"Mastering Distributed Collaboration: The CRDT and OT Handbook"

Sep 2, 2023·

16 min read

"Mastering Distributed Collaboration: The CRDT and OT Handbook"

Introduction: -

Hi there, I'm Gaurav, and I love all things related to designing systems and making software work seamlessly. You know, in today's world, we often collaborate and share information online. Whether it's working together on documents, drawings, or chatting with friends, our digital world is all about teamwork.

But here's the interesting part: Making sure that everyone's changes happen smoothly, especially when many people are working at the same time, is quite a puzzle. That's where CRDTs (Conflict-free Replicated Data Types) and OT (Operational Transformation) come into play.

CRDTs and OT are like the secret sauce behind the scenes. They make sure that when we edit documents together or draw on a shared canvas, everything just works. No mess, no confusion. Think of them as the superheroes who keep our online teamwork in order.

So, let's dive into this fascinating world of CRDTs and OT. We'll uncover how they make our digital collaborations possible and why they matter in our interconnected world.

CRDTs (Conflict-free Replicated Data Types) are like special tools for computers that help with sharing and updating information cleverly. They are super useful when lots of people are working on the same stuff at the same time, like when we edit documents together or chat in real time.

Here's the secret sauce: CRDTs make sure that all our changes fit together perfectly, without getting messy. They do this by making sure that it doesn't matter in what order we make changes; everything will end up looking the same for everyone.

Here are some cool things about CRDTs:

  1. Easy to Use: CRDTs simplify the process of working together on digital stuff. They make sure things stay simple and don't get tangled up.

  2. No Confusion: Even if we make changes at the same time, CRDTs ensure that everything will eventually look the same, so there's no confusion.

  3. Automatic Fixes: If there are any problems or conflicts, CRDTs usually fix them automatically. We don't have to worry too much about it.

  4. Works Well for Many People: CRDTs are great for situations where lots of people need to work together, like in online games or shared documents.

  5. Different Types: There are different kinds of CRDTs for different types of data, like numbers, sets, and lists. Each type knows how to handle its data.

So, in a nutshell, CRDTs are like digital helpers that make sure our teamwork on computers goes smoothly. They make sure our shared stuff stays organized and looks the same for everyone, no matter when or how we make changes.

Imagine you have a shared document that multiple people can edit at the same time. CRDTs help ensure that everyone sees the same version of the document, even if people are making changes simultaneously. Here's how they do it:

1. Commutative Operations: CRDTs use a clever trick. They make sure that all the actions people take on the document (like typing text, deleting, or adding things) can be done in any order, and it won't mess things up. This is like saying you can add your words before or after someone else's, and it will still make sense.

2. Tracking Changes: Each time someone makes a change, CRDTs keep track of it as a special type of operation. These operations are like notes that say what happened, such as "User A added text at position 10" or "User B deleted a word."

3. Sending and Receiving Operations: When someone makes a change, their "operation note" is sent to all the other people using the document. So, if you added a sentence, everyone else gets a note that says, "User Gaurav added a sentence."

4. Merging Changes: Now, here's the magic part. When someone receives a new operation note, they can add it to their version of the document without causing any conflicts. This is because CRDTs ensure that all these operation notes can be combined smoothly, no matter what order they arrive in.

5. Guaranteed Convergence: As everyone keeps making changes and sharing their operation notes, all the versions of the document slowly but surely become identical. This is called "convergence," and it means that everyone sees the same document, even if they started with slightly different versions.

6. No More Conflicts: CRDTs make sure that you don't have to spend time resolving conflicts manually. You don't have to decide whose edit "wins" when two people change the same thing at the same time. CRDTs handle it automatically.

In a nutshell, CRDTs work by allowing people to make changes to shared data in any order they want, tracking those changes as operation notes, and then seamlessly merging these notes. This way, everyone ends up with the same, consistent data, and it all happens without conflicts or confusion. It's like magic for keeping shared documents and collaborative applications in sync!

Imagine you're working on a project with friends, and you all need to edit the same document or use a shared to-do list. Here's why we use CRDTs for such situations:

  1. No Confusing Conflicts: CRDTs make sure that when everyone is editing at the same time, you don't end up with a big mess of conflicting changes. Without CRDTs, you might have to manually figure out whose changes to keep, which can be a headache.

  2. Guaranteed Agreement: With CRDTs, you're certain that, eventually, everyone will see the same version of the document or list, no matter how many changes people make. This means everyone stays on the same page.

  3. Simple and Smooth: Using CRDTs is like having a magical way to make sure all your edits fit together perfectly, no matter when or how you make them. It's like a puzzle where all the pieces automatically snap into place.

  4. Great for Collaboration: CRDTs are fantastic apps where multiple people need to work together on the same thing, like collaborative writing, project planning, or chat apps. They keep everything organized and everyone in sync.

  5. Less Stress: Since CRDTs handle the behind-the-scenes work of combining edits, you can focus on being creative or productive without worrying about edit conflicts or data inconsistencies.

In simple terms, we use CRDTs to keep collaboration smooth and stress-free in apps where many people work together on shared documents or data. They ensure that everyone's changes fit together nicely and that there are no conflicts, making teamwork a breeze!

Implementation of CRDTs:

  1. Data Structures: CRDTs are implemented using specific data structures that allow for conflict-free replication. These data structures are designed to ensure that operations can be applied in any order, and their effects will eventually converge to the same state across all replicas.

  2. Operation Tracking: CRDT implementations keep track of the operations performed on the data. Each operation is associated with metadata like a unique identifier and a timestamp to help with conflict resolution and ordering.

  3. Synchronization Mechanisms: CRDTs require mechanisms for efficiently transmitting and receiving operations between replicas. This often involves network communication to ensure that all replicas are aware of the changes made by others.

  4. Conflict Resolution (Optional): Although CRDTs aim to minimize conflicts, some implementations may include optional conflict resolution mechanisms for handling situations where concurrent operations cannot be easily merged.

Common Data Structures Used in CRDTs:

  1. Grow-Only Set (G-Set): A G-Set is a CRDT that represents a set of unique elements to which new elements can only be added. Once an element is added, it can never be removed. This structure is useful for applications where you want to ensure that data only grows, like tracking unique user IDs in a distributed system.

  2. Observed-Remove Set (OR-Set): The OR-Set is an extension of the G-Set, allowing elements to be removed while still maintaining conflict-free replication. It uses a two-step process: adding an element, and then marking it as removed. This allows for both additions and removals to be tracked.

  3. Last-Write-Wins (LWW) Element Set: In LWW CRDTs, every element has an associated timestamp. When there are conflicting operations (e.g., concurrent updates), the one with the latest timestamp "wins." This approach is often used when you need to prioritize the latest value over others.

  4. Grow-Only Counter: A Grow-Only Counter CRDT allows for incrementing a counter value without the possibility of decrementing it. It ensures that the counter always increases, which can be useful in scenarios where you want to track positive events.

  5. Integer Sequence CRDT: This type of CRDT is designed to represent an ordered sequence of integers. It enables concurrent insertions and deletions while maintaining a consistent order across replicas. This is useful for collaborative text editing and maintaining a history of events.

  6. Map CRDT: Map CRDTs are used to represent key-value pairs where keys are unique identifiers, and values can be updated concurrently. These CRDTs ensure that updates to different keys do not interfere with each other.

  7. Composite CRDT: In some cases, multiple CRDTs are combined to handle more complex data structures. For example, you might use a G-Set for tracking added elements and a LWW Element Set for removed elements to create a comprehensive set CRDT.

CRDTs' choice of data structure depends on the specific application requirements and the type of data being managed. Each data structure is tailored to ensure conflict-free replication, eventual consistency, and the ability to merge concurrent operations seamlessly. By using these data structures and careful implementation, CRDTs enable reliable and efficient distributed collaboration in various applications.

Application Of CRDT

CRDTs (Conflict-free Replicated Data Types) are employed in a variety of applications to facilitate collaborative editing, real-time synchronization, and distributed data management. Here are a few examples of how CRDTs can be used in different applications:

  1. Collaborative Text Editors:

    Example Application: Google Docs

    Use Case: In collaborative text editors like Google Docs, CRDTs are used to allow multiple users to simultaneously edit the same document. Each user's changes (insertions, deletions, formatting) are tracked as CRDT operations, ensuring that all users eventually see the same consistent document, regardless of the order in which edits are made.

  2. Chat Applications:

    Example Application: Slack, WhatsApp

    Use Case: In real-time chat applications, CRDTs can be used to manage message ordering and delivery. Users can send messages concurrently, and CRDTs ensure that messages appear in the correct order for all participants, avoiding out-of-sequence messages.

  3. Distributed Databases:

    Example Application: Riak

    Use Case: In distributed databases, CRDTs can be employed to manage distributed data and maintain data consistency even in the presence of network partitions or node failures. For instance, a CRDT can be used to handle distributed counters, ensuring that increments and decrements are correctly synchronized across replicas.

  4. Version Control Systems:

    Example Application: Git

    Use Case: CRDTs can be used in version control systems to handle concurrent changes made by multiple developers. Each developer's commits and changes can be represented as CRDT operations, allowing for distributed collaboration without conflicts.

  5. Collaborative Drawing Tools:

    Example Application: Figma

    Use Case: In collaborative design and drawing tools like Figma, CRDTs enable multiple users to work together on the same canvas. CRDTs ensure that shapes, lines, and other design elements are consistently synchronized across all participants' views.

  6. Distributed Key-Value Stores:

    Example Application: Cassandra

    Use Case: Distributed key-value stores often use CRDTs to manage distributed data structures like sets or maps. CRDTs help ensure that updates to these data structures propagate correctly across nodes, maintaining data consistency.

  7. Conflict Resolution in P2P Networks:

    Example Application: BitTorrent

    Use Case: CRDTs can be used in peer-to-peer networks to manage distributed resources, resolve conflicts between different peers, and synchronize data in a decentralized manner.

These examples demonstrate how CRDTs can be applied across a wide range of applications to handle distributed data in a conflict-free and eventually consistent manner. CRDTs play a crucial role in ensuring that collaborative and distributed systems function smoothly and maintain data integrity, even in challenging network environments.

Operational Transformation (OT): Transforming Collaborative Editing

Operational Transformation (OT) is a technique used in computer science and distributed systems to enable collaborative editing and synchronization of shared data in real-time. It plays a pivotal role in applications where multiple users need to concurrently work on the same document or data, such as collaborative text editors, drawing tools, and collaborative software development environments. OT ensures that these users can edit and interact with shared content without conflicts or inconsistencies.

Operational Transformation (OT) is a complex but powerful technique used in collaborative applications to allow multiple users to concurrently edit shared data while maintaining consistency and order. Let's break down how Operational Transformation works step by step:

1. Operations and Transformation Functions

  • In collaborative applications, users make changes to shared data by performing operations. These operations represent specific actions like inserting text, deleting characters, or formatting changes.

  • For each type of operation (e.g., inserting text at a particular position), there is a corresponding transformation function. These functions define how an operation should be adjusted when combined with other operations to avoid conflicts.

2. Sending and Receiving Operations

  • When a user acts, such as typing or making edits, their operation is generated locally.

  • These operations are sent to a central server or directly to other users who are part of the collaboration. The operations are timestamped to record the order in which they were created.

  • Other users receive these operations and apply them to their local copies of the shared data.

3. Transformation

  • Before applying received operations, users apply the associated transformation functions to these operations locally. The transformation functions ensure that the operations can be integrated without causing inconsistencies or conflicts.

  • For example, if User A inserts text at position 5, and User B deletes text at the same position, the transformation function will adjust one of these operations to prevent conflicts. It might change User B's operation to delete text at position 4 to avoid overlapping with User A's insert.

4. Applying Operations

  • After transformation, the adjusted operations are applied to the local copy of the data. This results in an updated version of the data on the user's device.

  • The operations are applied in the order they were received, ensuring that the changes happen in the correct sequence.

5. Convergence

  • As users continue to make edits and send operations, all users' local copies of the data eventually converge to the same state. This means that everyone sees the same content and structure, regardless of the order in which edits were made.

  • Convergence is a key objective of Operational Transformation, as it ensures that the shared data remains logically consistent despite concurrent edits from multiple users.

In summary, Operational Transformation works by allowing users to create operations representing their edits, sending these operations to others, transforming received operations to avoid conflicts, applying them in order, and ensuring that all users' copies of the data eventually reach the same consistent state. This process enables real-time collaborative editing in applications such as collaborative text editors, drawing tools, and chat applications while managing concurrency and maintaining data integrity.

Operational Transformation (OT) is of paramount importance in collaborative applications and distributed systems where multiple users work on shared data simultaneously. Its significance lies in its ability to address several critical needs and challenges, making real-time collaboration feasible and efficient. Here's why Operational Transformation is essential:

  1. Real-Time Collaboration: OT enables real-time collaboration by allowing multiple users to work on the same document or data concurrently. This is crucial in applications like collaborative text editors, where users expect to see updates instantaneously.

  2. Conflict Resolution: In collaborative environments, conflicts can arise when multiple users edit the same part of the document simultaneously. OT automates conflict resolution, ensuring that these conflicts are resolved seamlessly without requiring users to intervene manually.

  3. Order Preservation: Maintaining the order of operations is vital in collaborative editing. OT ensures that operations are applied in the correct sequence, preserving the logical structure and flow of the document.

  4. Customizability: OT is adaptable to various data types and application-specific requirements. This flexibility makes it suitable for a wide range of collaborative applications, from text editing to drawing tools to collaborative software development.

  5. Reduced Manual Intervention: By automatically handling conflicts and operation orders, OT reduces the burden on users to manage collaboration issues. This enhances the user experience by minimizing interruptions and frustrations.

  6. Efficiency: OT systems are designed to optimize network communication and minimize latency. This ensures that operations propagate quickly and that users see changes in near real time.

  7. Concurrency Control: OT implementations include mechanisms for safely applying operations concurrently without data corruption. This is essential in scenarios where many users are making edits simultaneously.

  8. Consistency Across Devices: OT ensures that all users' devices eventually converge to the same consistent state. This consistency is essential for applications where users need to access and work on data from different devices.

  9. Versatility: OT can be adapted to various collaboration scenarios and data structures. Whether it's collaborative text editing, drawing tools, or chat applications, OT provides a versatile framework for managing concurrent edits.

  10. History and Version Control: OT systems often maintain a history of operations, allowing users to review and revert changes. This is valuable for tracking edits and providing version control features.

    Operational Transformation is crucial for achieving real-time collaboration without conflicts, ensuring order preservation, reducing manual intervention, and providing versatility across a wide range of collaborative applications. It is a foundational technology that enables users to collaborate seamlessly and efficiently on shared data, making it an indispensable component in modern distributed systems and collaborative software.

  1. Collaborative Text Editors:

    • What It Does: OT is used in applications like Google Docs or Microsoft Word Online, where multiple people can edit the same document simultaneously.

    • How It Helps: OT ensures that everyone can type, edit, and format text in the document at the same time without causing chaos. It makes sure that all changes from different people blend seamlessly.

  2. Collaborative Drawing Tools:

    • What It Does: With tools like Figma, people can work together to create drawings or designs in real-time.

    • How It Helps: OT makes it possible for designers to add shapes, colors, and drawings at the same time without one person's work overwriting someone else's. It keeps the drawing looking great for everyone.

  3. Chat Applications:

    • What It Does: In chat apps like WhatsApp or Slack, you can have conversations with others in real time.

    • How It Helps: OT ensures that your messages appear in the correct order, even if you and your friend send messages at the same time. It keeps the conversation flowing smoothly.

  4. Version Control in Software Development:

    • What It Does: In software development using tools like Git, multiple programmers can work on the same codebase.

    • How It Helps: OT helps keep track of code changes made by different developers, making sure that everyone's code fits together without conflicts.

  5. Collaborative Task Lists:

    • What It Does: In shared to-do list apps, multiple users can add, edit, or check off tasks together.

    • How It Helps: OT ensures that tasks are updated correctly and that everyone sees the same list, whether you're adding new tasks or marking completed ones.

  6. Conflict Resolution in Databases:

    • What It Does: In databases that multiple user's access, like customer databases for online stores, OT helps manage updates and changes.

    • How It Helps: OT ensures that changes from different users, such as updating customer details or inventory, don't clash or create errors in the database.

In simple terms, OT is like a secret ingredient that makes sure everyone can work together on the same digital stuff without messing things up. It helps in collaborative writing, drawing, chatting, software development, managing tasks, and keeping databases organized, so everyone's efforts fit together perfectly.


In this blog, we've explored two important things: CRDTs and OT. CRDTs are like secret tools that help computer programs work together smoothly, especially when lots of people are involved. OT, on the other hand, is like a magic spell that ensures everything stays organized.

Now, here's the interesting part: These things we've learned about, CRDTs and OT, are really helpful not just for understanding tech but also for your career. LinkedIn, which is like a big online club for professionals, is a great place to learn more and connect with others.

Thank you for joining me on this journey of discovery with CRDTs and OT. Let's stay connected and keep the conversation alive!

Did you find this article valuable?

Support Gaurav Dhak by becoming a sponsor. Any amount is appreciated!