[kde-edu]: kwordquitz proposal

[Peter Hedlund]

Hi all,

Vladimir contacted me some time ago and I asked that he take his 
suggestions to the list for a broader discussion.

Unfortunately I get very conflicting suggestions from different people. 
Some want to turn KWordQuiz into a kids program and others want it to 
have more or less all the complex features of KVocTrain.

I think there is a place for three levels of programs when it comes to 
vocabulary/terminology/flashcard learning. A kids level, an intermediate 
student/adult level, and an advanced language focused level. I think 
that KWQ meets and should meet the intermediate level and KVT should 
meet the advanced level. The new edu programs in 3.5 might meet the 
first level, I haven't really analyzed them myself.

Work is currently ongoing in creating a general purpose library in 
libkdeedu/kdeeducore for handling vocabulary data based on our kvtml 
format originally designed for KVT. The library is in place, but needs a 
lot of work. KWQ and KAnagram are ported to use the library, but at 
least KWQ needs work to be functional again. KVT will be ported to the 
library in the near future. The goal is to be able to use the same 
vocabulary in several programs without data loss. With the library and 
also a future implementation of the Model/View concept of Qt4 it should 
be easier to create interfaces of varying degree of complexity.

What Vladimir suggests (and others before him) is to separate the 
vocabulary data from statistical data in the form of number of times 
queried, number of correct answers, and so on. The need for this is 
probably greatest in a school or other educational setting.


Vlad wrote:
> Hello
> 
> I have looked into trunk code and got some thoughts.
> 
> Menu and tool bar is overburden with options that can be moved to 
> configuration dialog.
> 
> 1 I think we have to move Vocabulary menu to the kconfig dialog. In fact 
> Leitner configuration information belong to configuration and not to 
> actions.

I agree, this was also Scott Wheeler's original criticism when KWQ 
replaced FlashKard.

> 
> 2 I think we have to change mode/type menu/toolbars. I like more how 
> kvoctrain do it. I can select from menu the type of query I want: flash 
> card, multiply choice or other. Mode - sequential, random or Leitner 
> should be selected in kconfig dialog. I think people change this not 
> very often. Also the direction from foreign{->,<-,<->}local language can 
> be selected from kconfig.

I only partly agree. I think it is quite common to start learning in 
sequential order and then move on to a random order.

> 
> 3 I want more options to the multiply choice. as in the kvoctrain.

Will be done. The library and kvtml allow for this.

> 
> 4 I want to configure how many flash cards I want to see. I have big 
> file (~3000 cards) and want to see only 50-100 card in query. This can 
> be placed in kconfig.

This is really what a Leitner system is supposed to do. KVT also has 
this already. In general I think users will break their data down in 
smaller vocabularies, e.g. a chapter in a textbook.

> 
> 5 I want to have actual log file about all my answers. Based on this I 
> can do some research on statistics about teaching process. It seems 
> silly for me to loose this information.
> 
> 6 I want to separate the log and the Leitner box data from the cards 
> database. Actually the using of this the data is different. Look in the 
> mail list and you will see teachers messages how want to use this soft. 
> But they can not. Because of the data structure. I want to store 
> information about questioned in separate files in separate directories. 
> So I can organize class teaching.
> 
> 7 I do not like the storage of updatable information (like Leitner box) 
> in the XML file. XML file is not panacea for all kind of data storage. 
> In can contain statical complicate data structures, but work bad for 
> updates and transactions. I want SQL. Log data can be placed in CSV file 
> or the SQL.

As said above I can see the use for this. But I also see a great 
increase in complexity for the user in maintaining statistical data, 
e.g. resetting or adjusting the data.

If you already see advantages in KVT and the purpose in general is 
language learning, maybe KVT is a better program for these advanced 
features.

Thanks,
Peter