Hi, sorry for being so long to come back to this.
On Tue, Mar 1, 2016 at 4:37 PM, Nicolas Gizardin <mode...@gmail.com
> I like Tagaini jisho very much but i would like to talk about homographs
> I'm surprised by the fact that, while using the kanji recognition training
> function, only one answer is accepted.
> For example, to the 後 kanji, i answered あと and the expected answer was のち.
Right, we are reaching the limits of the training feature here. There
are 3 different vocabulary entries that are written "後":
Since they all mean the same thing, discerning between them is not
easy. In the case of the latter, I think it should just not come
during the reading practice, since its kanji writing is an alternate
(i.e. not main) writing. So this as least can be fixed. But there will
still be an ambiguity between "あと" and "ご".
> Another example is: 今日は.
> I wrote きょうは but the expected answer was こんにちは which seems weird to me.
Here there should be no ambiguity, since the meaning part of the entry
is visible. So you know that you are suppose to write the entry that
means "Good day", and not "today".
> On some other occasion, the correct answer was 音読み and not 訓読み.
> What is the rule behind it ? Why is there only one possible answer even when
> there could have been several ?
It is not exactly a matter of onyomi or kunyomi. Reading practice is
done on vocabulary entries, not single kanji. It's just that some
vocabulary entries are made of one single character, so in this case
it is difficult to know which reading to use. Displaying the meaning
of the entry is supposed to lift the ambiguity, and works for the vast
majority of cases, but there will be exceptions like "後".
Still, you found one issue that can easily be fixed: alternate
readings should not be displayed. Fixing this will at least make sure
that のち (後) and こんにちは (今日は) don't appear anymore. I am fixing this at
Thanks! (and sorry again for the delayed reply >_<)