Anki Spaced Repetition Program

Anki add-ons are implemented as Python modules that Anki loads during its startup process.

Add-ons interact with Anki through two primary mechanisms:

1. Hooks: Add-ons can register themselves to be notified when specific actions occur (for example, a hook that triggers when the browse screen is loaded).
2. UI Modification: Add-ons can modify the user interface, such as adding new menu items, in response to those actions.

While the core UI is written in Python/PyQt, certain screens (like the review screen and editor) utilize TypeScript and Svelte. To develop add-ons, familiarity with Python is required.

The Anki Rust API is primarily composed of two main crates located in the rslib/ directory:

• anki: The core library containing the main logic.
• anki_io: Provides I/O utilities.

The table rows represent cards or notes matching your search.

• Selection: Drag the mouse or hold <kbd>Ctrl</kbd> (or <kbd>Command</kbd> on Mac) to select multiple rows. When multiple rows are selected, the editor is temporarily hidden.
• Selection Logic:
  • In Cards mode, a note is selected if any of its cards are selected.
  • In Notes mode, a card is selected if its note is selected.
• Current Item: Clicking a row selects it as the 'current' card or note.
  • In Cards mode, the current note is the note of the current card.
  • In Notes mode, the current card is the first card of the current note.
• Visual Indicators: Row background colors indicate status:
  1. Flagged: Uses the flag color.
  2. Suspended: Yellow.
  3. Marked (Notes mode only): Purple.

Anki uses a spaced repetition algorithm rather than a traditional flashcard 'cramming' approach. Instead of you deciding when to stop reviewing a card, Anki decides when you are most likely to forget it and schedules the review for that time.

• Spaced Repetition: If you answer a card correctly, the delay until the next review grows larger (from days to months or years).
• The 'Again' Button: If you press Again, the card will be shown again soon.
• Efficiency: Reviewing material multiple times in a short period (cramming) is less effective than following Anki's scheduled delays.

If you feel like you aren't seeing all your cards, it is likely because the algorithm has determined you have already mastered them or they are not yet due for review.

While shared decks are useful for saving time on simple factual subjects (like capital cities or flags), they should be used as a supplement rather than a replacement for external study material when learning complex subjects.

Key considerations:

• Context: Many shared decks lack the background explanations or context that the original creator already understood. This can make them difficult to use without external textbooks or teachers.
• Active Learning: Creating your own decks is often more effective for complex subjects (like languages or sciences) because the process of inputting information forces you to identify key points and build a deeper understanding.
• Language Learning: Avoid simply memorizing long lists of word-translation pairs; effective language learning requires exposure to real-world sentences and context.

Users own the cards and decks they create in Anki and are free to license or sell them. Anki does not implement Digital Rights Management (DRM) to prevent deck sharing.

If copyrighted content is found on AnkiWeb's list of shared decks without permission, users can contact the Anki team to request removal.

All fields within a oneof block are implicitly optional. The same caveats regarding default values in Python and Typescript apply to oneof fields.

In Python, you can use WhichOneof("field_name") to determine which specific field in the oneof group is currently set. It returns the name of the set field as a string.

message Foo {
    oneof bar {
      string name = 1;
      int32 number = 2;
    }
}

message = Foo(name="")
assert message.WhichOneof("bar") == "name"

Desired retention controls the probability that you will remember a card when it is scheduled for review. The default value is 0.90.

Impact on Workload

Adjusting this value has an exponential impact on your daily review workload:

• Higher Retention: Increasing retention (e.g., to 0.95 or 0.97) significantly shortens intervals, meaning you must review cards much more frequently.
• Lower Retention: Decreasing retention reduces workload but increases the number of cards you forget, which can be demotivating and may eventually increase workload if forgetting becomes too frequent.

Recommendations

• Keep the value between the Compute Minimum Recommended Retention and 0.97.
• Use the Compute minimum recommended retention (CMRR) tool to find a baseline that balances learning efficiency and workload.