15 — Removed / not shipped

This chapter is the tombstone register. It exists so v1 migrators and readers cross-checking against blog posts can find out, in one place, what's gone, what was proposed but not shipped, and what to use instead.

If a row says "(proposed earlier)," that means the surface was named in an early draft of the v2 spec and dropped before ship. The replacement is canonical; the proposed name never existed at run-time.

Removed / not shipped	What to do instead	Reference
dispatch-with (master)	(dispatch event {:fx-overrides {...}})	MIGRATION M-4
dispatch-sync-with (master)	(dispatch-sync event {:fx-overrides {...}})	MIGRATION M-4
dispatch-to (proposed earlier)	(dispatch event {:frame :todo})	002
subscribe-to (proposed earlier)	(subscribe query-v {:frame :todo})	002
frame-dispatcher / bound-dispatcher / bound-subscriber (proposed earlier)	(rf/frame-handle) — the keystone operation bundle; captures the frame at creation time and is safe to call during render and from async callbacks.	002
dispatcher	(:dispatch (rf/frame-handle)) or the dispatch injected in a reg-view body.	002
subscriber	(:subscribe (rf/frame-handle)) or the subscribe injected in a reg-view body.	002
bound-fn (CLJS macro)	(rf/frame-handle) — the keystone operation bundle. (bound-fn was first renamed frame-bound-fn, then EP-0024 Open Issue #8 retiered the frame-bound-fn / frame-bound-fn* closures to internal — frame-handle, or an explicit {:frame …} opt, expresses the real use cases.)	002
current-frame	Renamed to current-frame-id — returns a frame-id keyword.	002
get-frame-db	Renamed to app-db-value — returns the app-db VALUE (a plain map), nil for an unregistered / destroyed frame. The container accessor is the internal re-frame.frame/app-db-container.	002
enable-performance-api-tracing! (proposed earlier)	Performance-API instrumentation is gated on the compile-time re-frame.performance/enabled? goog-define, not a runtime toggle. See 11 — Instrumentation.	009
add-trace-listener / remove-trace-listener (proposed earlier)	register-listener! / unregister-listener!	009
register-trace-listener / unregister-trace-listener (no-bang, proposed earlier)	Renamed to register-listener! / unregister-listener! — the bang form matches the side-effecting nature of listener registration.	009
Bare [:my-view "args"] keyword-tagged hiccup	Var form [my-view "args"] (canonical) or [(rf/view :my-view) "args"] for late-binding by id.	004
h macro (proposed earlier)	Removed. Use the Var form [my-view "args"] or [(rf/view :my-view) "args"].	004
reg-global-interceptor	reg-frame :interceptors — frame-level is the canonical "global within this frame." For cross-frame observation, use register-listener!.	MIGRATION M-17
clear-global-interceptor	No replacement needed — re-register reg-frame with an updated :interceptors vector (absent-key semantics clear it).	MIGRATION M-17
reg-sub-raw	reg-sub for app-db reads; Pattern-AsyncEffect for non-app-db sources; state machines for lifecycle; the 006 adapter contract for bridging external reactivity.	MIGRATION M-18
re-frame.alpha/reg	The shipped per-kind registrars: reg-event / reg-sub / reg-fx / reg-cofx / reg-flow. (The v1 event trio reg-event-db / reg-event-fx / reg-event-ctx is not a v2 target — those are removed/withdrawn throwing stubs and migration inputs only, see the rows below and EP-0018; reg-event is the single event-registration form.) The re-frame.alpha namespace is dissolved.	MIGRATION M-23
re-frame.alpha/sub	Vector-form (rf/subscribe [::id arg]).	MIGRATION M-23
re-frame.alpha/reg-sub-lifecycle and built-in lifecycle policies (:safe, :no-cache, :reactive, :forever)	Sub-cache uses a single algorithm — synchronous ref-counting (dispose on derefer-count → 0). For specific edge cases, file a follow-up bead.	MIGRATION M-23
debug interceptor	Trace surface (Spec 009) + 10x / re-frame-pair	MIGRATION M-21
trim-v interceptor	Canonical map-payload call shape	MIGRATION M-21
on-changes interceptor	Flows (Spec 013)	MIGRATION M-21
enrich interceptor	Flows (derived state) / :schema (validation) / custom ->interceptor (escape hatch)	MIGRATION M-21
after interceptor	Registered fx (:fx [[:my-fx ...]]) for side-effects; custom ->interceptor for context-shaped work; vendor from v1 if the helper is wanted as a local utility	MIGRATION M-21
inject-cofx / inject-cofx* (coeffect-delivery interceptors)	Declare the facts a handler needs with :rf.cofx/requires registration metadata; the runtime supplies them. Register the supplier with value-returning reg-cofx. There is no interceptor that delivers a coeffect — see 01 — Core §reg-cofx and Guide — Effects and coeffects.	EP-0017
reg-event positional interceptor vector middle slot	Put the chain in registration metadata: (reg-event :id {:interceptors [i1 i2]} handler). If a call also has metadata, merge :interceptors into that map.	MIGRATION M-70
reg-event-db	Use reg-event (no alias). Destructure :db from the coeffects map and wrap the return in {:db …}: (reg-event id (fn [{:keys [db]} ev] {:db BODY})). A stale call raises the always-on hard error :rf.error/reg-event-db-removed naming reg-event.	EP-0018 / MIGRATION M-73
reg-event-fx	Use reg-event (no alias) — the identical shape under the bare name (coeffects in, effects out); just rename the call. A stale call raises :rf.error/reg-event-fx-removed.	EP-0018 / MIGRATION M-73
reg-event-ctx	Demoted to a framework-internal primitive. Express application full-context work as a registered interceptor (reg-interceptor with :before / :after, referenced by id from a reg-event registration's :interceptors chain — EP-0022). A stale public call raises :rf.error/reg-event-ctx-removed naming reg-interceptor.	EP-0018 / MIGRATION M-73
with-overrides (v1 macro name)	Renamed to with-fx-overrides.	MIGRATION M-50
add-marks / set-marks (imperative app-db path-marks)	Declare durable app-db classification on the frame: reg-frame :sensitive / :large {:app-db [...]} path maps. The underlying re-frame.marks/* fns remain internal/test helpers only.	EP-0015 / 015
redact-interceptor (positional payload-scrub interceptor)	Classify transient payloads on the registration: reg-event-* :sensitive / :large metadata. Projection happens centrally at egress, not via interceptor placement.	EP-0015 §7 / 015
Schema-attached :sensitive? / :large? as the app-db classification route	app-db classification is frame-owned (above). Per-slot :sensitive? / :large? Malli props remain the route for owner-local schema'd data — machine :data, resource data/params, HTTP :decode bodies.	EP-0015 §8 / 015
:rf.privacy/show-sensitive? / set-show-sensitive! (process-global on/off privacy toggle)	There is no process-global toggle. On-box visibility is a named :rf.egress/* profile per (tool, frame) — :rf.egress/local-redacted (default, fail-closed) / :rf.egress/local-raw (trusted-local opt-in).	EP-0015 issue 7 / 015

Why these went

A few one-line rationales for the larger cuts:

• The frame-affordance surface (dispatcher, subscriber, bound-fn, and — under EP-0024 — the frame-bound-fn / frame-bound-fn* closures) — the multi-frame surface had accreted ~8 affordances that were mostly patches around one root cause: ambient frame lookup doesn't survive async boundaries. The redesign collapses them so you choose by INTENT, not mechanism — hold a frame's ops as a value via frame-handle (the one public carry primitive), scope with with-frame / with-new-frame / frame-provider, override per-call with {:frame …}. dispatcher / subscriber folded into (:dispatch (rf/frame-handle)) / (:subscribe (rf/frame-handle)); bound-fn was renamed frame-bound-fn, and EP-0024 Open Issue #8 then retiered frame-bound-fn / frame-bound-fn* to internal (:tier :implementation) since frame-handle expresses the real use cases. Full design: Spec 002 §The multi-frame surface.
• reg-global-interceptor — the v1 surface conflated "global" (process-wide) and "global within this frame" (frame-wide). The frame-wide answer is reg-frame :interceptors; the process-wide answer is register-listener! (observation, not mutation). Splitting the two ended the confusion.
• reg-sub-raw — the v1 escape hatch covered four distinct use cases (app-db reads, async sources, lifecycle, external reactivity). Each now has its own surface — reg-sub, Pattern-AsyncEffect, state machines, the adapter contract — and the escape hatch isn't necessary.
• re-frame.alpha — the alpha namespace was an experiment to unify registration / dispatch under generic reg / sub / dis verbs. The unification didn't pay for itself — the per-kind macros read better at the call site and survive better in the linter — and the alpha namespace is dissolved. No APIs in this reference live outside re-frame.core (with the documented per-namespace exceptions).
• The five v1 interceptors (debug, trim-v, on-changes, enrich, after) — each was either trivially ->interceptor-able (so didn't earn its own surface), or replaced by a richer mechanism (flows, schemas, the trace surface).
• inject-cofx / inject-cofx* — EP-0017 made coeffect delivery a declared contract, not an interceptor side-channel. A handler lists the world facts it consumes in :rf.cofx/requires, suppliers are value-returning reg-cofx fns, and the runtime grades/records facts by id. Injecting a coeffect through the interceptor vector hid the dependency from handler-meta, the cofx graph, and the recordable-fact grading, so the interceptor form was removed. Full model: Guide — Effects and coeffects.
• The pre-EP-0015 privacy surfaces (add-marks / set-marks / redact-interceptor / schema-attached app-db marks / the process-global :rf.privacy/show-sensitive? toggle) — five overlapping ways to express one fact. EP-0015 collapsed them to one model: classification is declarative and local to the owner (frame config for durable app-db; per-slot schema props for owner-local schema'd data; registration metadata for transient payloads), projection is centralized at trust boundaries (project-egress + the closed :rf.egress/* profile enum), and sinks consume already-projected records. The full teaching is Guide ch.23.

See also

• MIGRATION.md — the AI-driven migration spec, with one rule per row above.
• 01 — Core — the surfaces that replaced the removed ones.