I was actually planning on using Excel@Japanese until I found out that it doesn't (didn't? Perhaps they fixed it: whatever, it doesn't matter now) play nice with Office 2007. But, whatever: Anki probably blows it out of the water, too, although admittedly not having to type up flashcards is pretty alluring.

I think the expression is 'horses for courses'.

Excel@Japanese really helped me prepare for the JLPT last year.

It's a shame to see it getting a dissing considering Mike is a member of the forums too

A new version, called Tango, is in the pipeline. Not too sure when it will be ready though.

Writing an importer shouldn't be too difficult. I don't have Excel but if OpenOffice can be used to export a CSV of the 'Words' sheet, importing should be pretty easy. Start by sending me their file so I can see how the 'last test' and 'first successful test' fields are stored.

Megaqwerty: Excel@Japanese, like most programs out there that come with premade material, bases the entries on edict. By the looks of it, Excel@Japanese has trimmed the definitions of JLPT3+4 to make them tidier (though a couple have been missed). JLPT1+2 definitions seem to be verbatim from edict.

The problem with studying like that is while it's useful for simple nouns and adjectives, it's not so useful for verbs and more complicated adjectives/adverbs/nouns, because even if you learn the word, you don't know how to use it. Using the Tanaka corpus can improve that somewhat, but the Tanaka corpus is full of strange translations and duplicate entries.

Synewave: from the posts on the forums, Mike seems to be aware of the problems with using Excel as a development platform. I'm assuming the previous posts were more a comment on Excel than Mike's (and the original author's) work.

Yes I really did mean using Excel makes it buggy, not that his work was a waste or anything like that.

Resolve?

Was wondering why for the past two weeks the graphing feature in my Anki copy hasn't been working at all. Would you like to have a look at the log file? (If yes, I'll mail it to you.)

Bringing it up on http://repose.ath.cx/tracker/anki/ is probably the best way. You can post the log file there too.

0.3.0 is getting there - most of the core code is done now and I just need to get the new add/edit cards stuff going.

I see. I've never used this sort of error-logging site before, but I've registered and posted my error log, and I hope I did it properly though.

P/S: Can't wait for Anki 0.3.0! I know software developers and programmers hate it when people ask this, but, well... umm, have you projected any date for its release?

P/P/S: If you need a mirror to host the file, I'd be happy to help.

In the "edit deck" screen, it would be nice to have a column giving the time/date of the last review, mainly to make it easier to find recently reviewed cards. IMHO that would be more useful than the "date created" column or the "time until next review" column.

The "time until next review" column is tremendously useful to me but I've never had any use for the "date created" column.

I like the date created column so I can find recently added cards that I realized I made a mistake on...

A great example of how everyone's needs differ, I guess :-) I was about to explain that the 'created' field is there so you can find the most recent and oldest cards.

brose: What part of your workflow requires you to see recently reviewed cards? Selecting the current card is not enough?

I think the grid display risks becoming too cluttered if there are too many fields, but I may add support for sorting on other attributes or customizable fields at some point in the future.

wzafran: Thanks, I got your bug report fine. I'm hoping to have an alpha version out for testing within 2 weeks, but it really depends on how busy work and life is in the mean time. I'll do my best!

During the review, occasionally I pass a card that I meant to fail, then I can't find it again, or there was something I wanted to go back and add to a card after I finish reviewing the deck, but then I can't remember which card it was. If you want to avoid clutter you could make the "date created" button toggle to the "date reviewed".

A preview of the changes coming in 0.3:

http://repose.cx/dump/sneak.png
http://repose.cx/dump/sneak2.png

Whoa, 0.3 looks like it has a ton of features I've been looking forward to, very cool. Thanks for adding the buttons to the review section, I have a tendency to hit "Y" when I use the keyboard.

Templates and card tags look very sexy as well.

Yes, I agree with suffah, 0.3.0 seems very sexy!

drooooool..................

Having only just finished RTK1 I've only today started investigating the options for an SRS to use for sentences. Anki is definately my favourite so far (and everyone here seems to like it!).
I especially like the auto-kana completion. That will save a lot of time. My only request would be to have some way to tell which kana go with which kanji. If you have a two kanji compound, but three kana for it you don't know what to read for each kanji. Perhaps having something like this:

有名な俳優です。
（ゆう・めい）な（はい・ゆう）です。

I can add it in for all my sentences, but then it's just as quick to type the whole thing again myself with the formatting than just alter the one provided...

How do you guys deal with the specific kanji readings?

The simple answer is that with a bit more experience, the divisions between kanji are obvious 99% of the time, as the 音読み readings are not words like うめい and はいゆ. As a rule, they're a single sound.

Playing devil's advocate, with a bit more experience, the readings of the kanji are obvious.

The point of the furigana is to provide readings of the kanji and thus it should effectively do that with middle dots so that the incredibly awesome tech of auto-completing readings is not lost on those who are not yet as awesome as your program.

The tools I use to generate furigana don't support it out of the box. I'd have to make some sort of hack that feeds the kanji characters at a time, and it would break if the kanji has a kunyomi. Perhaps I'll move to another parsing engine in the future, but not right now. Suggestion noted, anyway!

The reason I asked for the kanji/dot feature was because I'd seen that on alljapaneseallthetime.com

I've since emailed the guy who does that site to ask him how important he thought the dots were, and he said

Quote:It is not necessary at all. I just do it to help people. I never have dots in my own stuff.

So, I guess I won't worry about it

Has anyone had any luck importing mnemosyne decks into Anki? I've tried a few times but every time I do, nothing happens, and I get a bunch of errors when I close the program. What am I not doing right?

JimmySeal Wrote:Has anyone had any luck importing mnemosyne decks into Anki? I've tried a few times but every time I do, nothing happens, and I get a bunch of errors when I close the program. What am I not doing right?

I do it all the time:
1: Set up Anki for the import (create new deck whatever).
2: "Import Mnemosyne Deck" from menu.
3: Navigate to .mnemosyne folder
4: Select correct .mem file
5: Anki goes to work.

If your doing all that then you could try to figure out if its your mem file that is bugged or Anki. I would try export to xml then create new mem deck and import (all in mnemosyne ) then save and try to import that file to Anki.

That's weird. I'm doing exactly that, and no luck. I've tried it on a brand new Mnemosyne deck too (just add one card and save it). I get this error:

Code:

Traceback (most recent call last):
  File "ui\main.pyc", line 568, in onImportMnemosyne
  File "ui\importing.pyc", line 48, in importMnemosyne
  File "ui\importing.pyc", line 55, in importGeneric
  File "anki\importing.pyc", line 132, in importFromMnemosyne
  File "pickle.pyc", line 1370, in load
  File "pickle.pyc", line 858, in load
  File "pickle.pyc", line 1181, in load_append
IndexError: list index out of range

What version of Mnemosyne do you have? I have 0.9.7 .

Hi JimmySeal,

Could you email me your deck?

Edit: the error is in the library code, so a problem with different python versions or Windows vs Unix newlines wouldn't surprise me. I'm assuming both you and dilandau use the premade binary on Windows?