---
title: Usage
in_menu: true
sort_info: 40
---
Usage
=====
Note: Before starting JLDrill, please make sure you have installed
the software listed in the [Requirements](Requirements.html) and
[Install](Install.html) pages. Failure to do so will cause problems.
On some platforms JLDrill is a little bit tricky to install, so
please follow the instructions carefully. If you have difficulty,
feel free to contact me.
Starting JLDrill is typically done from a command prompt
(from Windows: Start->Programs->Accessories->Command)
type:
jldrill
Note: In the Debian/Ubuntu version of JLDrill an icon to start JLDrill
will automatically be placed in your applications menu.
Note: When using the gem installed version of JLDrill you must make
sure that the gem bin directory is in your path. Debian (and Ubuntu)
put this in the /var/lib/gems/1.8/bin directory, where 1.8 is the
version number for ruby.
Note: If you have a source release of JLDrill, you may start it by
running AppRun in the main directory. This will run that version
of JLDrill even if you have another version installed somewhere else.
If you know the name of the file you want to drill then you can also
write type it. For example:
jldrill myDrill.jldrill
Main Window
-----------
This is what you see when you first start JLDrill.
There's not much to do until you load a drill file.
Selecting a Drill File
-----------------------
You may open a drill file by clicking the "Open" button on the tool
bar. You may also access this functionality by pressing Ctrl-O, or
selecting File/Open from the menu bar. After doing so you will be
presented with the file open dialog.
There are many drill files available. They include hiragana, katakana,
JLPT vocabulary levels 1-4 and grammar. JLDrill will automatically choose the
directory that contains the default drill files.
Note: The JLPT files have some incorrect
translations. Please cross reference the vocabulary with the dictionary
to ensure that they are correct (see below). Also, the grammar
drill is a work in progress and contains only about half
of the material from Tae Kim's Japanese Guide to Japanese Grammar.
Conducting the Quiz
-------------------
After loading the quiz file you will be presented with your first
question. This will be shown in the top pane of the main window.
The first time you see an item you will be shown the kanji
and reading for the word. You should try to guess what the
english meaning of the word is. When you think you know,
or you've given up, you click on "Check" ("Z" on the keyboard).
If you were right in your guess, you click on "Correct" ("C" on the
keyboard). Otherwise you click on "Incorrect" ("X" on the keyboard).
After replying you will be presented with a new question. In this manner
you are introduced to new material.
Measuring Your Progress
-----------------------
The status of the quiz is displayed at the bottom of the screen. When you
first load a quiz, it will look like this:
First, the * on the left hand side means that the quiz has not been
saved yet. As you progress through the quiz, it keeps track of your
progress. If you save the file (and hence your progress), the * will
disappear.
Following the modification indicator is a count of the items in
each set of the quiz. JLDrill divides the quiz into 3 sets: The
New set, The Review set and the Working set.
###The New Set
The new set contains all the items that you have never seen before.
When you first start a quiz, all of the items are put in the new
set. Then one item is chosen to be displayed to you. So in the
example above, there are 672 items in the new set. Since one is being
displayed, you know that there are a total of 673 items in the drill.
###The Review Set
After you memorize an item, it will be placed into the review set.
We have just started and no items have been memorized, so you can see
that the review set contains 0 items.
###The Working Set
The working set contains the items that you are currently memorizing.
The maximum number of items allowed in the working set depends on the drill.
Usually, it is set to 10 items, but you can adjust this number in the options.
The working set is further divided into 3 levels (levels 1, 2 and 3). When
you first start to learn an item, it is placed in level 1. As you learn it,
it advances to the second and third level. After you master the third level,
it is considered *memorized* and is moved into the review set. In the example
above, you can see that the count is 1,0,0 which means that there is one item
at level 1 and zero items at levels 2 and 3.
The next thing on the status is an indicator of the item currently
being drilled. You will notice it says "1 -> 5.0 days". The number
on the left hand side indicates where the currently drilled item
came from. In this case, it is a level 1 item from the working set.
If it were level 2, then the number would be 2. Level 3 would be 3.
Items from the review set have a "+" in front of them. They will
be described later.
After the level indicator is the *potential schedule* for the item.
After you memorize an item, JLDrill will put it in the review set and
schedule it for review sometime in the future. The potential schedule
is the approximate amount of time that JLDrill will wait before showing
the item to you again. While the item is in the working set, JLDrill
keeps track of the number of times you successfully guessed the right
answer. It uses this value to create the potential schedule. The
value displayed is the scheduled amount of time that will be used
if you continue to guess the item correctly until it is memorized.
More will be said about this later.
Finally there is an indicator showing whether or not JLDrill is
in "review mode" or "new item mode". Since we have just started,
JLDrill will select items to put into the working set from the
new set, rather than the review set. It displays "New Items" to
indicate this. More detail on "review mode" and "new item mode"
will be discussed later.
Learning and Reviewing
----------------------
JLDrill's strategy for introducing items into the quiz is at the
same time complicated to describe and easy to use. If you want to
understand the nitty gritty details, along with a rationale for
the choices made, please see [the strategy documentation](Strategy.html).
But for those who just want to use the program, a brief description
of what is going on follows.
As has been mentioned, when you first start a drill, all items are
put into the new set. Then one item is selected (either the first
item in the drill, or at random according to the options set) to
be quizzed. This item is placed into level 1 of the working set.
Once you quiz the item and select either "correct" or "incorrect",
a new item is selected from the new set and placed into level 1
of the working set. This continues until the working set is full.
The size of the working set is dependent upon the drill, but by
default it is 10 items. You can see that every time a new item
is introduced, the number of items in level 1 in the working set
increases.
Here you can see that the working set has 10 items in level 1.
At this point, JLDrill will stop adding new items to the working set.
Instead, it will randomly select an item from the working set for
you to try again. You can think of the working set like a deck of
cards. It draws one card and then presents it to you. Then it draws
another card. Once all 10 items in the working set have been presented,
it reshuffles the deck and starts again.
Note: Occasionally after the working set has been shuffled it will
present the same item it just presented, showing the same item
twice in a row. Since it is randomly selecting items, it happens
from time to time. If you want to wait a bit longer before seeing
this item, you can simply hit the "N" key to get the next item.
### Going to the Next Level
If you get an item correct two times in a row, it will be *promoted*
to the next level. The second time you get the item correct, you will
notice that the number of level 2 items will increase and the number
of level 1 items will decrease.
Here you can see that there are now 9 items at level 1 in the working set
and 1 item at level 2.
The next item you get to this item you will notice something different.
In the top pane (the question pane), you will notice that the quiz now
only shows you the kanji. You must now remember the meaning *and* the
reading for the item. What happens if you get this one wrong now?
Every time you get an item wrong, it is *demoted* back to level 1.
Also, the *potential schedule* for the item is reduced. This is what
the item looks like after we've gotten it wrong and drill it again.
Here you can see we've drilled all the other items and gotten back to
our demoted item. From the "1 --> 4.2 days" in the status bar, you
can see that it is again a level 1 item and that the potential schedule
is now 4.2 days rather than 5.0 days. Each time you get an item
incorrect, the potential schedule will decrease.
Eventually we will start to get our item correct as we learn it.
When it is correct twice in a row, it will be promoted to level
2. When it is correct twice more from there, it will be promoted
to level 3 (i.e., it has been guessed 4 times in a row correctly).
The quiz then looks like this.
You will notice that the question pane has the english meaning now.
The answer pane has the kanji and the reading. You should be able
to say the Japanese word for the item. You should also know what
kanji it is using. Finally, notice that the Status bar now
contains "3 --> 4.2 days." indicating that this item is level 3 and
it has a potential schedule of 4.2 days.
When you have gotten a level 3 item correct 2 times in a row, it
will be considered "memorized". At this point it will be promoted
to the review set. How it is treated in the review set will be
discussed below, but having been promoted it leaves a spot in
the working set. This spot will be filled by an item from the
new set, unless you are in review mode (see below).
In this way, you review items over and over again until they
are memorized. When you memorize one item in the working set,
it is replaced by a new item from the new set. Slowly, over
time, you will eventually memorize all the items from the
new set.
### Review Mode
Once an item has been memorized in the working set, it is
promoted to the review set. Unlike the working set, the
review set items have a definite order. They are ordered
by the amount of time they have waited in their
"potential schedule".
Imagine you have 2 items. Item A has a potential schedule
of 2 days. Item B has a potential schedule of 4 days.
When the items have just been inserted, they have waited
for 0% of their potential schedule. But after one day,
Item A has waited 50% of it's schedule, while Item B
has only waited 25% of it's schedule. Thus Item A will
be placed before Item B. It has waited for more of it's
potential schedule and thus has priority.
Items in the review set are displayed when you are in "review
mode". You automatically enter review mode when you first
start the application. At any other time, you can enter review
mode by pressing the "Review Mode" button at the top right of
the window.
Note: The application won't enter review mode unless you have
at least the same number of items in the review set as you do
in the working set (default 10). That's why you don't
enter review mode the first time you start a quiz.
When the application is in review mode it continues to operate
normally. You are continually tested on items from the working
set. However, when an item from the working set is promoted
into the review set it leaves a gap in the working set. In
review mode, this gap is filled by an item from the review set.
In this picture the item has been promoted into the review set.
Then we quit the application and restarted it later in the day.
As you can see from the status line we have 12 items in the review
set, so the application automatically entered review mode. Eventually
it displayed this item. Here it is displaying a kanji problem.
Inspecting the status line, you can see that part of it has changed.
It now reads "+1, Today --> 4.2 days" where the level and potential schedule
were displayed previously. In this case, "+1" refers to the number of
times this item has been reviewed in review mode. This is the first time,
so it is level 1. Every time we get the item right in a row, the level
increases.
"Today" refers to the last time this item was reviewed. In reality, I
just quit the application and restarted it, so the last time we reviewed
the item was today. If we reviewed it last yesterday, it would say
"Yesterday", otherwise it just prints the date. Finally we can see the
potential schedule if we get the item correct.
Before I make a discussion of scheduling, I would like to discuss the
last item on the status line. It says "0%". When you are in review
mode, the application keeps track of the percentage of items you got
right in the last 10 attempts. This was my first review item, so I am
at 0%
The percentage is a guide to show you when review mode will end. If
you can maintain a success rate of 90% in the last 10 attempts, the
program will go into "countdown mode". In this mode, if you manage
to keep your success rate at 90% or above for 9 more items, then
review mode will end and you will go back to "new items" mode.
You can tell when you are in countdown mode because the percentage will
look something like "90% - 5", meaning that you have answered 90%
of the last 10 questions correctly and that you have 5 more questions
before review mode ends.
The system will continue asking you questions from the review set
until you get one wrong, or review mode ends. If you get one wrong,
it is placed into the gap in the working set and quizzing resumes
as normal. If the review mode ends, then the system goes into
new item mode and a new item is selected for the gap.
Scheduling
----------
In most spaced repetition applications, each item has a fixed date
when it is scheduled to be reviewed. When that date has passed, the
item is reviewed. JLDrill doesn't work like that.
Instead it has a "potential schedule". The items are then sorted
by the potential schedule. You then review items until you seem
to be getting at least 90% of them correct. The reason for this is
simple.
When the items are scheduled, JLDrill puts a lot of effort to
ensure that the items you are most likely to forget are placed at
the front of the list. Items that you are least likely to forget
are placed at the back. As you review, you will notice that your
rate of success gets better and better. Once your rate of
success is 90% or more, then there is little point to reviewing
any more. You will be better off to learn new items at this point.
While you are in review mode, when you get an item correct,
it is given a potential schedule of up to twice the amount of time
since the last review. For example, if we haven't reviewed the
item for 2 days and then get the item correct, it will be given
a potential schedule of around 4 days. In this way the gap between
successful reviews will get longer and longer.
But it's important to realize that the "potential schedule" is
not the actual time that it will be reviewed. It just represents
the order in which the items are sorted. When you actually review
the item will depend on how well you remember the items. When you
get to 90% correct, you stop reviewing.
In other words, JLDrill automatically adjusts the scheduling for
you. If you have difficulty remembering items, then they will
show up more often. If you remember them easily, then they will
appear less often.
However, you may have noticed that the item we reviewed today
started with a potential schedule of 4.2 days, and after we
successfully reviewed it once, still had a schedule of 4.2 days.
I had only just promoted it a few minutes ago, so the total
schedule should have been twice that -- or only minutes.
What happened was that the
item was scheduled for 4.2 potential days. But I reviewed
all the items for those 4.2 days in just a couple of minutes.
So the program adjusts for this and creates the new scheduled
date from the old scheduled date. In this case, we have
a few minutes plus 4.2 days, which ends up being about 4.2 days
(once the numbers are rounded off).
Earlier I said that the new potential schedule is "up to"
twice the amount of time that had passed. In fact the
multiplier is higher for younger items than older items.
An item that has waited less than a day since being reviewed
last will have it's schedule multiplied by 2. This multiplier
is reduced over time so that items which had a previous
schedule of 180 days or more have a multiplier of 1. This
means that it is scheduled to wait the same amount of time
as it has already. Forgetting an old item is expensive since
you have to start all over again. So it makes sense to
review old items a little more often.
All this talk of scheduling may be confusing. If you don't
understand, don't worry about it. The only important point
is that the program is designed to keep your items roughly
in order of the chance you have of forgetting them. When you
know all the items in the list with 90% chance of success,
it gives you new items.
Problem Types
-------------
JLDrill allows you to select what kind of problem types you
want to review. There are three types of problems:
* Reading Problem. This is the type of problem you see
in the first level of the working set. The Kanji and
Reading are displayed. You must guess the Meaning.
* Kanji Problem. This is the type of problem you see
in the second level of the working set. The Kanji
is displayed and you must guess the Reading and the Meaning.
* Meaning Problem. This is the type of problem you see
in the third level of the working set. The Meaning
is displayed and you must guess the Kanji and Reading.
You may select which types of problems you wish to
quiz for items in the Review Set. By default JLDrill
will present both Kanji Problems and Meaning Problems.
You can change this from the Options window (see below).
If you select more than one type of problem to review,
JLDrill will schedule both types of problem independently.
In any one session, you will see an item only once.
But if you reload the file then you will often see the
item again with the other problem type.
For instance, if you have both Kanji Problems and
Meaning Problems enabled, then JLDrill will create
a schedule for both. On the first day you may see
the Meaning problem. If you get it correct, it will
be scheduled sometime in the future. But the next
time you run JLDrill you are likely to see the
item again with a Kanji Problem. Since the schedules
randomly vary by +-10%, the two types of problems
for the same item will eventually diverge and there
will be little relationship between them.
If you get either of the problems wrong, the schedule
for both problem types will be reset and you will have
to start again.
Options
-------
If you press O or select Quiz/Options from the menu, you can
alter a few options for the quiz.
There are four sections to the options. These options are
associated with the quiz that you have loaded and will be saved
with that quiz. You may change the options for each of your
quizes separately.
The first section is the dictionary. This sets the name of the
dictionary that you want to load. JLDrill is usually installed with
the Japanese edict dictionary. If you have a different dictionary,
for example the Chinese cedict dictionary, you can change it here.
Hitting the "Browse" button will allow you to select any file on
you system. Dictionaries must be in the EDict file format.
Normally the dictionary won't load every time you run JLDrill.
But if you want the dictionary to be loaded automatically
for this drill, then you can select the "Autoload?" button.
The next section is for the new set. Here you can select
whether or not you want new items to be selected in the order
they appear in the quiz, or in random order.
Next is the working set section. You may select the size of
the working set. Remember that the working set is the area
that the item appears in after it has been introduced and
before you have memorized it. By default it is set to 10,
but you can alter it from 1 to 30. Having more items means
that it will take more time between seeing the same item
again while memorizing it. If you are having a very easy time
memorizing the items, increase this. If you are having a hard
time, decrease it. Useful values are probably between 5 and 20.
The "Promote After" slider indicates how many times you must
get the item correct before it goes to the next level. Remember
that there are 3 levels in the working set. If you set this
to 1, you will have to remember the item 3 times in a row to
for it to be memorized and moved to the review set. If it
is at 2, then you need to get it right 6 times in a row.
You can set it up to 5, but that is rarely necessary. I
usually use 2 for vocabulary quizes and 1 for kanji or
grammar quizes.
The final section is the review set. Here you can select
what kinds of problems you want to review. For each
problem, you can review from the reading, the kanji or
the meaning. Each problem type has a separate schedule,
so if you have kanji and meaning problem set, each will
be scheduled. However, you will only see each problem
type once in a session for each question.
Finally you can set the rate at which you "forget" items.
This option is useful if you haven't reviewed your items for
a while and they have gotten very old. Once the "review rate"
of a problem exceeds the value you set here, the item will be
"forgotten". That means it will be removed from the review
set and placed in a separate forgotten set. When you
have finished reviewing, the items will be selected from
the forgotten set, rather than the new set.
The "review rate" is the multiple of time that has passed
compared to the ideal schedule. That means that if an item
has a schedule of 10 days and 20 days have passed, the
"review rate" is 2. When forgetting items you should set
the "Forget At" level to the highest value that allows
you to complete the review (get out of review mode).
Feel free to experiment. Reducing the value of this
slider will "unforget" items that you have forgotten,
so you do no harm by trying it.
Setting "Forget At" to 0.0 means that no items will be in
the forgotten set. If you have really gone a long time,
you can set the value so that all items are forgotten.
This is better than resetting the drill because the items
will be presented in an order based on how likely you are
to remember them. Hardest items will be presented first.
Saving Your Progress
--------------------
Remember to save your drill before you quit. That way it will pick up
where you left off last time. The first time you load a drill, you
might want to use the "Save As" entry from the menu bar (File/Save As, or
Ctrl-A) to save your drill under a different name. This will allow
you to keep the original file unchanged.
At any time you can save the current file (using File/Save from the
menu, Ctrl-S, or clicking on the Toolbar). If you quit the
application, or open up another drill, JLDrill will always ask you if
you want to save your file.
Indicators
----------
Above the top panel in the drill is a bank of indicators.
Drills derived from edict entries have markers. These indicators will
light up when the word has a specific marker. In the example, the
word is a "usually kana" and an "intransitive" verb. The markers that
JLDrill has indicators for are:
###Usually Kana
The item is usually written with kana rather than kanji. In most of the
drills I have kept the kanji. If you don't want to drill them, you can
edit the item and remove the kanji. Personally, I find it useful to
memorize words from the kanji, so I've left it in. A future version of
JLDrill will have an option to avoid drilling the kanji for items
that are usually kana.
###Humble
The item is part of humble keigo. This means that in polite speech you
will use this word when referring to yourself or your circle (family,
company, etc).
###Honourific
The item is is part of honourific keigo. This means that in polite speech
you will use this word when referring to people outside your circle
(your boss, people you are talking to, etc).
###Polite
The item is used in polite speech. It's neither humble nor honourific,
but used in polite situations.
###Suru Noun
This noun can be made into a verb by putting "suru" (する) at the end
of it.
###Intransitive
This verb does not take a direct object. That is you can't use wo (を) with
it. If this item is not lit up, usually you can assume that the verb
is transitive, but the whole transitive/intransitive thing in Edict is
still a little shaky. It is always best to look at usage examples.
###Differs
This item was not found in the reference dictionary. This only lights up
when the reference dictionary is loaded.
Note: Vocabulary items from Edict sometimes have several meanings. In
reality, each meaning should have a separate set of indicators. Currently,
all of the indicators for all of the meanings are set. So a word could
potentially be transitive for one meaning and intransitive for another.
Please be careful. A future version of JLDrill will address this issue.
Kanji/Kana Popup Information
----------------------------
JLDrill has a handy popup reference for kanji and kana characters.
Whenever you hover over a kanji or Kana character
you will get information for that character.
Here you can see an example of a kanji character. Although you can't
see it in the picture, the mouse is hovering over the character on
the right.
In the left hand side of the window you can see the stroke order information
for the character. You draw the stroke labeled "1" first. The numbers are
positioned next to the end of the line where you start drawing.
In the right hand side of the window there is a lot of information.
First is a list of all the readings for that character. In practice they
are all there, although I've found some strange readings that are
missing. Underneath the readings are a list of english meanings for the
character.
Underneath that is the grade that the character would be learned in
a Japanese school, and the number of strokes for the character.
Finally there is a list of all the radicals for the character, the Japanese
name for the character and the english meaning of each radical. The official
bushu is the first radical in the list and is annotated with a \*.
If you hover over a kana character you will only see something like this:
On the left you can see the stroke order diagram. On the right you will
see the character followed by the official representation in roman
characters. In this example the character is the katakana し which
is represented by the roman characters "shi".
Because the pronunciation of the character is not always intuitive
from the roman characters, I have included my own interpretation of
the sound. In this case it is "shee". I have also included a few
English words that contain the same sound -- she and sheep. Please
note that I am a Canadian with a Canadian English accent. While I
have tried to choose words that correspond with most accents I know
of, the sounds might be wrong for your English accent. In a future
release I will provide recordings of all the sounds spoken by a
native Japanese speaker.
Note: The stroke order diagram is only shown if you install the stroke order
font. Please see [the installation instructions](Install.html) for more information.
Timer
-----
Sometimes when you review an item you haven't seen in a while, it
takes some time to remember it. But when you are thinking it's easy
to lose track of time and you can't quite tell if you remembered it in
a reasonable time frame. Since it is important to remember vocabulary
quickly it's better to say that you got an item wrong even if it just
took you a long time to remember. But how do you know if you've taken
too long?
When you review items in the review set, JLDrill keeps track of how
long it took you to remember them. The first time you review an item,
it records that information. The second time you review it, JLDrill
starts a timer. If you take longer than the previous time to answer
the question, the question pane turns red. If you spend significantly
more time to remember the item after this point, it's a clue that
maybe you should mark it as incorrect.
This action has no effect on the item, or on the statistics in an of
itself. It's only intended to inform you when you spend more time
thinking than is usual for that item. Also, it only happen on the
second and subsequent times you review the item.
Statistics
----------
JLDrill keeps track of some statistics to help you understand how you
are doing. In truth, most of these statistics are used to help me understand
if the scheduling algorithm is working properly. But you may also find them
interesting.
You can obtain the statistics by pressing Alt-S, or by selecting
Drill->Statistics from the menu.
On the left side of the window you can see some potential schedule ranges.
The first column shows the distribution of item durations in those ranges.
For example I have 52 items with a potential schedule of less than 5 days,
62 with a potential schedule of 5-10 days, etc.
The next column shows how successful I was in quizzing items in these
ranges today. The third column shows how many items I quizzed in those
ranges. For example, I reviewed 23 items with a potential schedule of
less than 5 days. I ended up with a 78% success rate for these items.
Finally there is a column with various statistics. The first items
shows how may items I have reviewed so far today. In this case it was
30. In that time I also learned 5 items from the working set.
Review took me on average 13.8 seconds per item (they were grammar
items which take me a bit longer to remember) and it took 117.3
seconds for me to memorize each of the 5 items from the working set.
While reviewing, my total accuracy was 83%. This is pretty typical
for me.
The next item show the ratio of learning time to review time.
Probably it's a useless statistic.
Finally I show the "rate" of the current item and the average "rate"
of all the items so far. The rate is the ratio of actual wait time
to potential schedule. For example, if I waited 10 days to review something,
but it had a potential schedule of 5, then my rate would be 2.0.
You'll notice that my rate is pretty high here. I tend to remember
the grammar drills pretty easily, so my actual wait time is much
longer than the potential schedule. For the vocabulary drill that
I am studying now, my rate is usually about 0.7, meaning that I
find it difficult to remember things. The actual values aren't important,
but it is fun to look at it.
Don't be too concerned about these statistics. Again, they are used
for determining what needs to be changed in the review algorithm.
If you are using the application consistently every day and notice
problems, though, please contact me.
Non-Drill Features
------------------
**Reference Dictionary**. This feature is useful in case
you want to check that the translation in the drill is correct. It
allows you to cross reference the word against a dictionary
and edit it if necessary.
Note: The JLPT files are known to have some inappropriate or incorrect
translations. When working with words you haven't checked, please load the
reference dictionary.
You can load a reference dictionary by pressing Ctrl-D (or selecting
File/Load Reference Dictionary from the menu) The reference
dictionary distributed with JLDrill is the EDICT dictionary from
Monash University. It takes a few seconds to load the dictionary.
A window with a progress bar will be shown. You are able to continue
drilling while the dictionary is loading, though.
After the Reference Dictionary has been loaded, JLDrill will look up
every item as you encounter it. If the item can not be found in the
dictionary, the "Differs" indicator in the upper right will be
highlighted.
You can also look up words in the Reference Dictionary by hovering
over them. Usually JLDrill looks up kanji when hovering over them.
But you can change to word look up mode by pressing the space bar.
After doing so, when you hover over a word a popup will appear
showing you some likely definitions for the word.
The right side of the popup contains the candidates of words
ordered by how likely they are to be correct. You should scan through
the list to see which one is best for the context of the sentence.
JLDrill will also deinflect words and look them up in the dictionary.
In this example, the mouse is over the expression 果さなければならない,
which means "must achieve". JLDrill deinflects the word to
discover that it is the verb 果す. The left side of the popup shows
a large version of the kanji for the most likely word.
You can also hover over the words in the Example window if the
reference dictionary is loaded (see below).
**Tanaka Corpus**. The Tanaka Corpus is a collection of Japanese
sentences with English translations. Most of the common words
in the dictionary have example sentences in the Tanaka Corpus.
To load the corpus press Ctrl-E (or select File/Load Tanaka Corpus from
the menu). It takes a while to load, but when it is finished,
it opens up a new window. Whenever you are reviewing a word that
exists in the corpus, example sentences will be shown in the window.
Here is an example for the word 甘える.
There are two main sections in the window: Checked Examples and
Unchecked Examples. Words in the dictionary often have multiple
"senses". These are distinct meanings for the words. In the Checked
Examples section, there will be one example for each sense of the
word. In the example, 甘える has two senses: 1) to behave like
a spoiled child and 2) to depend on another's benevolence. You can
see there is one example for each of these senses in the Checked Examples
section.
The examples are taken right out of the Tanaka Corpus with only
a little bit of formatting. The number on the left is the reference
number for the example. Each example has a unique number. After that
you see the word as defined in the dictionary. Often you will see
a number in square brackets like \[1\] following the word. This
is the number of the sense for the example. To find the definition,
simply look at the meanings in the main JLDrill window. After the
sense there is often a description of the exact text used. In this
case \{甘えた} and \{甘えて}. Verbs, expecially, will be inflected so
this allows you to find the word in the example more easily. The
last two lines of the example show the Japanese and English text
for the example.
The Tanaka corpus is a work in progress. There are close to 200,000
sentences of varying quality. In general, all of the Japanese is
correct. However, the English can sometimes be awkward or even
completely unrelated to the Japanese. So you must be careful. All
of the checked examples have been checked for accuracy (although there
may still be a few errors around).
The Unchecked Examples section shows examples that have not been
checked for accuracy. Again, the Japanese usage is almost universally
correct, but the English translation may be incorrect. They can
be very useful as long as this caveat is kept in mind. Since
the Unchecked examples can contain many, many examples they are
sorted by sense.
Because the examples can be difficult to read, you can hover
over any of the Japanese to either look up the kanji or look
up the word in the reference dictionary. To switch between
looking up kanji or words, press the space bar while the
Examples window is focused. At the moment, there is no indication
what mode you are in, but by hovering over a word you will
notice easily.
Note: You must load the reference dictionary (Ctrl-D) in order
to look up words by hovering over them. Please see the Reference
Dictionary section (above) for more information.
**Edit** item. Pressing E (or selecting Vocab/Edit
from the menu) will allow you to edit the current item.
In this mode you can modify any of the fields.
If the dictionary has been loaded you will also see a list of words
from the dictionary that start with the characters in the reading.
Double clicking on one of them will copy its information into the
fields.
When you are done, press the Set button and the item will be
changed in the drill. This will also close the Edit window.
If you wish to close the window without setting the item,
simply press the close button in the top right of the window.
Sometimes you will edit an item but realize that you haven't loaded
the dictionary. You can load the dictionary by pressing Ctrl-D. When
it is finished loading, it will automatically search for the item.
Note: If you continue with the quiz without closing the Edit
window, it will update itself with the new items as you
get them. Be careful though because if you have edited the
data without pressing the "Set" button, the information will
be overwritten.
While the mouse driven interface is helpful, this window is
meant to be used most times with the keyboard. When you
first enter the dialog, JLDrill will search for the item.
Normally it will find several choices that are close. However,
it will highlight the one that is closest to the item you have.
If JLDrill has chosen the correct one, you can simply press
the Enter key. If not, you can use the up and down arrow
to highlight the item you want. Finally you can press the
Enter key to put the information into the fields. After
that, the "Set" button will be focussed (it's hard to see
sometimes). At this point, press Enter and the item will
be set.
Pressing the Esc button will cancel the editing process.
Occasionally it's useful to see what your changes will look
like (or to check the kanji information). In this case you
can press the "Preview" button. It will take the information
in the upper form and show what it would look like.
The information is displayed in the main window.
Notice how the background on the main window turns purple.
This is to show you that the item doesn't actually exist;
it's merely a preview of what it would look like. In this
state, pressing the Check, Incorrect and Correct buttons
do nothing. When you are finished editing the item,
the original problem (possibly edited) will be displayed.
**Add** item. Pressing "A" (or selecting Vocab/Add from the menu)
will allow you to add a new item to the drill.
It adds it to the end of the new set. This feature works exactly the
same as the Edit feature except that there is an "Add" button rather
than a "Set" button.
When you first start adding an item the window is empty.
You can edit the fields as you desire. If you have loaded the
dictionary you can also search for items by pressing the "Search"
button. This works by searching for the reading only. As with
the edit window, if you haven't loaded the dictionary, you can
press Ctrl-D in the main window without closing the Add window.
When the dictionary is finished loading, it will search for
the item automatically.
You can then simply double click on the item that you want and it will
copy itself into the fields. When you click "Add" the item will be
added to the quiz at the end of the New Set.
When you press add, the Add window will not close. This is so that
you can add several items. To close the window, press the
close box in the upper right corner of the window. After an item
has been added, the main window will display it.
Notice that the background is blue. This is to distinguish it
from an edit preview (purple) or a regular quiz item (yellow).
This special preview is intended to help you keep track of
what you are doing. Like the edit preview you will not be able
to use the Check, Incorrect and Correct functions. If you wish
to return to your active quiz, simply press the Next button,
or press "N".
As with the edit window, this window has a useful workflow
when using the keyboard. When the window first opens, the
focus will be placed in the Reading field. If you type the
first few characters of the reading and press Enter, the item
will be searched in the dictionary. The focus will move
to the table, where you can use the up and down arrows
to highlight the correct entry, and finally Enter to copy
the information into the fields. This will move the focus
to the "Add" button. Pressing Enter again will add the item
and return you to the Reading field.
Sometimes you will get far too many items in the search table to find
the item you want. In this case you can help locate it by entering
information in the the other fields and then pressing Ctrl-S (or
hitting the Search Button). Note that the search is always done on
the Reading field. The information in the other fields is only
used to try to jump the cursor close to the correct item.
Finally, as with the edit window, the add window allows you to
preview your item. If you press the "Preview" button, it will
display the item as you have created it so far. Just like the
edit preview, the background will be purple and you will not
be able to use the Check, Incorrect or Correct buttons. If you
cancel the Add, the problem that was displayed before showing
the preview will be shown. If you add the item, then the item
will be shown with a blue background.
**Delete item**. Pressing "D" or selecting "Delete..." from the
Vocab menu will delete the currently displayed item. JLDrill
will first ask to verify whether or not you really want to
delete the item. Be careful since once you delete it you
can't get it back.
**Learn item**. Pressing "L" or selecting "Learn" from the
Drill menu will put the current item in the review set
(or mark it as correct if it is already in the review set).
This is useful if you have a drill that contains items that
you already know well and don't want to keep going over it
in the working set.
**Show all**. Will show a table of all the words in the
drill.
This window allows you to view and modify your quiz as a whole.
The current item in the main window will be highlighted in the
list. Using the arrow keys or the mouse, you can select
another item.
If you wish to display that item, press the "Preview" button
or press the "P" key. This will show the item in the main window
with a blue background.
If you wish to edit the item, you can activate the item in
the list (double-click, or the "Enter" key), or you can
press the "Edit" button.
You can also change the original order of the items in the
quiz. To do so, select an item in the list and press
the up or down arrows along with the "Shift" key. Or you can
use the "Move Up" and "Move Down" buttons. Note that this
changes the original order, not the current schedule. In
other words it changes the introduction order of unseen items
if the drill isn't random.
Note: The key to understanding the purple and blue previews
is that purple previews show items that don't actually exist yet
(i.e., they are items you are currently editing). Blue items
show items that currently exist, but are not being quizzed
right now.
The "D" key will allow you to delete the currently highlighted
item. JLDrill will ask you if you really want to delete
the item. Be careful because once the item is deleted you
can't get it back.
**Reset**. You can reset the drill so
that all the words are Unseen again. Press Ctrl-R or select
Drill/Reset from the menu to do this.
**Info**. Show information about the current drill.
This is just information that the drill author wants to show. It
might contain copyright information or other interesting things.
Pressing Ctrl-I or selecting Drill/Info from the menu will display
this dialog.
**Create a New Drill**. There are two ways you can create
a new drill. The first is by selecting "New" from the "File"
menu or pressing Ctrl-N. After you have a new drill you can
add items by using the Add function (Drill->Add, or the "A" key).
You can also create a new drill by
importing any UTF8 or EUC encoded EDict file. Just open the file (JLDrill
will automatically recognize the file type) and then save it to
another name. Going up one directory from the quizes, you will find a
directory called "dict". In this directory you will find the files that
generated the drill. Additionally you will find the Mainichi Shinbun
Frequency files. Finally you will also find the whole edict file
in case you want to drill yourself on all 180,000 words in the
dictionary.
**Merging Two Drills**. You can merge two drills by loading one
of them and the appending the other using the File->Append menu
item or pressing Ctrl-P. The files can either be JLDrill files
or UTF8 or EUC encoded EDict files. If the files are JLDrill files, the
status of the items will be preserved. The items in the second file
will be add after items in the first file. Also items in the second
file that duplicate items in the first file will be discard without
warning.
That's it. Hope you find it useful. Please see the
[Planned Features page](Planned.html) to see what's coming up next.