WordPress co-founder Matt Mullenweg has called for a delay to WordPress 7.0, putting an end to plans for a live release during Contributor Day at WordCamp Asia 2026 next week.
Mullenweg posted his decision in the #core-committers channel in WordPress Slack yesterday, citing the scope and status of the release — specifically, real-time collaboration (RTC) and its underlying database architecture — as the reason for the pause.
“Given the scope and status of 7.0, I think we should go back to beta releases, get the new tables right, lock in everything we want for 7.0, and then start RCs again,” he wrote. “Date-driven is still our default, but for this milestone release we want to target extreme stability and exciting updates, especially as AI-accelerated development is increasing people’s expectations for software.”
He said the delay was a one-off, and indicated he wanted WordPress to return to a scheduled cadence — targeting four releases a year from 2027, supported by faster, AI-enabled development.
The delay comes after core committers shared their frank concerns about the release’s readiness ahead of the RC1 release party last week. Mullenweg, posting in #core after the release party, acknowledged their concerns: “I’d rather have a really stable 7.0 than hit the WC Asia date, so let’s make sure this one is really rock-solid. We’ve got a lot of big changes.”
In the #7-0-release-leads channel in WordPress Slack yesterday, release lead Matias Ventura said he had spoken with Mullenweg about the delay and put the likely extension at around 3-4 weeks. He said a final timeline would only be set once the table design had been resolved and reviewed by Mullenweg.
Ventura said the immediate priority was working through the design of a new database table for real-time collaboration. The table had been proposed last month and was acknowledged during the RC period as the preferred long-term solution to RTC’s cache invalidation problems, but was deferred due to unresolved questions about its specific purpose, schema, naming, and fields.
Ventura said there was also an additional consideration to work through — whether the table should be designed to support use cases beyond real-time collaboration, including sync and Playground, to ensure the project was building the right primitives from the start.
“The release is in good shape but there’s been enough feedback from contributors that the post meta mechanism was ill-conceived, had some tradeoffs, and some amount of consensus that a table is the long-term solution we’d want for the project,” Ventura wrote.
He also noted that the timing could work in the project’s favor, suggesting the shift could be a good opportunity for extra testing on Contributor Day at WordCamp Asia.
Bluehost-sponsored core committer Jonathan Desrosiers moved quickly to stop the clock on active development, asking fellow committers in #core to hold off on committing anything unrelated to 7.0 in trunk until the implications of the delay were clearer.
He also asked Ventura to clarify what, specifically, was now considered unstable, and why the RTC fix that shipped in RC1 — a direct-database-query approach from WordPress VIP’s Chris Zarate that sidesteps the cache-clearing problem without touching core APIs — was not acceptable for 7.0.
“What we’re aiming at is unclear,” Desrosiers wrote.
Independent core committer Aaron Jorbin also raised concerns about the proposed timeline and the availability of contributors in the coming weeks.
“I would also be interested in understanding why 3-4 weeks is viewed as the amount of time necessary when there is no clear consensus on the DB table, the RC period is normally 3-4 weeks and we are entering an immediate period where a large number of folks have time off for religious holidays and WCAsia travel,” Jorbin wrote.
Ventura confirmed that while RTC was the primary reason for the extension, other features punted ahead of RC1, including client-side media processing, could be reconsidered if contributors had confidence in their readiness. “Client-side media is one long overdue in my opinion,” he wrote.
Meanwhile, the versioning question created its own headaches. Mullenweg called for a return to beta releases, but two release candidates have already shipped — RC1 on March 24, RC2 on March 26 — and reverting to beta versioning now is not straightforward.
PHP’s version_compare() function treats beta versions as lower than RC versions, meaning update mechanisms, Composer packages, and the WordPress Beta Tester plugin could behave unexpectedly, or break outright, if a beta were published now. Several contributors argued on Slack that the safest path now is to continue with RC numbering regardless of how the releases are treated internally — more RCs, as many as needed, rather than a versioning experiment that could confuse the update server and the broader ecosystem. Ventura acknowledged the question needed to be resolved quickly.
The delay is a blow to WordCamp Asia 2026, which kicks off in Mumbai, India, next Thursday. The WordPress 7.0 release had been planned to happen during Contributor Day, allowing contributors in the region to take part in the release party in real time. It would have been the first major release to ship during a WordCamp, following the model of WordPress 6.9’s live release at the State of the Word in December.
In a Q&A with The Repository in February, WordCamp Asia co-lead organizer Aditya Kane described the prospect of pressing a button in a hall full of contributors and watching WordPress 7.0 reach millions of websites within hours as “a bit like watching a rocket launch,” something “romantic.” That moment will not happen next week.
Ventura is expected to publish an announcement on WordPress.org later today about the delay.