Traide Changelog

Fixed an issue where certain mutations could fail to properly commit database transactions, potentially leading to incomplete data updates.

Restored proper access control for product, attribute, and order queries, ensuring users only see data they have permission to access.

Fixed several checkout mutation behaviors that were inadvertently changed, restoring expected functionality for checkout operations.

Resolved multiple issues affecting API query reliability and data consistency.

Fixed multiple API query and mutation issues that were affecting marketplace functionality, improving overall API reliability and response accuracy.

This release contains internal improvements and maintenance updates.

Fixed an issue where draft orders could be created without a customer email address in certain scenarios, which could cause problems with order processing and notifications.

Fixed an issue where the NauticalOrderEvent quantity field could return incorrect values in certain scenarios.

AI-powered code review​

Integrated Claude-powered code review into our continuous integration pipeline, providing domain-specific feedback on pull requests to improve code quality and catch issues earlier.

Performance improvements​

Made several performance improvements across the platform:

• Loading the seller list and product schema is now significantly faster, eliminating query inefficiencies.
• Webhook processing now uses less memory and handles high volumes more reliably.

Fixed an issue where first-time sellers would see a blank payout summary instead of their current payout information.

Fixed a race condition where concurrent payout deletions could cause errors when viewing seller payout summaries.

Fixed validation errors when creating or updating reference and file attributes, ensuring proper error codes are returned.

Fixed an issue where password-set emails were not being routed correctly based on user roles (customer, seller, staff).

Fixed several data loading issues that could cause incorrect results in certain edge cases.

Task priority support for background jobs​

Added support for task prioritization in background job processing, ensuring critical tasks are processed first and preventing worker starvation during high-load periods.

Performance improvements​

Made several performance improvements across the platform:

• Search index rebuilds now run daily instead of multiple times per day, reducing system load.
• Background job processing is now more efficient, preventing task queue buildup and burst-flooding.
• Improved thumbnail generation with automatic task expiration to prevent queue buildup.

Completed migration to modern GraphQL framework​

Completed the migration to Strawberry GraphQL, the modern Python GraphQL framework. This internal upgrade improves API maintainability and performance while maintaining full backward compatibility with existing integrations.

Fixed multiple issues with GraphQL filter handling that could cause crashes or incorrect results when using null filter values across all API modules.

Fixed an issue where clearing all selections in a multi-select attribute field would not properly clear the attribute values.

Fixed an issue where certain media and content query fields could cause combined errors when data was unavailable.

Added improved error handling and logging for Redis cache operations to prevent silent failures.

Fixed an issue where certain Stripe errors could cause payment webhook processing to crash unexpectedly.

Improved error handling when attempting to allocate more stock than is available, providing clearer feedback.

Fixed memory issues that could cause webhook processing to crash under high load by reducing pressure on the message broker.

Improved task processing stability by preventing sudden bursts of background jobs that could overwhelm the system.

Added better error handling and logging for cache-related operations to improve system reliability.

Fixed issues with product thumbnail generation, ensuring thumbnails are created on-demand when requested and resolving problems with the thumbnail regeneration command.

Fixed an issue where thumbnail generation tasks could be lost when workers were restarted, ensuring all product images are reliably processed.