# Removing the Opt-Out Converts a Deprecation Into a Deadline ## Claim A breaking change shipped without a feature flag or opt-out forces every dependent to migrate on the platform's schedule rather than its own, which turns what would otherwise be a gradual deprecation into a hard deadline shared by the entire ecosystem at once. <!-- atomicity: 1 idea ✓ --> ## Scope & conditions - Holds for platform-level changes affecting many independent downstream consumers (plugins, themes, integrations) who cannot coordinate their own migration timing. - The absence of an opt-out is what does the work here — a flagged, opt-in change with a announced future removal date still gives consumers a defer-and-test window this pattern removes. - Breaks where the platform provides an equivalent escape hatch through other means (a compatibility shim, a grace-period build target) even without a literal flag. - Does not by itself judge whether removing the opt-out was the right call — a hard deadline can be justified (security fix, architectural necessity) even though it concentrates migration cost onto one date for everyone. ## Evidence WordPress 7.1's post-editor canvas becomes an iframe with no flag to check and no opt-out, landing 19 August. Any block referencing the global `window` or `document` breaks on that date rather than on a schedule the plugin author controls. Named fixes exist (`ownerDocument`, `defaultView`, `useRefEffect`, `editorStyle` in `block.json`), but for a block depending on an unmaintained third-party library, the only path is `patch-package` — adopting the dependency's maintenance burden indefinitely because there was no window to coordinate a replacement. > [!quote] > If your custom blocks touch the global window or document, WordPress 7.1 will break them. ## Links - [[3. Resources/Permanent Notes/A Published Deprecation Schedule Makes a Fast-Moving Dependency Safe to Adopt]] — because contrasts; that note treats a published schedule as what makes adoption safe, this note describes what happens when that schedule collapses to a single date with no lead-time mechanism. - [[3. Resources/Permanent Notes/A Reversible Opt-Out Beats a Revert When Integration Runs Deep]] — because contrasts; that note argues for building the opt-out in the first place, this note describes the downstream cost when one isn't provided. - [[3. Resources/MOCs/Open Source Governance MOC]] — because applies-in; ecosystem-wide breaking-change governance. - [[WordPress Platform Design MOC]] — what closing an escape hatch does to everyone building on the surface - [[3. Resources/Permanent Notes/A Legacy Feature Should Become Obsolete by Choice Not by Force]] — because extends: removing discovery performs the same conversion one step earlier, turning a migration into a deadline users never agreed to. ## Tests / falsifiers - Measure the actual rate of directory-plugin breakage on 19 August against a comparable past WordPress release that shipped an equivalent change behind a flag — if breakage rates are similar, the opt-out's absence isn't the operative variable. - Find a case where a no-opt-out breaking change caused measurably less disruption than a flagged one, because the flagged version's grace period was simply ignored by most consumers until the deadline anyway — if common, "opt-out" matters less than "deadline visibility." ## Maintenance - Last tended: 2026-08-16 - Next review: 2026-11-16 ## Source - [[3. Resources/Literature Notes/WordPress 7.1 is afoot, Gutenberg 23.6, iframed post editor, Abilities API, Studio Code and more - Weekend Edition 371]] — reports the 19 August iframed-canvas deadline with no flag and no opt-out.