Hi Gerard,
I think the core issue here is a misunderstanding of how conversational context works in Narrova (and AI systems generally), so let me clarify that directly.
1. Conversations do not share memory or context
Each conversation thread is its own isolated context. There is no background memory that allows Narrova to “remember” decisions, reasoning, or story discussions from a different conversation unless that information is explicitly reintroduced.
So requests like:
“Can you please recover our current status of the storyform we were working on?”
will not work if that information is not present in the current conversation’s context. Narrova cannot infer or reconstruct prior reasoning unless it is provided again via:
- Uploaded files
- Exported conversations
- Explicit story or storyform context created inside the thread
This is documented here:
https://platform.dramatica.com/docs
2. What you’re seeing with the old thread
From what you describe, the thread itself still exists, but it no longer behaves correctly. This usually means it was created under an older system configuration and is no longer compatible with the current version of Narrova.
That explains why:
- The thread still appears in your list
- You can open it
- But new replies disappear or fail to persist
This is a known symptom of a deprecated or unstable conversation state, not something you caused.
3. Why you can’t simply “pick up where you left off”
Even if the thread still exists somewhere, Narrova can’t reliably resume work from a conversation that is months old and no longer functioning. The system has no safe way to re-establish that earlier conversational context in a live, up-to-date session.
So while it may feel like “everything is still there,” the practical reality is that the system cannot re-enter that state as-is. In those cases, the correct approach is not recovery inside the old thread, but starting fresh with context reintroduced intentionally.
4. The recommended way forward (best practice)
Since the discussion and reasoning matter more than the final storyform, the recommended workflow is:
From there, Narrova can continue cleanly and reliably.
This is the same approach that worked for you previously when a conversation became stuck.
5. Going forward
For long-running projects, conversations should be treated as working sessions, not permanent memory. Periodic exports or saved notes are helpful—especially when work spans weeks or months, and development of the platform has necessitated a new conversation.
If you’d like help setting up a clean new conversation or structuring Storyform context efficiently, support is happy to help. But resuming directly inside a non-functional, older thread isn’t something the system can reliably do.
I strongly recommend reviewing the documentation above, as it explains these workflows and constraints in more detail.