Nice.  Its a problem to answer 一月　（いちがつ）　and 一月　（ひとつき） though.

alantin wrote:

I think I found a little mistake though.

==========================
お姉さん - おねえさん
(Hepburn: oneisan / Nihon-shiki: oneisan)
older sister
==========================

Isn't "お姉さん" translitterated in Hepburn as "oneesan"..?

Edit:
Well.. I just noticed that it takes hiragana input so that fixes it!

Thanks, I fixed up the oneesan and one's like it, should be working correctly now.

alantin wrote:

A great little app!!
Some little problems with words with more than one reading,
(行く = いく / ゆく, 一日 = いちにち / ついたち, 七 = しち / なな) but I'm lovin' it!

I'm trying to add some example sentences in the next version which I hope by context will help sort out which pronunciation it is looking for.  I agree at the moment the definition is pretty vague as to which one it wants, and if you turn off the definitions, there is absolutely no way of knowing what it wants as one user pointed out to me.  Still, I'd hate to take them out, I think those readings are important.  I'll see if I can't reword the definitions for the next release and together with the example sentences hopefully it'll be more clear.  If not, I'll get rid of them then.

radical_tyro wrote:

Hi joeellis,

I appreciate your work and it's a nice site! I haven't had much time to play with it, but I wonder why someone might use this instead of Anki with JLPT lists.

alantin wrote:

I think, having to write the pronunciation makes you much more active than if you were just reviewing with anki. Actually I think the strong kinesthetic component in this app makes it (atleast for me) way superior to anki with the JLPT lists in regard to learning the vocabulary as individual words.

I think it is perfect for a kinesthetic learner but perhaps for someone, who has little or no kinesthetic aspect in his learning style, it would be of not much use.

alanthin's explaination is spot on I think.  Having to actually type it out forces me to really think about the answer.  A good example where it helps would be voiced words like 天国.  If I use regular flashcards or Anki and see 天国 I would see "てんごく" and think, "oh easy enough, got it."  Now I take the JLPT and 天国 comes up and I have four choices:
1.てんこく
2.てんくに
3.てんごく
4.てんぐに

Now, whereas I thought I knew the word, I'm suddenly confused.  I'm pretty certain it's not 2 and 4, but I just can't decide between 1 and 3.  It's such a slight difference.  And this confusion not only takes up time but it's pretty maddening because it's a "on the tip of my tongue" kind of situation.  So I'm hoping that by actively typing in the answer,  the meaning, and more importantly the reading itself, will stick in your brain, you can reduce such frustrating moments come JLPT time

Last edited by joeellis (2008 November 05, 7:03 pm)

wow. I've gotta say, I've been wanting something like this ever since they introduced that excel file that functions similarly, yet doesn't (or didn't) work with OpenOffice at the time I tried it (a year or two ago). Finally, someone made a working version one better, one that we can access from any computer without opening a huge file that takes a long time to load on slower computers. Serious props to you!

Hey, the tanuki corpus might be handy in the development of this tool.

I'm missing the sentences power in this app.

mentat_kgs wrote:

Hey, the tanuki corpus might be handy in the development of this tool.

I'm missing the sentences power in this app.

I think he could use the Tanuki corpus. Check out the license that came with the files:

Tanuki license wrote:

This work is licensed under the Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License.

I didn't realize so many people want to be able to type their answers in their SRS. In the case of Anki, it's on the bug tracker as a very low priority enhancement. But if you go to the page:

http://code.google.com/p/anki/issues/detail?id=18

and star it, maybe Damien will realize more than a few users want this. Of course the quickest way to get the feature implemented is to write a plugin but no one has done so thus far, afaik.

It's cool though that joeellis has made this site for everyone to use.

Been using the site for about an hour now, starting from the bottom (JLPT4) just to make sure I know everything, and right now my statistics are:

Words Seen: 347
Successes: 1311
Failures: 17

This doesn't seem like the Leitner system to me at all. If I input a word correctly, I don't need to see it 30 seconds or even a minute later, especially if there are more words I haven't even seen yet. Maybe if I failed the word once before due to a typo or brain fart or whatever reason, okay throw it back at me, but the maximum number of words I should need repeated for that reason is 17. If it worked that way, I'd have been done with the JLPT4 stack 30 agonizing minutes ago. At first the words were fun, but now I find myself bored out of my mind going through piles of words I've passed 3,4,5 times already, actually getting excited when I see one one of the 10 remaining untested words I've been digging for.
Also, I assume these words don't expire or anything like our cards on RevTK, because even if they did, I'd have seen them 10 times before in my dig for new words, so they'll probably never expire if I use the site everyday. If possible, please make the cards behave like kanji cards on this site, showing up once before being held in a box until expiration to be released on us again.

-edit-
I moved up to JLPT3 without waiting around to see the 8 or so cards that never came up in JLPT4, and that broke the monotony for a while. I'm now a little over halfway through the JLPT3 deck (291 complete, 274 remaining, 16 failed), and although I keep getting the same cards a lot, it averages at about 1.44 displays for each card, which is slightly more bearable than 3.78. That average is sure to get higher as I near the end of the deck and try to see all the unseen cards, but I'm sure I'll quit for the night before it gets too frustrating. I just hope by the time I "finish" the level 1 section the site will have some "replay value" by holding cards until expiration.

btw, in the JLPT3 deck, the word 応接間 is defined as "drawing", when it should be "reception room".

Anyway, I don't want to sound all nit-picky when I think the site is a great start to something I could see myself using even more than RevTK (since there's tons more cards to review at ReadTK), so I'll just say thanks for starting the project.

Last edited by mantixen (2008 November 06, 12:29 am)

radical_tyro wrote:

In the case of Anki, it's on the bug tracker as a very low priority enhancement. But if you go to the page:

http://code.google.com/p/anki/issues/detail?id=18

and star it, maybe Damien will realize more than a few users want this.

A dumb question!
How do I star it?!

joeellis wrote:

I'm trying to add some example sentences in the next version which I hope by context will help sort out which pronunciation it is looking for.  I agree at the moment the definition is pretty vague as to which one it wants, and if you turn off the definitions, there is absolutely no way of knowing what it wants as one user pointed out to me.  Still, I'd hate to take them out, I think those readings are important.  I'll see if I can't reword the definitions for the next release and together with the example sentences hopefully it'll be more clear.  If not, I'll get rid of them then.

I don't think getting rid of them alltogether is such a good idea..
How about just making it accept both options in these cases?

alantin wrote:

radical_tyro wrote:

In the case of Anki, it's on the bug tracker as a very low priority enhancement. But if you go to the page:

http://code.google.com/p/anki/issues/detail?id=18

and star it, maybe Damien will realize more than a few users want this.

A dumb question!
How do I star it?!

There is a little star directly to the left of where it says:

Issue 18:       Type in text boxes for the answer

Wow, thanks to everyone for the great comments!  The algorithm definitely needs work, but I didn't realize it repeated so often.  I'll really look into trying to reduce that as the algorithm is my top priority.  Once I get that down, it's much easier to improve all the other issues on the site.  And to mantixen, I'll make the cards expire

I also agree that each level should have some sort of "end" or at least let you know when you've gone through all the cards in that level (and make sure it actually will display all the cards of a given level for you!).  Ahh, back to the drawing board! 

By the way, I really should open up a bug tracker of some sort, but in the meantime, if you have any problems with the vocabulary definitions or anything, I'm correcting each one as I see it, so just let me know, they are usually corrected within a day or so.

alatin wrote:

I don't think getting rid of them alltogether is such a good idea..
How about just making it accept both options in these cases?

Ah sorry, by get rid of them, I mean reword the current definitions to something less vague

thomasritz wrote:

First of all, it's a great app!

For me, 外 should be そと, not ほか, with definition 'outside', not 'another / other'.

I also had problems when 四 sometimes showed up as よん, but sometimes as し. How should one know which reading is requested?

Japanese is a messy language - 外 can be そと (outside) or ほか (other)... 

HINT:  I don't know how 外 is presented, but I have found that if you pay attention to the parts of speech in the definition (like suf for suffix often implies some kind of 音読み), it helps me guess the reading required. 

IMHO, the app isn't perfect, but it's really fantastic for practicing readings.

Tobberoth wrote:

Wow, I had no idea 外 could be used for ほか, during my year in Japan I only ever saw 他 used.

One quick way to check is to type ほか on your computer and scroll through the list of possibilities. 

BTW, if you study Hanzi, 他 is actually used for "he" while 外 really means "other or outside", so I was actually shocked the other way when I saw 他 used for "other" in Japanese!  This is a little irrelevent as Kanji-use can have nothing to do with the original Hanzi-use, but it's sometimes useful to know.

Japanese is just a mess of a language.  For example, when 斬る with a definition like "to behead" came up, I totally guessed きる thinking of 切る, of course.  WTF - I was right!!! 

Once in a while these kinds of guesses just work in Japanese like guessing のぼる for some kind of raising or climbing, etc... (there are actually a bunch of のぼる with similar meanings - likewise for あらわす).

I don't know about this history of Kanji use so I can't say for sure, but I would bet that many of these words existed in the verbal-sense, before Kanji was used.  I would conjecture that as Kanji were introduced, one reading would be assigned to different Kanji to distinguish different uses of that word like きる.

Looking it from another perspective, a common Kanji like 上 has weird readings beyond うえ and じょう - you will commonly see かみ in place names (this also applies to しも for 下 and other readings like うお for 魚 which most people would expect to be さかな or ぎょう)

And don't get me started on some freak readings like 素人 しろうと when any normal student would be expecting にん, じん or　ひと. 

This is why I just skipped RTK2 and went straight to learning vocab.
It's a mess.

Last edited by kfmfe04 (2008 November 07, 4:18 am)