- Core Identity
XDAG's core architecture is built upon two complementary mechanisms:
References organize blocks into a DAG, while time-sliced PoW main-block mechanisms establish the ledger's ordering framework.
References define how blocks are connected, forming the ledger's directed acyclic graph structure. Time-slicing and PoW-based main-block mechanisms provide the temporal organization and ordering basis for the ledger.
Together, these mechanisms form XDAG's distinctive approach to ledger organization.
XDAG should therefore not be understood merely as a PoW project that uses a DAG data structure. Its defining technical characteristic lies in the combination of a reference-driven DAG structure and time-sliced PoW ordering.
Reference-Driven DAG and Time-Sliced PoW Ordering are the two central elements of XDAG's technical DNA.
- Reference-Driven DAG: Organizing Blocks Through References
In XDAG, blocks are connected through references, forming a Directed Acyclic Graph (DAG).
Unlike a linear ledger that relies primarily on sequential block connections, a DAG allows blocks to establish richer relationships through references. Block relationships are not restricted to a single linear sequence, but form a network structure according to the rules defined by the protocol.
In the current design, Transaction Blocks (TX Blocks) and Link Blocks both participate in the DAG structure, but serve different purposes:
Transaction Blocks (TX Blocks): Carry transactions. A single TX Block can contain multiple transactions.
Link Blocks: Participate in the structural relationships within the DAG and do not serve as transaction containers.
References: Establish connections between blocks and provide the relational foundation of the DAG.
XDAG is therefore neither a system in which every transaction corresponds to exactly one block, nor simply a linear ledger formed by linking blocks one after another.
Understanding XDAG begins with understanding how references connect blocks and how these relationships form the overall DAG structure of the ledger.
- Time-Sliced PoW Ordering: Establishing Ledger Order
A DAG defines how blocks are related, but relationships alone do not automatically establish a definitive transaction order.
XDAG uses time-slicing and PoW-based main-block mechanisms to establish the ledger's ordering framework.
Under the current protocol design, main blocks follow an approximately 64-second cadence and participate in the formation and selection of the main-block chain through PoW. This chain provides the temporal and ordering foundation of the ledger. References between main blocks and other blocks connect the DAG structure to this ordering framework.
The mechanism consists of three complementary elements:
Time-slicing: Provides the temporal framework for main-block production and ledger organization.
PoW: Provides the proof-of-work basis for main-block production and selection.
The main-block chain and its references: Connect block relationships within the DAG to the ledger's ordering framework.
Together, these elements form XDAG's temporal organization and ordering mechanism.
It is important to distinguish ordering from confirmation. Time-slicing does not mean that all transactions within the same interval are confirmed simultaneously. Nor does it mean that every block in the DAG must be referenced directly by a main block.
Block inclusion, transaction validity, ordering, conflict resolution, and confirmation status must ultimately follow the specific rules defined by the protocol.
- Transaction Representation and Ledger Organization
XDAG organizes transactions within Transaction Blocks, and a single TX Block can carry multiple transactions.
Through references, TX Blocks participate in the DAG structure and establish ordering relationships with the main-block chain according to the rules defined by the protocol. Transaction representation, block relationships, and ledger ordering therefore work together as parts of a unified system.
To understand XDAG's transaction organization, three layers should be distinguished:
Transaction layer: How transactions are organized and carried within TX Blocks.
Relationship layer: How references connect blocks to form the DAG.
Ordering layer: How time-sliced PoW main-block mechanisms establish the ledger's ordering framework.
Each layer has a distinct responsibility, but they are not independent of one another. Together, they define XDAG's technical structure for transaction representation, block relationships, and ledger ordering.
Transaction validity, ordering, double-spend conflict resolution, and confirmation status must be determined by the protocol's rules. They should not be inferred solely from the existence of a reference path within the DAG.
- Preserving XDAG's Technical Foundation
XDAG's evolution is not about becoming another Kaspa, nor about following whichever technical architecture happens to be popular.
Different DAG projects may adopt different approaches to block organization, ordering, and consensus. Comparing these approaches helps clarify their respective principles, strengths, and trade-offs. It does not mean that every project should converge on the same technical direction.
XDAG should remain committed to its own foundation:
A reference-driven DAG with time-sliced PoW ordering.
This technical foundation should guide future architectural decisions and engineering development.
P2P networking, data synchronization, storage, recovery, and node efficiency can all improve over time. These improvements, however, must serve the core ledger mechanism rather than introduce unnecessary architectural complexity in pursuit of technical sophistication.
Changes affecting ledger determinism, transaction ordering, or consensus rules require a clearly identified problem, sound technical justification, and rigorous testing and review.
Technology should continue to advance without sacrificing XDAG's fundamental identity.
- Minimalism and Understandability
XDAG should remain committed to minimalist design.
Minimalism does not mean leaving out necessary functionality. It means that every mechanism must have a clear purpose, a well-defined boundary, and a sufficient reason to exist.
Problems that can be solved through simple and reliable rules should not be burdened with unnecessary architectural complexity.
The same principle applies to AI-assisted development.
AI can assist with code analysis, implementation, and testing, but it cannot replace developers' understanding of the protocol, architecture, and system behavior. Every change entering the core system must remain explainable, reviewable, testable, and maintainable by human developers.
Code that runs is not necessarily code that is correct. Passing tests does not, by itself, prove that an implementation complies with the protocol.
System complexity must never exceed our ability to understand, verify, and control it.
This is not merely an engineering quality requirement. It is fundamental to XDAG's long-term security, reliability, and sustainable evolution.
- Conclusion: Preserve the Core, Continue to Evolve
XDAG's technical DNA can be summarized as follows:
A reference-driven DAG, ordered through time-sliced PoW, continuously improving security, efficiency, and maintainability while preserving architectural simplicity and deterministic ledger rules.
This defines XDAG's technical structure and establishes a guiding principle for its future development.
XDAG does not need to become another project, nor should it lose its direction while pursuing technical trends.
Meaningful evolution comes from deeply understanding and strengthening its own technical foundation, making the system more reliable, efficient, open, and maintainable.
Preserve the core. Continue to evolve. Keep XDAG true to itself.
Top comments (1)
XDAG 技术 DNA
以引用关系驱动的 DAG 与时间切片 PoW 排序
XDAG — A Reference-Driven DAG with Time-Sliced PoW Ordering
XDAG 的核心技术架构建立在两个相互配合的机制之上：
以引用关系组织区块，形成 DAG；以时间切片 PoW 主块机制建立账本的排序框架。
引用关系定义区块之间的连接方式，构成账本的有向无环图结构；时间切片与 PoW 主块机制则为账本提供时间组织和排序依据。
两者相互配合，构成 XDAG 独特的账本组织方式。
因此，XDAG 不应仅仅被理解为一种采用 DAG 数据结构的 PoW 项目。它的核心技术特征，在于引用关系构建的 DAG 结构与时间切片 PoW 排序机制之间的结合。
Reference-Driven DAG 与 Time-Sliced PoW Ordering，是 XDAG 技术 DNA 的两个核心组成部分。
在 XDAG 中，区块通过引用关系建立联系，并由此形成有向无环图（Directed Acyclic Graph，DAG）。
与单纯依靠前后顺序连接的线性账本不同，DAG 允许区块通过引用关系建立更丰富的关联。区块之间的关系不再局限于单一的线性连接，而是按照协议规则形成网络化的结构。
在当前设计中，交易块（TX Block）和链接块（Link Block）共同参与 DAG 的构成，但承担不同职责：
交易块（TX Block）： 承载交易，一个交易块可以包含多笔交易。
链接块（Link Block）： 参与 DAG 的区块关系结构，不作为交易承载容器。
引用关系（Reference）： 建立区块之间的连接，构成 DAG 的关系基础。
因此，XDAG 并不是简单的“一笔交易对应一个区块”，也不是传统意义上由区块逐个串联形成的线性账本。
理解 XDAG，首先需要理解区块之间的引用关系，以及这些关系如何共同构成账本的 DAG 结构。
DAG 解决的是区块之间如何建立关系的问题，但区块之间存在关联，并不意味着交易顺序已经自动确定。
XDAG 通过时间切片与 PoW 主块机制，为账本建立排序框架。
按照当前协议设计，主块以约 64 秒为基本时间节奏产生，并通过 PoW 机制参与主块链的形成与选择。主块链构成账本时间组织和排序的基础，主块与其他区块之间的引用关系，则将 DAG 结构与这一排序框架联系起来。
这一机制包含三个相互配合的部分：
时间切片： 为主块产生和账本组织提供时间框架。
PoW： 为主块的产生与选择提供工作量证明基础。
主块链及其引用关系： 将 DAG 中的区块关系纳入账本排序框架。
三者共同构成 XDAG 的时间组织与排序机制。
需要明确的是，排序不等同于确认。时间切片不意味着同一时间片中的所有交易同时完成确认；DAG 中的区块也不意味着都必须被主块直接引用。
区块的纳入、交易的有效性、先后顺序、冲突处理及确认状态，最终都应遵循协议定义的具体规则。
XDAG 通过交易块组织交易，一个交易块可以承载多笔交易。
交易块通过引用关系参与 DAG 结构，并按照协议规定的规则与主块链建立排序关系。由此，交易承载、区块关联与账本排序形成相互配合的整体。
理解 XDAG 的交易组织方式，需要区分三个层次：
交易层： 交易如何被组织并承载在交易块中。
关系层： 区块如何通过引用关系连接，形成 DAG。
排序层： 时间切片 PoW 主块机制如何为账本建立排序框架。
三个层次各有职责，但并非彼此独立。它们共同构成 XDAG 从交易组织、区块关联到账本排序的技术结构。
交易的有效性、先后顺序、双花冲突裁决及确认状态，应由协议规则共同确定，不能仅凭 DAG 中存在某条引用路径就作出判断。
XDAG 的演进，不是为了成为另一个 Kaspa，也不是为了追随某一种流行的技术架构。
不同的 DAG 项目可以采用不同的区块组织方式、排序机制和共识设计。比较不同技术路线的意义，在于理解各自的原理、优势与取舍，而不是让所有项目走向同一种技术方向。
XDAG 应当坚持自己的核心设计：
以引用关系驱动 DAG，以时间切片 PoW 建立排序。
这一技术内核应当成为未来架构设计和工程演进的基本依据。
P2P 网络、数据同步、存储、恢复和节点运行效率，都可以随着技术发展持续改进。但这些改进应当服务于核心账本机制，而不是为了追求技术复杂度而引入不必要的架构。
尤其是涉及账本确定性、交易排序和共识规则的改动，必须建立在明确的问题分析、充分的技术论证和严格的测试审查之上。
技术可以不断进步，但不能在演进过程中失去自己的核心身份。
XDAG 应当坚持极简设计原则。
极简并不意味着缺少必要的功能，而是要求每一项机制都有明确的作用、清晰的边界和充分的存在理由。
能够通过简单、可靠的规则解决的问题，不应被不必要的复杂架构取代。
这一原则同样适用于 AI 辅助开发。
AI 可以参与代码分析、编写和测试，但不能代替开发者对协议、架构和系统行为的理解。进入核心系统的每一项代码改动，都应当能够被人类开发者解释、审查、测试和维护。
代码能够运行，不代表设计一定正确；测试通过，也不意味着实现必然符合协议要求。
系统复杂度不能超过人类理解、验证和控制它的能力。
这不仅是工程质量要求，也是 XDAG 长期保持安全性、可靠性和可持续演进能力的基础。
XDAG 的技术 DNA，可以归纳为：
以引用关系驱动 DAG，以时间切片 PoW 建立排序，在保持架构简洁与账本规则确定性的基础上，持续提升系统的安全性、效率和可维护性。
这是 XDAG 对自身技术结构的定义，也是未来发展的基本原则。
XDAG 不需要成为另一个项目，也不应在追逐技术潮流的过程中失去自己的方向。
真正有价值的演进，是在深入理解自身技术内核的基础上，不断完善系统，使其变得更加可靠、高效、开放和易于维护。
保持内核，持续演进。
让 XDAG 始终是 XDAG。