Version 1.14.0 is now available - Please download it at the top post
This version adds support for the full Niai Visually Similar Kanji. Previously Item Inspector supported Lars Yencken visually similar kanji which is a subset of it.
To display the Niai visually similar kanji just select a popup element.
You may configure which Niai kanji show up with these settings in the Content tab of the page.
The threshold is a real number between 0 and 1 that indicates how similar you want the kanji to be. 0 returns all kanjis and higher the number the fewer but more similar the kanji that are shown.
Niai uses six data sources, three main sources and three alternate sources. When unchecked only the main sources are used. When checked both the main and alternate sources are used.
Don’t ask me what is the difference between main and alternate. All I know is @acm2010 script Niai Visually Similar script works this way and I made sure Item Inspector displays the kanji identically to this script. This way people using both scripts won’t see discrepancies.
This version also reorganizes the Leech Training emulation. The reason is to make the feature easier to discover and use. The drawback is it departs from how things were done in version 1.13.0. @yukinet this may affect you.
Now Leech Training emulation no longer uses the random Selection button. It now has its own button.
Leech Training emulation now has its own section in the settings. It no longer share settings with ordinary Random Selection and popup configuration of LY visually similar kanji.
These settings are the same as they were in version 1.13.0:
- Size of Training Selection: Self-explanatory, was the Random Selection setting.
- Only Once Before Next Review
- Add Similar Items
You can now select the source of visually similar data used for adding visually similar items with the Type of Similarity option.
There is a progression where Wanikani has few similar items but they are highly similar while Niai has the most similar items but similarity is ranging from good to non-existant depending on the item. Lars Yencken lies in the middle, both in terms of number of similar items and quality of similarity.
There are only so many similar items and when we try to expand their numbers the quality of the similarity drops.
Now the visual similarity threshold for Lars Yencken and Niai is separate in Leech Training from the threshold for the popup elements. Same for the use of Niai alternate source. You can configure your popups to behave differently from the Leech Training.
There has been a reorganization of how Item Inspector download the non-Wanikani data it uses. I have opted for a “lazy” approach where the data is not downloaded until it is used for the first time. This makes the initial startup a bit faster and the latency from the incremental download that occur later is very acceptable.
The defaults settings for new tables are arranged not to cause data downloads when the same information is never used in other tables. This way people who are consciously avoiding the data downloads in their table settings won’t have to chase the settings every time they add a table.