Polygon's current support guide records July 3, 2026 as the sequencer sunset. Ordinary bridging no longer works. A separate claims interface uses exit certificates and a read-only RPC for eligible EOA-held assets, with transactions signed on Ethereum. Contract-held assets, including DeFi positions and multisigs, are excluded. The interface is scheduled to retire after December 31, 2027; the guide says assets are not forfeited, but recovery without its interface and support may be impractical. Older instructions to bridge normally are therefore historical, not current operating guidance.

The proof-generation API changelog supplies another concrete consequence of retirement. Its version 2.0 removes zkEVM proof-generation endpoints because the underlying chain infrastructure no longer answers, while leaving the PoS endpoints unchanged. For a developer, this is more useful than a vague statement that Polygon changed direction. A service can continue supporting one Polygon network while dropping another. A maintained repository or functioning corporate website does not establish that a particular retired chain's RPCs, bridge or application integrations still work.

The July 20, 2022 introduction framed the project around a difficult engineering combination: proving execution with zero-knowledge techniques while preserving the familiar Ethereum programming environment. Its sweeping language expressed the team's ambition. Compatibility still had to be implemented and tested, and the announcement itself anticipated that early versions would be incomplete. What made the project significant was the attempt to bring existing code and tools into a validity-proof system, rather than require every developer to abandon their established application stack.

The distinction was not obvious to everyone. The day after mainnet launch, SerWhaleLord asked r/0xPolygon whether zkEVM was separate from the existing proof-of-stake chain or merely supported it. The replies then moved into speculation about which token might benefit. That sequence shows how naming confusion can become investment inference before the architecture is understood. This account treats the question as community evidence, not the replies as authoritative token economics. Polygon's brand covered several systems with different roles and dependencies.

The October 10, 2022 public testnet announcement described the implementation as Type 3 in an EVM-equivalence framework and identified Type 2 as the goal. That qualification is more informative than repeating the word equivalent without a date. Developers were invited to find errors, test capacity and help prepare for audits. A testnet lets people challenge assumptions with working software, but test transactions and volunteer interest do not establish production demand. The public invitation was part of engineering the system, not proof that the final tradeoffs had disappeared.

The November 2022 audit forum post described sessions organized with Scroll and the Ethereum Foundation to explain architecture, proving approaches and compatibility to auditors. This is a record of knowledge sharing around an unfamiliar class of software. It also shows why an audit cannot be reduced to a badge: reviewers first need to understand what the system is supposed to prove. The post opened a discussion for questions. It did not certify every future upgrade or turn every contract subsequently deployed on the network into audited software.

Polygon announced public Mainnet Beta on March 27, 2023, with Vitalik Buterin performing a symbolic first transaction. The same announcement described privileged security arrangements for the early stage and a planned progression toward fewer special permissions. The symbolism and the qualifications belong together. Public access meant users could transact and developers could deploy; it did not mean all operator powers had already disappeared. The beta label also carried a practical warning that unfamiliar software could encounter failures even after testing and external review.

The separate open-source announcement selected AGPL v3 for the code and explained why Polygon considered openness part of scaling Ethereum's public infrastructure. Publishing source allows inspection, modification and reuse under a licence. It does not automatically create independent operators, eliminate upgrade authority, or pay the cost of running a service indefinitely. That distinction became especially important later: useful code could remain available even when this particular public network no longer made sense for its sponsor to maintain.

A software legacy and an operating chain have different lifetimes.

The prover explainer describes transactions being ordered into batches, their data being made available on Ethereum, and proofs being generated and aggregated before verification on Ethereum. Parallel proving was intended to prevent a growing queue from turning the prover into a bottleneck. The final proof wrapped earlier work into a succinct form that was cheaper to verify. The key relationship is between a specific state transition and the rules checked by the proof. This is execution verification; the name zero knowledge does not mean ordinary users' transactions became private.

The archived node repository distinguishes trusted, virtual and consolidated states. A quick result supplied by the trusted sequencer was not the same stage as data submitted to Ethereum or a state subsequently proved there. Its glossary also explains resource limits that could make a transaction invalid for the zkEVM even when familiar Ethereum concepts suggested it should execute. These implementation distinctions give readers a better vocabulary than instant finality. The meaning of a confirmation depends on which stage the observer has actually checked.

In the March 2023 security discussion, Yarko_De.Fi welcomed the staged approach, security council and delayed upgrades as a sensible path toward maturity. Jackson also expressed anticipation. These are attributable endorsements of a proposed process, not proof that all users accepted privileged intervention or that the final decentralized stage arrived. The record is useful because it shows a real argument for temporary safeguards: some participants believed limited emergency authority was preferable to exposing early users to an entirely inflexible implementation.

The node's v0.3.2 release notes expose the work behind a smooth application interface. They list a data-streaming service, synchronization fixes, a race condition affecting batches, and compatibility with a specific prover and bridge release. These are concrete engineering changes, rather than a new headline throughput number. A public network relied on release coordination and operators running compatible components. Successful proving mathematics alone could not ensure that a node stayed synchronized or that its database handled concurrent work correctly.

The February 2024 Elderberry post described testnet availability, fixes for out-of-counter errors and further changes after Etrog. It linked the proposal and versioned components so operators could inspect what was being scheduled. The initial post's anticipated mainnet date was conditional, which makes it unsuitable as proof of the eventual activation date. A release history should distinguish the day an upgrade is proposed, the day it is available for testing, and the later report that it actually completed.

The March 14 Elderberry thread records the human cost of late changes. Operator ian said a same-day mandatory prover revision disrupted preparations and complained that independent node runners were not given enough notice. Yorick replied that their nodes were also down despite updating in time. Ian additionally alleged deliberate exclusion, but the thread supplies no proof of that motive. The supported conclusion is narrower and still serious: the release process produced frustration and operational difficulty for identifiable community operators.

A March 12, 2024 announcement explained that users could see gas estimates substantially above what they ultimately paid after a fee-calculation change. Custom RPC methods were offered to improve those estimates. This was a usability problem produced by the relationship between execution and data costs, not necessarily evidence that users were charged the alarming number on screen. The incident matters because a technically correct final fee can still discourage users if the wallet asks them to approve a confusing estimate first.

The March 28 outage report traces a synchronization problem following an Ethereum reorganization to inconsistent exit-root handling and invalid batches. Recovery involved new software, resynchronization, an emergency upgrade and a period when withdrawals were halted. The report also warns that some apparent transactions were reorganized or not processed. It explicitly says Polygon PoS and other named infrastructure were unaffected. This is why network identity matters during an incident: a failure in one deployment should not be silently expanded into a claim that every Polygon chain stopped.

In April 2023, AverageRedditLad published an enthusiastic explanation of zkEVM and its early DeFi applications. The post mixed the author's understanding with material read elsewhere and invited corrections. Jekpopulous2 replied that liquidity was too thin for the activities they wanted and called for a campaign to stimulate use. The contrast is instructive. A person can admire the engineering and still find too little usable liquidity. Neither enthusiasm nor an individual liquidity complaint establishes the experience of every application or the long-term result.

A July post by the same author interpreted growth statistics as evidence that Polygon zkEVM was leading its category. Frozengrandmatetris challenged the ranking and pointed to a comparison source. This account does not reproduce the old metrics as current facts or endorse either ranking without its measurement scope. The episode preserves a common community habit: turning a few attractive indicators into a broad claim of leadership. Different measures, such as value secured, active users or trading depth, answer different questions.

The September 2024 Eggfruit announcement proposed moving the sequencer toward cdk-erigon and encouraged infrastructure providers to transition away from the legacy node. A later reply on October 1 reported successful completion. Keeping both messages distinguishes a deployment record from a schedule. It also shows that this chain was not a single frozen technical object from launch to retirement. Its clients and operational requirements evolved.

A description of the original Go node can illuminate the design without proving that the same client remained the production choice throughout its life.

The original zkevm-node repository was archived in February 2025 and points readers toward cdk-erigon. That archive is valuable historical evidence, but its instructions should not be read as a current recommendation to start a new production deployment. Repository state answers a specific question about software maintenance. It does not, by itself, tell us when a network stopped, how users should recover funds, or whether related teams continue developing other technology. Those conclusions require the corresponding operating and migration records.

Marc Boiron's June 2025 announcement acknowledged delayed development, awkward execution limits, weak differentiation and a slow strategic response. It promised a transition period and reduced support. This was the team's explanation, not an independent causal audit. Replies asked how migration assistance would work.

The April 2026 reminder told protocols to unwind positions and users to leave before the planned July 1 sunset. It distinguished ordinary wallet holdings from contract-held funds. That warning is historically important, but its proposed date and language about assets being abandoned should not override later recovery guidance. Versioned notices can disagree because plans or support arrangements change. A reader trying to recover funds needs the latest applicable instructions, while a historian should preserve what users were told before the operating deadline.

Pasqualou's April 2023 question asked how to verify a particular L2 transfer on Ethereum instead of merely trusting a wallet display. That concern remains relevant to this retired chain's historical importance. Verifiable infrastructure is only useful to ordinary people when the relationship between an action, a batch and a proof is understandable. The record of Polygon zkEVM contains mathematical progress, sincere enthusiasm and real operational disappointments.

Keeping those together explains more than either a victory narrative that ends at launch or a failure narrative that erases the engineering.