There are a few things that should be implemented into the SRS of Logseq:
An easy way to add audio/image from āassetsā folder. Currently to access a linked audio file in multiple platforms is a pain due to the full path url requirementā¦ also plugins do not help much on Android or IOS. It could be easily added to the slash command like ā/ Assetsā or ā/ Mediaā etc. Would be nice to be able to search the files by name through the slash menus when inserting. Also, consider to provide at least 2 options for playback - a full player and single play button (like in Anki)
Cloze hierarchy should be a thing. Like in Anki, āclozeā should have a way to hide and show parts of a sentence independently, nut just collectively. (cloze 1, cloze 2 / cloze #1, cloze #2) Currently if I want to hide 3 information in a sentence, I have to make 3 cards otherwise, I hides all of them at once.
Provide a way to review a keyword tagged cards separately. Hashtags behave like decks, but one must write a query to be able to review 1 specific ādeckā and not every card in the database.
I think audio in flashcards is craved by all of us who intend to use Logseq in their studies (not just language studies), and should be implemented with and easier access by the assets folder (not a full path url, thatās only suitable for desktop users and unusable on other platforms).
Please make this happen! It would be amazing to have my language notes and language flashcards (with audio) in the same app and be able to look up notes on grammar, without jumping form one app to another.
Thanks for reading this post, vote if you too think itās needed in Logseq!
Yeah been missing a native add local asset too. Theres a plugin called localassets that does it via filename (and filename only), but a call to shell (window to choose file) would natively would be great for seeing the assets folder in thumbnails instead of remenbering filenames
I also vote on the cloze. For now theres a āworkaroundā, indenting copies of the sentence under it and then deleting the areas on the parent sentence; So sentence would be redacted with whatever youd type to signal something there was ocluded, and unfolding it would be the āopen clozeā
lorem (ā¦) est parent
lorem ipsum (ā¦) est
lorem ipsum dolor est
Its cumbersome, involve more steps, redundant data and the reveal would be linearā¦ but at least its an way for now.
Thereās now built-in audio support. You can drag-and-drop an audio file and itāll be added to the assets folder. However, you still need to add a ! in front of the code to actually render the audio player. Like this: 
One thing Iād like to see next is to enable cards within cards. What now happens when I nest a card within another card, that all cards are shown when reviewing the main card. The āchild cardsā show fine; itās just the main card containing child card(s) that shows everything:
It seems that Logseq does not support card ālearningā phase: if you click āForgottenā, this card wonāt be redisplayed until tomorrow. This has negative effects on card retention. Hope this can be fixed soon.
The flashcard algorithm is broken. SM-5 does not work well, the author of the paper acknowledged they didnāt include all the ātricksā needed to make it work. This is why Anki does not use it, by the way. Is there any chance one of the great developers of Logseq could implement SM-2 or, even better, FSRS?
