I’ve been watching Cardano’s development rhythm for years, and every so often a roadmap update lands that feels less like another incremental tweak and more like a genuine shift in ambition. The latest one, centered on the Dijkstra upgrade, is exactly that kind of moment. Instead of promising everything at once, the team has split the work into two clear phases that stretch from late 2026 into 2027. Linear Leios arrives first, followed by Ouroboros Peras. The dates are still estimates, of course, but the technical scope looks locked in, and the community is already moving into readiness mode. That alone makes this worth a closer look.
Cardano Dijkstra Sets A Clear Two-Phase Path
What stands out immediately is how deliberate the structure feels. Phase 1 introduces the new Dijkstra ledger era together with Linear Leios and Nested Transactions. Code completion is targeted for the fourth quarter of 2026. Phase 2 then activates Ouroboros Peras through a separate intra-era hard fork aimed at the second quarter of 2027. The published schedule is careful to note that these are estimates rather than guarantees. Testing, governance votes, and real-world network behavior can still push the final dates later. Still, having a written sequence that the core teams appear to agree on is progress in itself.
Intersect’s mid-August update framed the current focus as ecosystem readiness, testnet reporting, and support for alternative node implementations. That last point matters more than it might seem at first glance. Cardano has long relied heavily on its Haskell node. Bringing other clients into production-grade status reduces single-implementation risk and gives the network more resilience. Amaru, the open-source Rust implementation, is already validating and syncing to the tip. Mainnet block production for Amaru is currently targeted for November 2026, which would place it right around the expected Phase 1 window.
Linear Leios As The Core Scaling Move In Phase 1
Linear Leios is the headline feature of the first hard fork. It does not throw away Cardano’s existing Praos security model. Instead it adds Endorser Blocks that can reference extra transactions and have those transactions certified by stake-based committees. The idea is to raise throughput while keeping the base consensus guarantees intact. I’ve always appreciated when a design tries to improve capacity without rewriting the security assumptions from scratch. That approach tends to age better.
Phase 1 also brings Nested Transactions, new serialization structures for transactions and blocks, PlutusV4 adjustments, and a set of new protocol parameters. Critically, it installs the codec extensions and parameters that Peras will later need, but it does not turn Peras on. That separation keeps the first upgrade focused and reduces the surface area that has to be tested all at once. In practice it means developers can ship the ledger changes and the scaling pieces now, then layer the faster settlement protocol on top once the foundation is proven.
Earlier this year the van Rossem hard fork moved mainnet to Protocol Version 11 and laid some of the technical groundwork for Dijkstra and Leios. That earlier step makes the current roadmap feel less like a sudden leap and more like the continuation of a longer sequence. The pieces are starting to connect in a way that looks intentional rather than opportunistic.
Ouroboros Peras And Faster Settlement In Phase 2
Ouroboros Peras is scheduled for the second phase. The protocol adds a voting layer that lets committees of stake pool operators vote on recent chain tips. The practical result is a path to settlement that can be faster than the standard Praos chain-depth rules. Once Phase 1 is live and the necessary ledger structures and parameters are in place, Peras can activate through its own Preview and Pre-production deployments followed by another mainnet governance action.
The current target sits in the second quarter of 2027. That date will almost certainly move if Phase 1 slips or if the governance process takes longer than expected. Still, the sequencing is clear: Phase 1 first, then Peras. Trying to do both in a single hard fork would have created a much larger testing burden and a higher chance of unexpected interactions. Splitting them feels like the more responsible engineering choice.
Amaru And The Push For Node Diversity
One of the quieter but more important threads running through the Dijkstra work is the push toward greater client diversity. Amaru is already relay-capable and can validate and synchronize with the chain tip. Its development tracker lists a general block-producer release targeted for the end of September, a Dijkstra-compatible block producer milestone near the end of October, and a Leios-compatible release toward the end of November. These are development milestones rather than guaranteed mainnet activation dates, but they give a sense of the pace.
Having a production-ready Rust node would meaningfully reduce Cardano’s reliance on the Haskell implementation. In a network that values decentralization, that kind of diversity is not just nice to have; it is part of the long-term resilience story. Intersect is also building a Dijkstra readiness tracker that covers testnet performance and broader ecosystem preparation, and it is inviting alternative node teams into the weekly Hard Fork Working Group. That collaborative tone is useful. Hard forks go more smoothly when more implementers are involved early.
Governance Must Move Before The Protocol Can
Dijkstra introduces new protocol parameters that Cardano governance cannot adjust unless those parameters appear explicitly inside the Constitution’s guardrails. Input Output therefore plans a narrow constitutional amendment that simply adds the relevant parameters and their permitted ranges. The proposal does not touch governance roles, voting thresholds, or core constitutional principles. It is deliberately limited in scope.
The current target is to submit that governance action no later than Epoch 655, which begins on September 11. Community discussion is already under way through Intersect’s Constitutional Amendment Portal. By mid-August the portal had already received four initial submissions. Getting this amendment through on schedule is a prerequisite for the rest of the Dijkstra timeline. Without the constitutional update, the new parameters cannot be activated even if the code is ready.
I’ve found that governance steps are often the part of a roadmap that gets under-estimated. Code can ship on a technical schedule; on-chain votes and community discussion move at a different rhythm. Keeping the amendment narrow should help, but it still needs careful communication so that stakeholders understand exactly what is changing and what is not.
What The Near-Term Work Actually Looks Like
Right now the focus sits on finishing Dijkstra Phase 1 code, establishing clear readiness criteria, and moving the first phase through Preview and Pre-production environments. Only after those stages succeed does a mainnet governance vote become realistic. Intersect’s latest update states that the agreed scope and target dates remain unchanged. That consistency is reassuring. Roadmaps that keep shifting every few weeks tend to erode confidence.
The main caveat remains timing. The Haskell node team is working toward Phase 1 mainnet delivery by the end of 2026, yet the formal roadmap describes the fourth quarter as a code-completion target. Governance and community testing can still push the actual activation later. Peras stays targeted for the second quarter of 2027, but only after Phase 1 is successfully deployed. In other words, the sequence is solid; the calendar is still flexible.
The agreed technical scope and schedule remain in place, with planning now shifting toward ecosystem readiness and support for alternative node implementations.
That kind of measured language is typical of mature blockchain development. Big promises are easy. Delivering a hard fork that actually improves throughput and settlement without breaking existing applications is harder. Cardano has historically preferred the slower, more deliberate path. The Dijkstra plan continues that pattern.
Why The Phased Approach Matters
Perhaps the most interesting design decision is the explicit separation between Linear Leios and Peras. Shipping both in one hard fork would have looked more ambitious on a slide deck. It would also have multiplied the testing surface and increased the chance of unexpected interactions. By installing the codecs and parameters in Phase 1 and activating the voting layer later, the teams can validate the scaling changes in production before adding the settlement acceleration. That is good engineering hygiene.
Nested Transactions and the new serialization formats are less flashy than Leios itself, yet they matter for long-term maintainability. Cleaner structures make future upgrades less painful. PlutusV4 adjustments continue the steady evolution of the smart-contract environment. None of these pieces alone would justify a hard fork, but together they form a coherent ledger-era transition.
I’ve always believed that the projects that survive the longest are the ones that treat upgrades as a continuous process rather than a series of dramatic events. Cardano’s current roadmap reflects that mindset. The Dijkstra name itself is a nod to the computer scientist Edsger Dijkstra, whose work emphasized correctness and structured design. Whether intentional or not, the naming feels appropriate.
Node Diversity Beyond Amaru
Amaru is the most visible alternative client right now, but the broader invitation to other node teams is equally important. A healthy ecosystem should support multiple independent implementations that can produce and validate blocks. Single-client dependence has bitten other networks in the past. Reducing that risk is a quiet but valuable side effect of the Dijkstra work.
The readiness tracker Intersect is building should help surface performance data and ecosystem gaps early. Weekly Hard Fork Working Group meetings give implementers a regular forum. Those process details rarely make headlines, yet they often determine whether a hard fork feels smooth or chaotic when the time comes.
Constitutional Guardrails And Parameter Safety
The decision to keep the constitutional amendment narrow is worth emphasizing. Adding specific parameters and their allowed ranges without altering governance roles or voting thresholds reduces the political surface area of the vote. Stakeholders can evaluate the technical necessity of the new parameters without also debating broader changes to the governance system. That separation should improve the odds of timely passage.
Epoch 655 beginning on September 11 gives a concrete deadline for submission. Community discussion through the Constitutional Amendment Portal is already under way. Early engagement matters. Last-minute surprises tend to slow governance processes more than well-signaled, limited proposals.
Looking Past The Immediate Timeline
Once both phases are live, Cardano should have higher throughput from Linear Leios and faster settlement from Peras. Those are meaningful improvements for users and developers who have waited for better capacity characteristics. The network will still face the usual challenges of adoption, application quality, and competitive pressure from other ecosystems. Better base-layer performance does not automatically create demand, but it removes one set of constraints that have been discussed for years.
I tend to be cautious about projecting price or market share from technical upgrades alone. Markets move on many factors. What does seem clear is that the engineering path is more structured now than it has been at certain points in the past. A written two-phase plan with explicit dependencies and readiness criteria is better than open-ended promises.
The next few months will show how well the code-completion targets hold and how smoothly the constitutional amendment proceeds. If Phase 1 lands near the end of 2026 and Peras follows in 2027, Cardano will have delivered a significant scaling and settlement upgrade on a multi-year horizon. That outcome is still contingent on testing and governance, yet the roadmap itself looks more grounded than many earlier versions.
Practical Implications For Builders And Operators
For application developers the most immediate relevance is the set of new transaction and block structures plus PlutusV4 changes. Nested Transactions in particular may open different patterns for composing complex operations. Teams should plan to test against Preview and Pre-production environments once those become available for Dijkstra. Waiting until mainnet activation to discover integration issues is rarely pleasant.
Stake pool operators will need to track the Amaru milestones if they want to experiment with alternative clients, and they will eventually participate in the governance votes that activate each phase. The Peras voting layer will introduce new committee responsibilities once it is live. Understanding those mechanics ahead of time will reduce surprises.
Exchanges and infrastructure providers should watch the serialization changes carefully. New formats can create temporary friction during the transition window. Early engagement with the readiness tracker and the Hard Fork Working Group is the practical way to stay ahead of those details.
A Note On Expectations And Reality
Every major blockchain upgrade arrives with optimistic timelines and then meets the friction of real-world testing and governance. Cardano is no exception. The current Dijkstra schedule is more disciplined than some previous roadmaps, yet the disclaimer that dates are estimates remains important. Code completion in the fourth quarter of 2026 does not automatically equal mainnet activation in the same window. Governance and community testing sit in between.
That realism is healthy. Projects that over-promise and then repeatedly slip tend to lose credibility. Projects that under-promise and then deliver on a clear sequence tend to build more durable trust. The Dijkstra plan currently sits closer to the second category. Whether it stays there depends on execution over the coming months.
I’ve watched enough upgrade cycles across multiple networks to know that the quiet, process-oriented work often matters more than the flashy feature announcements. Intersect’s emphasis on readiness tracking, alternative node support, and a narrow constitutional amendment suggests that the teams understand this. The next test will be whether the community can move the governance pieces on schedule while the engineers finish the code.
Final Thoughts On The Road Ahead
Cardano’s Dijkstra upgrade is not a single dramatic event. It is a two-phase sequence that begins with Linear Leios and Nested Transactions in the Dijkstra ledger era, then continues with Ouroboros Peras roughly half a year later. Amaru’s progress toward block production adds a useful layer of client diversity. A limited constitutional amendment is required before the new parameters can be used. The calendar targets sit at the end of 2026 for Phase 1 code and the second quarter of 2027 for Peras activation.
None of those dates are guaranteed. Testing and governance can still shift them. Yet the structure of the plan itself feels more coherent than many earlier versions of the Cardano roadmap. The separation of concerns, the attention to node diversity, and the narrow framing of the constitutional change all point toward a more mature delivery process.
For anyone following the network, the practical next steps are straightforward. Watch the readiness tracker. Track the constitutional amendment discussion. Pay attention to Preview and Pre-production performance data once it appears. And keep an eye on Amaru’s milestone dates. Those signals will tell us more about the real trajectory than any single announcement.
In the end, scaling a blockchain while preserving its security assumptions and decentralization properties is hard work. Linear Leios and Peras are attempts to improve throughput and settlement without discarding the core design. Whether they succeed at the scale the community hopes remains to be proven in production. The roadmap, however, is clearer than it has been in some time. That clarity is itself a form of progress.
The coming year will show how well the estimates hold. If Phase 1 lands near the planned window and Peras follows in 2027, Cardano will have completed a meaningful multi-year upgrade cycle. If delays appear, the phased structure at least gives the community a transparent way to understand where the friction is occurring. Either outcome will be more informative than vague promises. For now, the plan is on the table, the scope is agreed, and the work is under way.
