Lea las últimas noticias y actualizaciones en la sección de noticias:
NoticiasV4 update specific notes
General update considerations
We always aim to ensure continuity of users ability to use MEMORY with all updates. We strive to release every update with miniminal alteration to current user's usage.Algorithm Update
Existing users should read the document to see the updates to the algorithm in V4Read about the algorithm update
Please remember: If you go back to the old version, any features such as rich-content question cards won't be visible in the original version. Creating a course on the new version with the new features (e.g. manual mode or rich questions) won't be visible to other users and may have unexpected effects.
OVERVIEW OF FEATURES
- Introductory guided tour for the create course screen
- Improved course export now exports multi line questions easier, and includes both images and custom audio
- You can how add rich content cards to levels - dress up your course and give users an introduction to your course. Level content is accessible during learn mode.
- Design your flashcards and questions: Questions and answers can now have rich text and images added to them
- Select your questions during learning mode
- Add stars or remove stars during learning to add questions to saved sets
- Go back a question whilst learning
- Retry a question whilst learning - NOTE: Retrying still considers the previous answer as incorrect
- MEMORY is now available in Spanish with other languages to come. Selecting a language does not affect your courses.
- New bulk import: Use an interactive Excel-like spreadsheet to add questions (optionally switch back to textual input)
- One by one view: Edit questions one by one without loading the entire level (useful for courses with lots of questions per level). If you load questions with 200+ questions per level, you may often experience screen lag. This functionality covers that.
- Drag and drop questions to reorder them in card view
- Print PDF now works in many more languages - with supporting fonts
- Filter questions whilst you edit
- You can now shuffle all questions in user settings - to reset the 'due now' time to 'now'. Doing so does not affect history, but it allows you to reconfigure your intervals according to spaced repetition learning algorithm adjustments.
- Fixed course password entry on course dashboard screen saying you are not logged in as the correct user (fix is retrospectively applied in v3)
- Flashcard preview mode now respects course creator's question ordering
- User's default question creation order is preserved throughout the site, e.g. Print PDF (in correct order, or Select Questions to study)
- We removed automatic capitalization of multiple choice questions - by popular request
- Listening mode: For same language courses, this now acts as a spelling test instead of typing the answer, you type what you hear
- Speaking mode now functions as a "say what you hear" mode for same-language courses
- Listening mode now has a higher chance of being typing mode instead of multiple choice
- NEW: Manual mode: Select exactly how you wish your questions to be served, which order, which question type, answer type and languages
- Question types are now smarter - based on your answer history for a question, we use a more sophisticated formula to detect which type of question to serve in smart mode
- Higher guarantee of question types returning a better type in smart, custom and manual modes - using new weighted random algorithm
- Less waiting: High XP users should no longer noticeably slow down when doing things such as completing sessions and viewing graphs.
- Improved in-learning speed: no more janky flashcards, smoother experience
- Less bandwidth - faster page loads and automatic optimised images
- Faster page loading on card view mode
- Server side rendering - the site is now rendered initially by a server to improve response times
- The Skeleton bandwagon - out with the 'ajax spinners', in with the Facebook style Skeletons whilst components are loading
- Less javascript - faster page loads from code splitting