What an audience sees in a live programme is a composed experience made from the speakers, the presentation and the layout chosen for that moment. If live delivery, the recording and later replay each assemble their own version, they can quietly drift apart. cRUD Morph keeps one Program Output and lets every consumer use it.
TL;DR
- Program Output is the composed programme of a media session, with an identity and timeline owned by cRUD Morph.
- Live delivery, recording and replay are consumers of that output, not separate compositions.
- Lineage is recorded from normalised provider receipts, and a provider job is never treated as complete without one.
- Slides stay protected presentation assets plus synchronised presentation state, and captions, polls and chat keep their own timelines. Program Output composes the experience without flattening them into video.
Three copies of the same moment
When the live stream, the recording and the replay are produced by different paths, one composes the live layout, another records whatever reached it, and a third packages a file for later viewing. Each path makes its own decisions, so the replay can show a different layout from the live session, or the recording can miss a change that viewers saw.
When that happens, nobody can say with confidence which version is the programme.
One output with one identity
In cRUD Morph, the Program Output domain owns the output’s identity and timeline. There is one Program Output for each media session, with immutable versions and a single authority over what the programme is at any point.
Having one identity matters beyond tidiness. It gives every later step something exact to refer to. The cRUD Morph workflow shows where Program Output sits between delivery, recording and reuse. A recording can say which output it captured, and a replay can say which recording it came from.
Consumers, not copies
Live delivery, recording and packaged playback consume the same Program Output through their own bindings. The recording is taken from that output rather than composed a second time, and it does not depend on any particular live-delivery path.
Replay follows the same discipline. It is packaged from the finalised source of a ready recording, so what people watch later traces back to the programme that was produced. On-demand delivery of a verified uploaded source is bound to the output in the same way.
Lineage from receipts, not assumptions
Composition, recording and egress run on provider infrastructure, and media moves from provider to provider. cRUD Morph records lineage from normalised provider receipts and never treats a provider job as complete without one.
Where receipts are missing, the gap is reconciled rather than filled with an assumption. Partial failures do not silently mark work complete, and public state can report requested, accepted, processing, succeeded, failed, cancelled or unknown, so a product never has to claim more than the evidence shows.
Keeping structured elements separate
Program Output composes the authoritative experience, including where the presentation appears and which slide is current, but it does not turn a presentation into continuous video. A slide remains a protected presentation asset plus synchronised presentation state, and captions, polls, chat, Q&A and reactions each keep their own synchronised timeline, so a gap or a delay in one does not block the others. The output refers to that state rather than flattening it into frames, and the product keeps ownership of its interaction content.
What this gives a product team
- One reference for the programme across live, recorded and replayed experiences.
- A traceable path from the session to every recording and replay derived from it.
- Honest status when a step is still in progress or its outcome is unknown.
- Interaction content that stays with the product rather than inside a video file.
Questions to ask
- Is the recording the same programme the live audience saw, or a separate composition?
- Can a replay be traced back to the session and recording it came from?
- How does the system know a recording job actually finished?
- What does the product see when an outcome is not yet known?
A session from live programme to replay
Imagine a session that opens with the host and one speaker, moves to a presentation, returns to a two-person discussion and closes with audience questions. The live audience should see each approved change in that order. The recording should preserve the same programme, and the replay should be prepared from that recording without inventing a fourth version of the session.
- Authorise the session
The customer product defines the session, roles, permitted sources and required outputs.
- Create one programme timeline
cRUD Morph records each approved composition change against one Program Output identity.
- Bind live delivery and recording
Both consumers use the same output while keeping their own delivery and readiness state.
- Prepare replay from verified work
Replay begins only after the recording source is ready and its lineage can be traced to the session.
A failure does not erase this relationship. If recording is still processing, replay remains unavailable rather than pretending to be ready. If a provider receipt is missing, the state remains unknown and is reconciled. If a destination fails, the failure belongs to that destination without changing the identity of the programme seen elsewhere.
A decision checklist for one programme
- Can every output be traced to one session and Program Output identity?
- Does the recording preserve the approved programme rather than assemble a new one?
- Can the product distinguish processing, ready, failed and unknown states?
- Do slides, captions and interactions keep their useful structure outside the video?
- Can one consumer fail without making every other output untrustworthy?
Summary
One Program Output keeps live delivery, recording and replay telling the same story. In cRUD Morph, the output has one identity and timeline, every consumer binds to it, lineage comes from receipts rather than assumptions and structured interaction stays on its own timelines.
About this article. It explains how cRUD Morph works. It makes no performance, quality, availability or cost claim. Those results are measured for each engagement against a matched baseline.