Is there any interest in a comprehensive Cangjie (input method) resource?

Hey Forumosa,

I’m in the middle of building some Cangjie learning software and I was wondering if it would be useful to the community.

First a link to a simple landing page (just a landing page for now! Read on for why it’s just a landing page):

Some background about Cangjie:

Cangjie, developed in the 1970s by Chu Bong-Foo (朱邦復) with input from NTU and the ROC, is a way of typing Chinese that relies on graphical decomposition rather than phonetic decomposition (my link has a simple example). That means that each key on your keyboard corresponds to a fragment of a character instead of a sound. You then use the keys on your keyboard to piece together a character fragment by fragment.

Why might you want that?

Well once you’re proficient with Cangjie, you can type faster than with phonetic methods because every Chinese character has a unique sequence of Cangjie keys so you don’t need to type a word and then hunt and peck the correct character from a drop-down menu of potentially hundreds of choices.

You can also type characters you don’t know how to pronounce as long as you have the character in front of you.

Finally, if you’re a Chinese learner, Cangjie also offers an additional opportunity to practice graphical recall of Chinese characters that can help considerably when it comes to general reading and writing of characters, since every time you type out a character you’re effectively practicing how to piece it together.

Then some background about my motivations:

I was looking a while ago for . There’s great software for learning how to touch type English (see e.g. Typing Club), but I couldn’t find any good ones for Cangjie, especially in English.

Unfortunately my post turned up no answers. And it didn’t seem like other forums had many answers either. Since programming is my livelihood, after trawling the web some more, looking at both English and Chinese sources, I decided I would have to make this software for myself if I wanted to use it.

As I began coding, I realized if I’m going to the trouble of making this software, I should see if others are interested.

So I’ve set up a website to canvas interest.

Basically the more interest I get, the more time and polish I’ll put into this to release to others. If people are willing to pay, then I’m going to treat this as a real production software project and try to make a definitive high quality Cangjie resource.

Assuming enough interest, the roadmap is a soup-to-nuts resource that teaches

  • The principles of Cangjie decomposition, i.e. in what order a character should be decomposed
  • The Cangjie primary forms (and muscle memory practice for them)
  • The Cangjie secondary forms (and muscle memory practice for them)
  • Graphical decomposition of many different characters into their primary and secondary forms
  • And has typing exercises (starting from single characters, then phrases, then sentences) in the same way that English touch typing is taught.

Unfortunately this represents a very large time investment and so I wanted to make sure this is something people actually want, before I went and sunk my time into it.

And I mention this on the website, but the email addresses I’m collecting to gauge interest won’t be used anywhere else for any other purpose and will of course be deleted on request.


As I mentioned, I had a friend who was an associate of Chu Ban-fu back in the 1980s when there was a Cangjie institute/commune here in Taitung. He wrote a book in English on the Cangjie system, just as Zhuyin swept to overwhelming popularity. (An orphanage here also trained teenagers in the method, for later jobs in computer entry.) It’s a good system, but, like QWERTYY, Zhuyin may be just too entrenched.

I’m interested, I’ll definitely take a look. It would be badass to be able to do it. My character recall for writing could use a lot of help too, if this could help, bonus.

I looked at an app, but I realize that this is just a wild fantasy now. My character recall is nowhere near good enough.