A detailed explanation of how to create vocabulary files - containing both simple and complex questions. Wall of text incomming...
Let's start with the easy stuff - the main screen of the program.
The file parameters let you select one more CSV files, just select them all and click "open".
On the right side of the input file parameter you can explicitely select which value separator you used when saving the data file - usually it should not be necessary to do so, the program will simply try all separators and see if the results look good.
Below the files are the questions the program will use. Right now the simple questions are always in english. With the complex questions you can define your own text.
On the right of the big process button are a few options you can set - the first, "delete temporary files", is already set. During processing the program will create several files that are deleted afterwards.
If you'd like to see them, uncheck this box. The files contain the raw parts of the final vocabulary file - they can be used with Holghoj, though it is not recommended.
By default, gherwI' will turn all "resources"(i.e. words in the resource file) into questions. If you've supplied a complex question file and don't want it mixed with regular question, you can turn it off with the second checkbox.
The third checkbox only is usable together with the second. When using complex questions you can define all possible answers, and if you don't want simple questions, no additional resource is going to be used by Holghoj, so leaving them in is going to waste space.
However, be aware that Holghoj uses the seemingly unused words to provide fake answers - so if your complex questions only provide two additional answers, Holghoj would find a third from the unused resources.
If you remove all unused resources, Holghoj will probably only show the default "no picture found" image for undefined fake answers.
The last checkbox causes gherwI' to produce a HGM file instead of a HGMZ file. There shouldn't really be a reason for this, but if HGMZ files don't work, or you have a large collection of media files and created several smaller question files, it may be worth thinking about HGM files.
Resource file (simple questions)
Let's move on to the input files. You have a template XLS file, which can be opened in Excel or OpenOffice (and other programs).
I've tried to make the file self-explanatory, with help of tooltips - most headings will show additional information if you hover above them.
As you see, lines starting with a # are comments - feel free to place them wherever you feel like it.
The paths are taken relative from where the program is called.
Grammar type has a drop-down list of available options - I recommend using it, the program needs this to correctly process the entries.
Each line represents a single word or idea, with several facets of it. Technically, each line can become up to four "resources", the ID you supplied will be appended by a single letter indicating the type (T for Text, A for Audio, V for Video and P for Picture).
Text resources contain both languages, media resources have no language.
Holghoj will know that these represent the same word and not display two facets of the same word at the same time.
The rewards are shown as soon as the points threshold is reached - simple questions always give 100 points. If you want to finetune the points, you have to use complex questions.
You may specify a media type for the reward, if you don't, the first available will be taken.
Note that the first column, "resource only" is useful for having a media resource that is only shown as reward, never as question or answer.
With complex questions you have to provide all information yourself, gherwI' will not be able to help you.
Again, lines with # are comments, and most headings have tooltips with more information
Here too you can specify language A and B. You can also set if the questions are supposed to be shown in the way they're in the file, or randomized. Remove the "FixedSorting" if you want them to be randomized.
An important difference from the resources file is that it has a special line that can be repeated: "QuestBlock".
In this line you define the category of the questions - this is used to get fitting fake answers - as well as the grammar type of the word you're looking for. This too is used to find fitting fake answers.
Also here are the points for best and good answer - that way you can finetune how many questions have to be correctly answered to get a reward.
The lines below are the actual questions. It's probably easiest to list the columns:
The benefit of these complex questions is that you can control exactly what is displayed - you might, for example, have an audio file of a sentence as hint, and have the program show other resources each containing a part of the sentence and ask for the subject, object, or another grammatical part of it.
- ID - they're not really important to you, so just have them count up
- Hint ID - this is the resource that is shown as a hint, accompanying the question text
- Hint language - for resources with language (i.e. text) you can specify which language is to be used
- Best Answer ID - obviously required, this is the most correct answer
- Good Answer ID - optional, this would be the second-best answer which must be allowed as the single answer as well.
- Question Text - This is shown next to the hint. If you leave it empty, the last question will be repeated.
- Fake Answer IDs - you can specify that some resources are shown as answers. If you leave it empty or don't provide enough (3 or 2), Holghoj will add the rest itself.
- Forbidden Answer IDs - if you know that some resources are misleading (i.e. could be considered correct but aren't), you can specify them here to prevent Holghoj from using them as fake answers.
Then repeat the process with the same sentence but move to another part.
I think that's about it. Any questions? Mail me.