WordPress 5.8 and the Long Road to Full Site Editing

Watch a platform learn from its own history, because it is happening in public, and the lesson applies elsewhere. WordPress 5.8, released this month, advances the full-site-editing project. The block paradigm extends from post content to the entire site: headers, footers, templates, everything. It moves in three deliberate steps. Blocks now power widget areas, the sidebars and footers that stayed classic through the Gutenberg era. theme.json arrives as a configuration layer for design control. And template editing previews in a limited, opt-in form.

What is notable is as much the cadence as the content. The 2018 launch (the compressed endgame, the earned objections, the Classic Editor install graph) taught the project something visible. 5.8 shows the learning. Capabilities ship in absorbable increments. Opt-in comes before default. Escape hatches publish alongside changes. The full-site-editing future arrives as a road rather than a bang.

We take the month’s real headline to be the quiet one. So this post leads with theme.json, carries the compatibility notes the release actually requires, and ends with the timeline honesty the FSE discourse currently lacks.

theme.json: design constraints as configuration

The quiet headline, unpacked for the design-governance audience that has been waiting for it. theme.json is a single file in which a theme declares its design system. The color palette. The typography scale. The spacing units. Which options editors may adjust and which are locked, per block type if desired.

Declared once, the constraints spread everywhere. The editor’s controls show the palette, not a color picker’s infinity. The front end renders from the same tokens. The drift between what designers governed and what editors could actually do closes structurally rather than by process.

Readers of our design-systems coverage will recognize exactly what this is. The token layer, arriving natively in the platform that runs the plurality of the web. Decisions made once and distributed, expressed as a configuration file rather than a governance meeting.

For the multi-site organizations we serve, the implications are the systems business case, arriving in WordPress. Brand governance as supply rather than review. The fifty-department site constrained at the source. The handoff from design system to build, once documentation and hope, literally becomes a file. The practical advice: theme.json adoption is worth it now for maintained custom themes. Incremental adoption is supported, so start with the palette and typography layers. It belongs in every new-build spec we write from this month. Agencies still shipping themes governed by CSS conventions alone are shipping the previous decade.

The compatibility notes, which are this release’s actual work

• Block widgets carry the surface: the widget-area conversion touches every site with classic widgets, which is most of them. The breakage lives where it always lives: the custom widgets, the theme’s widget-area assumptions, the plugins that inject into sidebars. The escape hatch is official and honorable. The Classic Widgets plugin restores the old screen indefinitely, and staging-first testing is the usual advice, unchanged.

• Template editing is a preview, not a default: opt-in, theme-dependent, and labeled as such. The advice is the beta practice we apply everywhere. Evaluate on staging, deploy nothing, file the findings.

• WebP lands natively: February’s image-format project just got easier for WordPress sites. Core now accepts and serves WebP uploads, which moves the format work from plugin-bridged to platform-supported. The CDN-negotiation rung remains the bigger win, and the two compose cleanly.

• IE11 support ends with little fanfare: 5.8 drops Internet Explorer 11 support, a milestone worth noting after a decade of browser-support cleanup. Sites with residual IE audiences should note the platform’s decision as the planning input it is.

FSE timeline honesty, for planning purposes

The expectation-setting the discourse needs, stated plainly for the client-planning meetings we sit in. Full site editing is a multi-year road, not a next-release event. Block themes remain pioneer territory. The theme architecture FSE requires is young, the ecosystem is thin, and the pioneering sites should be the ones with pioneering reasons.

Existing themes face no deadline. Classic themes are supported indefinitely on every current signal, and the 5.0-era lesson about announced horizons versus panic applies. ‘Wait for maturity’ is a respectable choice we are actively advising for most of its book. Adopt theme.json now, since it works in classic themes and pays immediately. Watch the FSE releases with staging-tier curiosity. Let the pioneer cohort sand the edges.

This is the maturity model we celebrated with Drupal 9. Steady, absorbable change over big-bang trauma. It is arriving in WordPress’s culture, visibly, because of its own 2018 experience. Platforms that learn from their own transitions show exactly the stewardship the durability rubric weights heaviest. The road is long and the pace is honest. The destination (the fully block-built, token-governed site) is component thinking arriving at the web’s largest install base, at the speed that base can absorb.

Where design systems meet platform maturity

The release rewards zooming out. Two long stories intersect in this month’s changelog. The design-systems story: tokens, constraints, governance as supply. And the platform-transition story: cadence, compatibility, the physics of moving a third of the web. 5.8 is the month they merge. The governance mechanism we have wanted since the customizer era is now shipping through the transition practice the platform learned from its hardest release.

For the practice, the takeaways are concrete and already in our templates. theme.json in every spec. The widget-compatibility pass in every 5.8 upgrade. The FSE watch on the annual review calendar rather than the quarterly panic one.

The post ends with the observation the whole month supports. The platform is learning in public, at scale, from its own recorded history, which is the same practice we sell to readers and apply to ourselves. Steady change, well managed, escape hatches published. The platform transition playbook, working as written. We judged the 2018 release harshly where it earned it, and we are glad to call this one the correction it visibly is.