Anki user here, for about 5 years now studying Korean and to a lesser extent Japanese and miscellaneous topics.
This is cool, but I will never base my learning on a closed-source product. I simply don't trust anyone with data / workflows this important to me.
Anki doesn't have great UX, but it is -dependable-. It works, day in, day out, and I've had zero issues with it. I trust that in 5-10 years, my anki decks will be there and work exactly the same.
edit: I'm not trying to be negative. I think there is a lack of spaced repetition-based flashcard apps with great UX. However I personally value stability and guarantee of longevity over nice UX. If the developer of Anki quits tomorrow, it being open source ensures that it'll live on with new maintainers (even if that's just me).
I'm glad this comment was the first thing I saw. My first thought was "is it free software?", because it's a clear downgrade and not worth using over Anki if it's proprietary.
I studied Japanese, and went from 0 to JLPT N2 in 4 years. I attribute a lot of that to Anki/Ankidroid.
This is how I did it.
Usually when you start a language, you start with a textbook. It might be Genki Vol 1, or Nihongo So Matome 1, etc. I created two decks - a kanji deck, and a vocab deck. These decks were only for this single textbook. Every week when new kanji or vocab was introduced, I'd add those to the deck, and become familiar with them over the next week (while at the same time, reviewing existing vocab/kanji).
When that semester was over, and we started the next textbook in the new semester, I'd create two new decks. And so on and so on.
I think it's important to create the decks yourself. It gives you control, and makes you feel more invested in the process. For vocab, I'd always include example sentences. For Kanji, I'd actually have representations in two different fonts. One in the standard kanji font on your computer, but another in HonyaJi, a "hand-written" kanji font. For me, I just felt it helped when writing out the kanji to see how it looks handwritten.
The great thing about Ankidroid is the whiteboard feature. Use it!
If you're not studying a language with logographic characters like Chinese or Japanese, then great, you only need to worry about vocab!
(As for grammar, I never relied on Anki for that. Practice speaking, practice writing. Find local "conversation classes", where you meet up with native speakers. They can practice their English with you, and you can practice your new language with them. As for writing, check out a website called Lang-8. Native speakers look over your work for free, and in return you just take a few minutes to read over their stuff in English).
I concur with all of these points. I've also written a bit about Anki and updating some of the default settings for better retention. [0] Making a good template for cards changes everything, particularly with adding context / usage sentences on the back of the card. Also, switching to using the native definitions once you can instead of the english translations helps immensely.
For me, I -hated- how long it took to input words into Anki, so for ~7 months I used Evita's Korean Deck which has about 5.5k words. I eventually burned out and got tired of the lack of context in the cards, but that helped me learn a lot of day-to-day vocabulary. I tried this a bit with a Japanese 2k deck but I think with logographic language it doesn't really work as well as slowly inputting words that you know the context/definition for.
Nowadays I write down words from books/movies/etc that I don't know in my journal, and every once in a while I collect those into a personal anki deck. I actually am working on a CLI program that helps automate the process of searching a word's definitions and creating a card. It outputs all of the selected definitions into a csv, which I can then bulk import into Anki. [0]
It's important to do it every day. Anki by default allots 20 new words per day, but this is quite a lot unless one plans to spend a while every day on anki. I find 5-10 words is more tenable, and simply 1/day is better than 0/day if one burns out.
It's not either one or the other. Anki is a supplement to input received from native speakers and media. The idea of spaced repetition is to help one efficiently remember the thousands and thousands of words needed for a sophisticated vocabulary.
I definitely don't recommend anyone to try and learn a language from e.g. Anki only.
Indeed it's not one or the other, but if Krashen's "comprehensible input" theory [1] is (mostly) true, the role Anki (or any SRS system) or grammar books plays would be very small (if not non-existent), and many polyglots' opinion on Anki seems to back it up.
This is cool, but I will never base my learning on a closed-source product. I simply don't trust anyone with data / workflows this important to me.
Anki doesn't have great UX, but it is -dependable-. It works, day in, day out, and I've had zero issues with it. I trust that in 5-10 years, my anki decks will be there and work exactly the same.
edit: I'm not trying to be negative. I think there is a lack of spaced repetition-based flashcard apps with great UX. However I personally value stability and guarantee of longevity over nice UX. If the developer of Anki quits tomorrow, it being open source ensures that it'll live on with new maintainers (even if that's just me).