Amy Kamala has issued a call for testing for web hosts and database administrators to test the Real Time Collaboration feature planned for WordPress 7.0, with feedback needed by May 4.
Initially, WordPress 7.0 was set to be released during the Contributor Day at WordCamp Asia 2026, but was postponed to further refine the Real Time Collaboration feature. The new release date for WordPress 7.0 as of now is May 20, 2026.
Why broader testing is being requested
Amy Kamala said additional testing is required within a limited timeframe to help guide architectural decisions for the real-time collaboration feature in WordPress 7.0.
She highlighted that the feature has already undergone testing, but only across a limited range of environments, making broader participation necessary, “ While thorough testing has been done, it’s been done in a limited set of environments, and it’s crucial to have data covering as wide a spread as possible of WordPress environments to set the feature up for success.”
Kamala said the feature is not limited to large sites with multiple users working on content creation, and described it as “ It’s a big step forward in the phase 3 roadmap that opens the door to many new possibilities, some of which aren’t known.”
She also stressed that contributing to testing at this stage helps build a strong foundation for what comes next, adding that upcoming possibilities include “agentic collaborators” as well as future phase 3 features such as “suggestion mode.”
How hosts can test and submit results
A dedicated “Distributed Real-time Collaboration Performance Tests” suite has been created that hosting providers can run and submit the results back to WordPress.org. The scripts require only a small set of environment variables and depend on four tools: bash, cURL, WP-CLI, and patch.
The small number of environment variables to be configured are the path to the WordPress installation (WP_PATH), a label describing the hosting setup (ENVIRONMENT_NAME), and reporting credentials (REPORTER_API_KEY).
The ENVIRONMENT_NAME field is intended to group results across similar hosting types, with examples including shared hosting tiers, VPS, and managed eCommerce environments. The guidance specifies that version-level details, such as PHP or database configurations, should not be included in this label, as they are collected separately.
Hosting contributors are advised not to include technical details such as PHP or database versions in this field, as that data is collected separately. The suite measures the performance of four different data storage approaches for the real-time collaboration feature.
Kamala also highlighted that the most relevant results will come from setups that reflect actual customer environments, encouraging hosts to test configurations that mirror real-world usage rather than default or newly provisioned installs, “ The most useful data you can contribute is data that represents your actual user base, not your new-install setup. If most of your customers are on PHP 7.4 with no object cache on shared infrastructure, that’s the configuration needed to see results from.”
The post notes that the primary method for participating is through the test runner, which is designed to automatically submit results, “ The simplest and best way to participate is by using the test runner to submit test results to WordPress.org (specifically make.wordpress.org/hosting), which the script will attempt automatically when the necessary environment variables are configured correctly.”
A WordPress.org bot account is also needed, and detailed steps on how to submit results are given on GitHub.
How test data will be handled
Kamala also clarified that results from this testing effort will not be published individually, differing from how PHPUnit hosting test data has been shared in the past, “ Unlike the PHPUnit Hosting Tests, the results submitted through this repository will not be shown publicly on a per report basis.”
The decision is linked to the temporary need for the data ahead of WordPress 7.0 and the limited time available to publish it, “ This route was chosen because of the ephemeral nature of the need for this data (to make an informed decision leading up to 7.0), and because of the very short timeframe to publish this (building a front end interface would have added time).”
The results will be combined and reviewed without linking specific data to individual contributors, with analysis also taking into account differences in hosting environments and server setups.