I'm nearing the end of RTK1 now so am thinking about the next step, and specifically which SRS system to use. The obvious choices to me are of course Anki and Trinity, I don't have any long term experience with either though so am finding it difficult to decide which to use. Whichever one I pick will have a lot of time and effort invested into it so I want to pick wisely. Has anyone here used both extensively, and/or switched from one to the other? Which do prefer and why?

I have found the stacks really motivating with reviewing RTK so that is a big plus for me with Trinity, I also like the idea of being to review vocab separately from sentences, as well as being able to keep track of which readings have been covered in the Kanji section.

I lean toward Anki however as I like the option of being able to test myself from English to Kanji or reading to Kanji in the future, just by adding the missing cards. (at first I'm just going to go Kanji to reading/meaning.)

[ファブリス]

Use whichever fills your needs.

Trinity is just barely started and is "frozen" for several weeks because I'm "refactoring" everything so even if you needed a simple import/export feature I can't build it now. So you should probably use Anki now.

If Trinity starts looking really sexy and you want it, you can always import your cards into it when the time comes

I personally prefer Trinity. It treats vocabulary with love and care. Having a failed stack helps to keep troublesome words in focus. Associating vocabulary to sentences or vocab lists or the great China Wall is simply brilliant.

But Anki is great too. I used it for a good while and was happy with it. Especially syncing is flawlessly implemented. And the scheduling algorithm schedules cards based on time since last review, not due date... nice touch!

I guess Anki is like a multifunctional swiss army knife; it's robust and will let you tinker around to your heart's content. Trinity is like a specialized kitchen knife; it's got fewer functions but does these functions well.

I've sort of made my decision to use both....Anki for all sentence purposes. And Trinity for Vocab, simple vocab like nouns and adjectives anyway. The intention being that I will add any such simple words to Trinity as I come across them in my continued studies, reviewing them before I review sentences. Anyone see any problems with this?

Do you have any idea how you will implement the import feature for Trinity? How will vocab from sentences be treated, as well as updating the Kanji section?

[ファブリス]

tomusan Wrote:Do you have any idea how you will implement the import feature for Trinity? How will vocab from sentences be treated, as well as updating the Kanji section?

Good question. I don't have means to do an automatic parsing of the sentences right now (extraction of vocabulary). Furthermore, all this vocab would turn into flashcards.

But it is still my intention that there is a flexibility there which is useful, wherein you can add a sentence with place names for example, and easy vocab for which you don't want to swamp yourself with additional flashcards. But for the sake of simplicity I am trying to keep the straightforward system where if you add a word reference into a sentence or vocablist, it automatically translates into a flashcard.

Thus, automatically extracting words and creating all those flashcards would not be optimal. Next is the possibility to mark words in sentences to extract separately, eg. *like this* but if the user do this, might as well use the Trinity interface after importing, as it will get better and more convenient as I develop it.

Perhaps it's fine like this. There are always compromises one way or another. So the first import version will likely be just import sentences with no attached vocabulary.

This is because I'm not aware of other software which creates sentence/words associations which could be imported into Trinity. If those programs exist and you use them, then of course it's possible to import those associations and create the vocab flashcards.

As for vocab lists, again good question. Format is not a problem (CSV, tab separated, etc.), but the user may end up with huge lists of vocab. If the plan is just to review vocab in a SRS fashion that will be ok. But the most useful aspect of the vocablists is to practice them separately, to prepare for tests or just review textbook lessons for example. In that case, short of importing separate short lists, the user will end up with a very long vocab list that they will probably not want to edit directly. Such long lists are not ideal because the UI will need paging etc. Which for the user is not great either. But I think I will add a simple delete button to the flashcard list so it will still be possible to delete specific entries this way.

Kanji wise when vocab will be imported the kanji counts, reading and kanji pages will work as expected. But for sentences the characters will be ignored so far as there are not attached words.

The kanji themselves could be counted towards the readings targets etc. but taking the assumption that only the words specifically added by the user represent what the user knows or wants to know, then kanji appearing in sentences should not automatically count towards readings targets etc.

Sorry for interrupting your interesting discussion, do you know where I can find more information about trinity, a website?

All information on trinity I know of is here

http://kanji.koohii.com/vocab/documentation.php

There is an alpha version you can try out running on the site atm.

[ファブリス]

jokoto Wrote:Sorry for interrupting your interesting discussion, do you know where I can find more information about trinity, a website?

See the Trinity sticky topic in the Feedback forum.