and deployed on the network (starting early 2018) the rate at which mined DCR was going to exchanges decreased, and a growing proportion was staked or held. PoW miners are not mixing their mined DCR, so far at least. 60% of the PoW rewards went to exchanges, 23% had unknown outcomes (moved 5 hops without triggering an outcome), 10% was used to buy tickets, 6% is unspent within the first 5 hops without triggering an outcome.
Contributor: Chjango Unchained, Ecosystem Developer at Cosmos
Be on the lookout for the biggest launch to come to the Cosmos Hub since mainnet launch itself: Stargate. Stargate, to the uninitiated, is the name of a major software upgrade that contains several significant milestones, namely:
Inter-Blockchain Communication—*The* IBC that was first specified in the Cosmos whitepaper that the industry has been anticipating with bated breath (need I say more?)
Protobuf migration—Up until Stargate, the Cosmos stack relied on an artisanal serialization format called Amino, compared to Ethereum's use of RLP. While Amino optimized for correctness, the migration to Protobuf would optimize for performance, giving all Cosmos SDK-based blockchains a ~100x boost when they make the upgrade.
State sync—This feature allows new nodes to sync to the blockchain tip 200x faster than they currently can.
Tendermint lite clients—Tendermint lite clients are to trusted Cosmos (PoS) lightweight nodes as BIP 37 Simplified Payment Verification (SPV) nodes are to trust-minimized Bitcoin (PoW) light nodes. The operative distinction is in the trusted versus the trust-minimized dichotomy. Put simply, while a SPV node would not need to trust the full node that getting its data from, a Tendermint lite client _would_ need to trust the validator set that it's getting its data from. With the inclusion of this feature and the feature outlined in #1 (IBC), this implies that Cosmos would have completed the delivery of all of its promised outcomes that were specified in the original whitepaper. Tendermint's lite client design is arguably the most viable way of implementing all BFT-based Proof-of-Stake lite nodes. I fully anticipate the rest of the PoS ecosystem to ultimately converge on this design. More about Tendermint lite clients here.
Chain upgrade module—This automates the process of software upgrades after Cosmos governance votes them in. The current state of upgrades on the Cosmos Hub are painfully manual and difficult to scale without this new feature. A loose timeline for the Stargate upgrade sets us up for an end of July start. Read an in-depth outline about the Stargate upgrade here.
Game of Zones, a nearly 2-month long competition aimed at stress-testing IBC within an incentivized test network, has concluded (synopsis here). Indeed, the Game of Zones experiment was an exercise in global coordination with multiple different ecosystem players, and it concluded with these key contributions, in no particular order:
IRISnet—The shining heroes of GoZ were uncontrovertibly the IRISnet team, who found and reported bugs that were causing consensus failures in a critical component of IBC, the Relayer, before the game even started. Despite GoZ setting off on a rocky start, this does highlight the strength of the rest of the Cosmos ecosystem, where incentive-aligned projects with strong technical teams have stepped up to buttress development support wherever it was needed.
Sentinel—Sentinel is highlighted here for modifying and improving upon the Relayer to achieve the requisite throughput that was needed to handle more packets per transaction across IBC. More on that here.
stake.fish—Stake.fish is *the* largest Cosmos validator with 6.86% of the voting power in the entire network. Its founder co-founded F2pool, one of the largest Bitcoin mining pools in the world—no surprises there why stake.fish is one of the most successful non-exchange/hedge fund validators in Cosmos. Their winning GoZ project was a 90's nostalgia-inducing one. They built tamagotchis into one Cosmos zone with food siloed into another. In order to feed those tamagotchis, you had to transfer the digital food over IBC from the food zone to the tamagotchi zone so that the cute little pets would not starve. (Source)
- Cosmic whales are afloat in the Cosmos. With a newly launched block explorer, we're now capturing data that shows that 67% of all ATOMs in circulation are held in just 63 addresses. This is a very small number of whales, considering there were more than 1000 participants in the public fundraiser that concluded in 2017. The address with the largest holding has 8.6% of all ATOMs! (Source)
- Voting power distribution: Voting power distribution has always been at the forefront of governance discussions. The data shows that voting power has become more decentralized. But not by much. In January, when we gave our very first Cosmos update, 72% of voting power had been concentrated amongst the top 25 validators. In just under the 6 month mark, the top 25 validators have 71.36% of the voting power. This is a small change but an incremental improvement nonetheless. Of course, because delegations are non-sticky due to having an instant redelegation feature, voting power can easily shift from validator to validator. (Source)