Flashcard overhaul wishlist
- Thread starter etm001
- Start date
Ok. Any idea when we are likely to get the option to be tested on example sentences without having to enter it all manually on a custom card? I suppose for starters you could just give an option for "show" example sentence in the test options menu, that seems like the option requiring the smallest change.
If you simply convert a dictionary-based to a custom card you should be able to edit the definition to strip out everything but the one example sentence you want to test on - obviously not a perfect solution but a lot easier than manual entry anyway.
Honestly I've given up on trying to predict release dates for things, particularly with the twin potential monkey wrenches of Android 5 and iOS 8 looming in just a few months.
Honestly I've given up on trying to predict release dates for things, particularly with the twin potential monkey wrenches of Android 5 and iOS 8 looming in just a few months.
Hi Mike,
I wish Pleco's overhauled flashcard system had as an option a feature included in Wenlin's by default: to repeat during the test the flashcards answered wrongly, and to keep repeating them randomly during that same session —whether or not one answers them correctly the next time they come up— in order to reinforce the learning of those items which, for whatever reason, one finds more difficult to learn.
I wish Pleco's overhauled flashcard system had as an option a feature included in Wenlin's by default: to repeat during the test the flashcards answered wrongly, and to keep repeating them randomly during that same session —whether or not one answers them correctly the next time they come up— in order to reinforce the learning of those items which, for whatever reason, one finds more difficult to learn.
That is actually what I was just about to ask...it seems like right now it is not possible to replicate Anki's default behavior, so wrongly answered cards get queued back into the same session?
Additionally the overall SRS algorithm that's used by default also seems a bit weird to me. When I answered correctly a couple times, the 3 options that I get are usually things like 12 days - 12 days and 8 hours - 12 days and 16 hours - and stuff like that. Shouldn't these intervals be spaced much further apart? Otherwise it seems to make no sense to even prompt for the quality of the response. I assume I can already change this somewhere, somehow? I'm not sure how, though...
Additionally the overall SRS algorithm that's used by default also seems a bit weird to me. When I answered correctly a couple times, the 3 options that I get are usually things like 12 days - 12 days and 8 hours - 12 days and 16 hours - and stuff like that. Shouldn't these intervals be spaced much further apart? Otherwise it seems to make no sense to even prompt for the quality of the response. I assume I can already change this somewhere, somehow? I'm not sure how, though...
Only possible with the 'weighted' review mode, which is kind of wacky otherwise (don't really want so much randomness) and is due for removal soon.
The dates listed actually aren't even reliable predictors of what the intervals will be - we recently removed them from the iOS version for that reason (we'd only added them recently and didn't think too many people would object). But narrow spacing like this isn't that unusual - the response factors mostly into the card's difficulty factor for computing future repetition intervals, so it's the cumulative effect of a card getting a lot of 4's or a lot of 6's rather than any one answer.
The dates listed actually aren't even reliable predictors of what the intervals will be - we recently removed them from the iOS version for that reason (we'd only added them recently and didn't think too many people would object). But narrow spacing like this isn't that unusual - the response factors mostly into the card's difficulty factor for computing future repetition intervals, so it's the cumulative effect of a card getting a lot of 4's or a lot of 6's rather than any one answer.
Hi,
Desired functionality: when creating custom cards, allow me to indicate that a portion(s) of the card definition should *not* be displayed when taking a definition-based test (i.e., displaying the definition and having to guess the correct word). After selecting to reveal the answer, the hidden text can be displayed.
Explanation: Most textbooks will include usage examples/sentences for vocabulary words. I include these examples in the card text when creating custom cards. The downside is that these examples are displayed along with the word's definition when taking a definition-based test. For now I just do my best to not look at the examples when taking this kind of test. But it would be great to have the usage examples shown to me after I select to reveal the answer.
Desired functionality: when creating custom cards, allow me to indicate that a portion(s) of the card definition should *not* be displayed when taking a definition-based test (i.e., displaying the definition and having to guess the correct word). After selecting to reveal the answer, the hidden text can be displayed.
Explanation: Most textbooks will include usage examples/sentences for vocabulary words. I include these examples in the card text when creating custom cards. The downside is that these examples are displayed along with the word's definition when taking a definition-based test. For now I just do my best to not look at the examples when taking this kind of test. But it would be great to have the usage examples shown to me after I select to reveal the answer.
Hi,
I tried Skritter Chinese for a week and I must say, it is quite good. In comparison to Pleco, the audio prononciation is more neat.
But in the end, Pleco can do pretty much the same thing except of one little detail:
If I wrote a chinese character wrong, the correct character with pinyin and sound is displayed for you in Skritter. You have to correct yourself in order to continue. This has a big advantage in the learning process. In Pleco, if I'm reviewing 50 char and did e.g. 10 wrong and have to repeat it at the end of a session, I usually forgot the ones I did wrong.
I tried Skritter Chinese for a week and I must say, it is quite good. In comparison to Pleco, the audio prononciation is more neat.
But in the end, Pleco can do pretty much the same thing except of one little detail:
If I wrote a chinese character wrong, the correct character with pinyin and sound is displayed for you in Skritter. You have to correct yourself in order to continue. This has a big advantage in the learning process. In Pleco, if I'm reviewing 50 char and did e.g. 10 wrong and have to repeat it at the end of a session, I usually forgot the ones I did wrong.
HW60
状元
When trying to catch up a backlog, the overdue interval of the flashcards in combination with the card score maybe a good information to decide which cards to catch up first. The overdue interval seems to be visible only in the Top Fields of the Flashcard Interface. It would be helpful to search for overdue cards, and even more helpful to add an Overdue Card filter for flashcard sessions.
I wish there were separate fields for the dictionary definition and user notes, so that I could add a tiny sticky note "only used in mainland China"/"usually pronounced XX" to a flashcard without editing its definition. Right now I'm hesitant to add notes to cards because it means the (now "custom") PLC/CEDICT definition will never receive updates, and I can also never update the card to a fancy dictionary if I'll ever need more than PLC (well, at least not without losing my notes again).
Probably just a programmer's OCD.
Probably just a programmer's OCD.
@Kamuro - not quite sure if I understand; we display the correct character on a wrong fill-in-the-blanks answer too. Is it that you want to be forced to write it again?
@HW60 - we actually already prioritize SRS cards by how overdue they are; is there any reason you'd ever want to find them aside from wanting to review them first?
@jlnr - definitely coming, custom fields are long overdue actually. (for other stuff too, like Cantonese)
@HW60 - we actually already prioritize SRS cards by how overdue they are; is there any reason you'd ever want to find them aside from wanting to review them first?
@jlnr - definitely coming, custom fields are long overdue actually. (for other stuff too, like Cantonese)
HW60
状元
I have quite a lot of overdue cards and cannot recover from this backlog within reasonable time. Therefore I am working since several months with a combination of two profiles: one is standard SRS with only several categories, the other is SRS with a score limit. The standard SRS is for learning new cards in new categories and for slowly adding more and more categories to reach the standard SRS with all my categories. The standard SRS with a score limit helps not to forget cards completely, starting with a low score limit and increasing the upper limit more and more. Both profiles work together with the same cards.
I think that the standard SRS with a score limit can be improved with a overdue upper and lower limit and a score limit at the same time and would like to test it if it really works. If for instance my upper score limit is 500 and there should be more cards to review than I like to, I would like to test the cards with a high overdue first. It could be even a good idea to sort all due cards by score/overdue (increasing) and start reviewing the cards with a low score/overdue first.
Probably this is not the kind of problem a Pleco beginner has. But I have almost 14000 flashcards now and will never catch up. But I like my actual system and think it could be improved. Maybe there are more users who intend - not necessarily by their own will - not to catch up.
Honestly, any solution that requires manually manipulating a lot of variables is likely to go over the heads of 99% of our flashcard users; we do need to do something for people who have too many cards due for review, but as I've said in a few other places, that solution is going to need to be largely automated in order to be worth adding. We don't plan to make our flashcard system any less powerful - quite the opposite - but we do need to make it less complicated, or else the only people left using it will be the ones who've been using it since 2008.
HW60
状元
I fully agree with the goal of a less complicated flashcard system and cannot recommend my system described above for other users because it is by far too complicated, needing Excel to keep track etc. But actually I like my system very much: I do not loose cards out of my focus which I had learned some time ago, I slowly catch up, I can learn new cards, and I can determine how many cards I want to review today, which can mean no cards today without causing any greater problems. I think my system can become a little better using the overdue information, and would like to test it.
Pleco's flashcard system should only ask the user "how many cards to review today?", when the backlog is apparently to big to catch up in one or two sessions, and otherwise should not even ask that question.
Probably my post was still too early - I have no idea how Pleco flashcards finally will work and just wanted to make a proposal for the usually hidden technique, not for an even more complicated user interface.
I'm sure I've seen this somewhere in the forums, but no search will bring it up:
Hands-free Flashcards!
... I dream of the day when I can use TTS in both English and Chinese and have audio flashcards and be running on a treadmill and practicing as I exercise... Okay, actually I never run on a treadmill, but other people do, and I'd still totally use them when walking to classes, the bus, etc.
However, this brings up the question of how you would select all of the different choices, hands free. Supposedly, back in the day Ankidroid tried to do this with a Bluetooth gaming controller called "Zeemote". It's wireless, hopefully ergonomically designed (I've never actually used it), and is used with one hand. There must be other wireless Bluetooth game controllers or something out there that you could also use... Have you had any other ideas of how to do hands free? Maybe even by voice, although to be honest, from what I've seen, it would be pretty annoying to use voice commands, because they might mess up. Furthermore, it would really limit where you studied, and background noise on a train, for instance, might be a problem. I'd much rather some kind of controller thing...
One more suggestion pertaining to audio flashcards.
First let me say that I doubt you will want to do this, but I'll ask anyways. The tts (the one I have is Hui) can be really nice for some sentences, and not so nice for others. I'm sure you guys are working on this, but I'm wondering you'd also consider letting people import audio with their flashcards. No matter how good tts is, I'll probably almost always go for the native audio if I have it (which, of course, when I don't, tts is great!). What I'd really like to use it for is to splice movies up into small fragments (using subs2srs) and making audio flashcards with the audio. For some movies, it's so amazingly natural - no textbook perfection for me, no sir. It really helps for hearing (and trying to mimic) how people actually speak. What would be awesome is if we could import any audio, to the front or the back of the card. That would be totally cool.
Hands-free Flashcards!
... I dream of the day when I can use TTS in both English and Chinese and have audio flashcards and be running on a treadmill and practicing as I exercise... Okay, actually I never run on a treadmill, but other people do, and I'd still totally use them when walking to classes, the bus, etc.
However, this brings up the question of how you would select all of the different choices, hands free. Supposedly, back in the day Ankidroid tried to do this with a Bluetooth gaming controller called "Zeemote". It's wireless, hopefully ergonomically designed (I've never actually used it), and is used with one hand. There must be other wireless Bluetooth game controllers or something out there that you could also use... Have you had any other ideas of how to do hands free? Maybe even by voice, although to be honest, from what I've seen, it would be pretty annoying to use voice commands, because they might mess up. Furthermore, it would really limit where you studied, and background noise on a train, for instance, might be a problem. I'd much rather some kind of controller thing...
One more suggestion pertaining to audio flashcards.
First let me say that I doubt you will want to do this, but I'll ask anyways. The tts (the one I have is Hui) can be really nice for some sentences, and not so nice for others. I'm sure you guys are working on this, but I'm wondering you'd also consider letting people import audio with their flashcards. No matter how good tts is, I'll probably almost always go for the native audio if I have it (which, of course, when I don't, tts is great!). What I'd really like to use it for is to splice movies up into small fragments (using subs2srs) and making audio flashcards with the audio. For some movies, it's so amazingly natural - no textbook perfection for me, no sir. It really helps for hearing (and trying to mimic) how people actually speak. What would be awesome is if we could import any audio, to the front or the back of the card. That would be totally cool.