“You Didn’t Open”: Clarifying What Did Not Happen in Interactive Fiction
Separate “You didn’t open the cabinet door” into the requested action, the actual attempt, and the resulting state. Then identify exactly what did not happen. A negative-feedback revision card helps distinguish three cases: no contact, a pull that did not open the door, and a door that was already open.
Introduction
To revise “You didn’t open the cabinet door,” first establish how far the character actually got, then describe the door’s current state. No contact, a pull that did not open the door, and a door that was already open call for three different responses. Simply swapping “didn’t” for “couldn’t” will not remove the ambiguity. Separate the attempt from the result before deciding what to negate.
Distinguish the request, the attempt, and the change of state
When a player enters “Open the cabinet door,” they are making a request. Whether the character reaches out or pulls the handle concerns the actual attempt. Whether the door changes from closed to open concerns the result. These three may align, or the action may stop at different stages. Feedback must not treat the player’s input as an action that has already happened.
“You didn’t open the cabinet door” negates the event as a whole without explaining what happened along the way. A reader might understand that the character did nothing, or that they tried and failed. If the door was already open, the same sentence can misleadingly suggest that it is still closed.
There is also a difference between “It is open now” and “You opened it this time.” The former describes a state; the latter says that this action caused a change. When the cabinet door was already open, the target state already holds. That does not justify writing “You successfully opened the cabinet door” and crediting this action with a preexisting state.
This article demonstrates revisions through three independent scenarios involving the same cabinet door. The characters, conditions, inputs, and responses are original fictional teaching examples. They are not real user cases, measured results, or records of hands-on testing, and they do not represent DramaFork product capabilities.
Use five fields to identify what is being negated
You can copy this negative-feedback revision card directly. Fill in the facts before composing the sentence; do not start by choosing a negative expression that sounds smooth.
| Field | What to enter |
|---|---|
| Requested action | What the player asks the character to do, and to which object |
| Actual attempt | The actions that definitely happened this time; explicitly enter “None” if there were none |
| Resulting state | The object’s state before and after, and whether it changed |
| What is negated | Whether the attempt never happened, the expected change did not occur, or the action needs no repetition |
| Accepted follow-up input | A specific input currently accepted, without promising an unknown outcome |
Under “Actual attempt,” record only established facts. If the record says only that the character touched the handle, you cannot add “You pulled hard several times.” Nor should you fill in “Resulting state” by reasoning backward from the outcome: a door remaining closed does not mean it is locked, much less that a key is hidden nearby.
The final field helps the author check whether the response provides a valid action. It does not require every failure to add a clue or open another route. If the work provides no follow-up action, mark it “To be confirmed.” Resolve the gap in the rules before writing the prompt; do not invent an available command just to complete the card.
Fill out the card for three cabinet-door scenarios
For this example, all three independent scenarios accept “Open the cabinet door.” In Scenario A, the character is too far from the cabinet, and this input does not automatically move them closer. In Scenario B, the character has actually pulled the handle, but the door remains closed for a reason not yet established. In Scenario C, the door was already open, and repeating the request does not perform the opening action again. The follow-up inputs below also belong solely to this example’s setup.
| Field | A: No contact | B: Pulled, but not opened | C: Already open |
|---|---|---|---|
| Requested action | Open the cabinet door | Open the cabinet door | Open the cabinet door |
| Actual attempt | None; the character neither approached nor reached out | Grasped the handle and pulled outward once | None; did not repeat the opening action |
| Resulting state | Closed before; still closed now | Closed before; still closed after the pull | Open before; still open now |
| What is negated | The attempt to open the door did not begin | The change from closed to open did not occur | Opening need not be repeated; the existing open state is not being negated |
| Accepted follow-up input | Approach the cabinet | Examine the cabinet door | Look inside the cabinet |
A possible response for A is: “You are still far from the cabinet and have not touched the handle; the cabinet door remains closed. You can enter ‘Approach the cabinet.’” The first clause places the stopping point before contact, and the next states the object’s condition. Do not write “You can’t reach the handle” here: the example establishes only that the character is far away, not that they have already reached toward it.
A possible response for B is: “You grasped the handle and pulled outward once, but the cabinet door did not open. It remains closed. You can enter ‘Examine the cabinet door.’” Acknowledge the attempt first, then negate the result. The response does not claim that the door is locked or promise that examining it will reveal a solution. The available evidence establishes only that the door did not open.
A possible response for C is: “The cabinet door was already open, so there is no need to open it again. You can enter ‘Look inside the cabinet.’” There is no need to force a negative sentence into the response. Confirming that the target state already holds and then providing the next step allowed in this example is more accurate than saying “Opening failed.”
These three responses need not be the same length. A must explain why the action did not begin; B must retain the pull that actually happened; C must correct the premise that the door needs opening again. A uniform sentence pattern that removes these differences would reintroduce confusion.
“Couldn’t” can suggest failure, but cannot establish an attempt
Here, “You couldn’t open the cabinet door” more readily suggests an obstacle or lack of success than “You didn’t open the cabinet door.” It still does not specify what the character did. The word “couldn’t” alone cannot be expected to tell the reader that the character pulled the handle.
If the attempt matters later—for example, if the character subsequently says “I pulled it earlier”—write it out: “You pulled the handle, but couldn’t pull the cabinet door open.” If “pulled once” is already explicit, “did not open” also works in the second half. What matters is which facts the complete sentence preserves, not assigning the two expressions rigid, mutually exclusive uses.
Similarly, “It hasn’t opened yet” can create an expectation that the action will be completed, while “It won’t open” can suggest that opening it is currently impossible. If all you know is that one pull did not succeed, prefer “It didn’t open on this attempt” or “After you pulled, the door was still closed.” Avoid turning a single outcome into an ongoing restriction.
When space is tight, B can be shortened to “You pulled the handle. The cabinet door is still closed.” This preserves the attempt and the state, at the cost of omitting a follow-up input. If the next step is clearly listed among nearby options, the repeated prompt can be dropped. Otherwise, retain actionable guidance.
Check revisions against the facts; do not use prose to patch the rules
When revising, first copy down the original feedback and the input that triggered it, then fill out the card. Next, choose a core response: identify where the action stopped if it never began; preserve the action and explain the result if an attempt failed; confirm the existing state if the goal was already met. Only then adjust tone and length.
Check the revised sentence against the card. Have you invented a reach in A, erased all action in B, or incorrectly described C as a newly successful opening? Then verify that every follow-up input really is available and that the response implies no unestablished cause, consumption of turns or attempts, or guaranteed success.
If the author has only an “unsuccessful” result and cannot confirm whether the character attempted anything, the response must provisionally stay with the known state, such as “The cabinet door is still closed.” This cannot establish how far the action progressed. The missing process must be confirmed through the story’s rules or action records; polishing the prose cannot supply it.
This card is suited to revising textual responses after a single operation. For consecutive actions, simultaneous actions by several characters, or a narrator who deliberately withholds facts, you must separately establish the time span and viewpoint covered by the response. Suspense can be preserved by withholding why the door did not open. But if the character did pull it, that attempt should have a clear place in the feedback.
