New GRIDNEToken Mobile App - Release Notes

GRIDNET OS - Main
1

We are thrilled to announce the latest update to the GRIDNEToken mobile app. This release represents a major leap forward in functionality, security, and user experience. Our team has worked tirelessly to implement cutting-edge technologies and innovative features that set a new standard in the blockchain mobile app space.

1. Networking Enhancements

QUIC Protocol Integration

  • The app is now fully QUIC protocol capable, completely replacing the previous UDT protocol.

  • We have successfully implemented Microsoft’s QUIC protocol on both iOS and Android devices. This is a significant achievement in cross-platform compatibility, as it required extensive optimisation and adaptation for mobile environments.

  • Technical details: The QUIC implementation provides multiplexed connections over UDP, offering reduced latency and improved congestion control compared to TCP-based protocols. This allows for faster and more reliable data transfer, especially in mobile networks with varying signal strengths.

Improved Transaction Broadcasting

  • We’ve completely overhauled our transaction broadcasting mechanism to ensure maximum reliability and efficiency.

  • Transactions are now immediately stored in a local SQL database upon user consent. This ensures that no transaction is lost, even in cases of unexpected app termination or device shutdown.

  • Our newly developed internal AI subsystem and advanced scheduling algorithms work in tandem to manage transaction persistence. These systems continuously monitor the state of each transaction and make intelligent decisions about when and how to rebroadcast if necessary.

  • Technical deep dive: The scheduling system uses a combination of exponential backoff and network condition monitoring to optimise rebroadcast timing. It takes into account factors such as network latency, node responsiveness, and transaction priority to ensure the highest probability of successful propagation.

Onion Routing Support

  • We’ve implemented full support for Onion Routing, including the ability to use a web-browser as the final endpoint via a web-socket-based connection.

  • The onion-routed session key establishment between the mobile app and web-browser has been significantly improved, enhancing both security and performance.

  • Technical explanation: Our implementation uses a multi-layered encryption approach, where each node in the routing path only knows its immediate predecessor and successor. This ensures that no single node has complete information about the data path, significantly enhancing user privacy.

2. Security and Cryptography

Revised Cryptographic Protocols

  • We’ve implemented fully revised custom cryptographic protocols based on Elliptic Curve Integrated Encryption Scheme (ECIES).

  • A new custom Diffie-Hellman based session establishment protocol has been integrated, providing robust key exchange mechanics.

  • We’ve introduced an innovative mode where the session key is derived as a by-product of Authenticated Encryption with Associated Data (AEAD) container decryption. This approach combines encryption and authentication in a single step, improving both security and performance.

  • Technical details: Our ECIES implementation uses custom cryptography Curve25519 for elliptic curve operations to allow for both key establishment, encryption and signatures with same private key. The Diffie-Hellman key exchange uses ephemeral keys to provide forward secrecy, ensuring that past communications remain secure even if long-term keys are compromised.

PIN Protection

  • Users can now set a PIN for enhanced security. This adds an additional layer of protection beyond the existing biometric and password options.

  • Technical implementation: The PIN is never stored directly. Instead, we use a key derivation function (specifically, PBKDF2 with a high iteration count) to create a key from the PIN. This derived key is then used in combination with device-specific information to encrypt sensitive data.

Improved Nonce Management

  • We’ve vastly improved our autonomous nonce management system to prevent replay attacks while allowing for rapid transaction sequences.

  • The app now intelligently increments nonces locally in advance of network confirmation. This allows users to issue multiple transactions in quick succession without waiting for network validation of each nonce.

  • A sophisticated fallback mechanism reverts to the network-provided nonce after a configurable timeout period, ensuring consistency with the blockchain state.

  • Technical deep dive: Our nonce management system uses a combination of local caching, predictive incrementation, and network synchronisation. It maintains a local high-water mark for nonces and uses atomic operations to ensure thread-safety when incrementing. The system also includes a recovery mechanism to handle cases where local and network nonces diverge.

3. User Interface and Experience

GRIDNET Halo Enhancements

  • The GRIDNET Halo has been completely revamped with advanced, reactive shader mechanics.

  • New touch gestures have been implemented: Swipe left for Transaction History, swipe down to lock the app, and swipe up to unlock.

  • The Halo now dynamically reacts to current network conditions and state, providing visual feedback on connectivity and transaction status.

  • Technical implementation: We’ve utilised OpenGL ES 3.0 for shader programming, allowing for complex, real-time visual effects while maintaining excellent performance. The shader programs dynamically adjust based on incoming network data, creating a responsive and informative user interface element.

Advanced Rendering Techniques

  • We’ve incorporated cutting-edge, on-hardware, on-GPU custom rendering of UI elements, working seamlessly alongside standard Qt controls.

  • This is particularly noticeable in the Hexi Screen, where Hexi’s thoughts are displayed with fluid, high-performance animations.

  • Technical details: Our custom rendering pipeline uses Vulkan on Android and Metal on iOS for low-level GPU access. We’ve implemented a custom scene graph that integrates with Qt’s existing rendering system, allowing for high-performance custom elements alongside standard UI components.

Improved Layout and Stability

  • The app’s layout has been vastly improved, with a focus on intuitive navigation and efficient use of screen real estate.

  • We’ve enhanced support for back and menu buttons on Android platforms, addressing long-standing Qt fullscreen bugs through custom JNI-level mechanics.

  • Technical deep dive: Our JNI implementation hooks directly into the Android activity lifecycle, allowing us to intercept and handle system events before they reach Qt. This gives us fine-grained control over the app’s behaviour in response to system events, resulting in a more stable and responsive user experience.

Computer Vision Enhancements

  • Our QR code recognition system has been optimised to achieve recognition times of less than 10ms under 89% of lighting conditions classified as ‘good or semi-good’ in our virtual 3D vision testing environment.

  • Technical implementation: We’ve used a combination of OpenCV for image preprocessing and a custom-trained convolutional neural network for QR code detection and decoding. Hexi processing on mobile is aided by a relatively small neural network which optimised for mobile devices using TensorFlow Lite, allowing for rapid inference even on lower-end hardware.

4. New Features and Functionality

Transaction History View

  • We’ve introduced a new Transaction History View, accessible by swiping left on the GRIDNET Halo.

  • This view displays all transactions associated with the current active identity, including those not issued from the current device.

  • The history and account balance are automatically refreshed every few seconds to ensure up-to-date information.

  • Technical details: The Transaction History View uses a virtual list view for efficient memory usage when displaying large numbers of transactions. It implements a local caching mechanism that stores transaction data in an SQLite database, with a background service that periodically syncs with the blockchain to fetch updates.

Sub-Identity Toggling

  • Users can now toggle between sub-identities, allowing quick switching between unrelated wallet addresses based on a single Master Private Key.

  • A simple index increment in the Settings view updates all internals, including the Hexi AI subsystem.

  • Technical implementation: We use BIP32 hierarchical deterministic wallets to derive sub-identities. Each sub-identity is derived from the master seed using a different derivation path. The app maintains a mapping between user-friendly labels and derivation paths, allowing for quick switching without exposing the underlying cryptographic complexities to the user.

AI Assistant (Hexi) Improvements

  • Hexi’s AI algorithms have been enhanced to provide a more immersive and responsive experience.

  • Hexi now monitors which fields users interact with and tracks the state of dispatched tasks, allowing for more contextual and helpful responses.

  • Technical deep dive: Hexi’s AI system uses a combination of rule-based expert systems and machine learning models. The rule-based system handles task tracking and immediate responses, while a transformer-based language model (fine-tuned GPT variant) generates more complex, contextual responses. The system uses transfer learning techniques to adapt to individual user behaviour over time.

Advanced Local Storage

  • We’ve implemented a self-maintained, advanced SQL-based storage subsystem for efficient data management.

  • This system is used for storing and managing data retrieved from the network, including comprehensive transaction tracking and state updates.

  • Technical details: The storage system uses SQLite as its backend, with a custom ORM layer for type-safe database access. We’ve implemented a multi-version concurrency control (MVCC) system to allow for concurrent read/write access without locking, significantly improving app responsiveness during data-intensive operations.

5. Performance and Stability

Networking Subsystem Overhaul

  • Our networking subsystem has been completely revised and is now fully reliant on the QUIC protocol.

  • The app can now maintain multiple connections with multiple nodes simultaneously, significantly improving network resilience and transaction propagation speed.

  • Transaction broadcasting has been improved to occur across multiple nodes concurrently, increasing the likelihood of rapid inclusion in the blockchain.

  • Technical deep dive: Our QUIC implementation uses a custom congestion control algorithm optimised for blockchain traffic patterns. We’ve implemented a novel peer discovery and selection algorithm that takes into account node reputation, geographic distribution, and network conditions to maintain an optimal set of connections.

Transaction Monitoring and Management

  • The app now continuously monitors the state of broadcasted transactions, providing real-time updates to the user.

  • An intelligent re-broadcasting system automatically retries failed transactions based on network conditions and transaction priority.

  • We’ve introduced a new notification system in the Transaction History View that provides detailed status updates for each transaction.

  • Technical implementation: The transaction monitoring system uses a combination of local state tracking and selective network polling. It implements an event-driven architecture where changes in transaction status trigger updates throughout the app. The re-broadcasting system uses a machine learning model to predict the optimal timing and node selection for retries, based on historical data and current network conditions.

6. Additional Improvements

  • Extensive stability and responsiveness improvements across a wide range of Android and iOS devices.

  • Optimisation of memory usage and battery consumption through intelligent background task scheduling and efficient use of system resources.

  • Implementation of crash analytics and automatic error reporting to help us quickly identify and resolve any issues that may arise in real-world usage.

We’re incredibly proud of this major update to the GRIDNEToken mobile app. Our team has pushed the boundaries of what’s possible in a blockchain mobile application, delivering a product that we believe sets new standards for security, performance, and user experience. We deeply appreciate your continued support and feedback as we strive to deliver the best possible product to our users.

As always, we welcome your thoughts and suggestions. Please don’t hesitate to reach out to our support team with any questions or feedback you may have about this new release.

1 Like