Wubi question

[Publius]

@imron,have you tried RIME? It's cross-platform and open-source. The MacOS version is called Squirrel. I'm using the Windows version. But it doesn't seem to support 无重码自动上字.

[imron]

I haven't.  I'll see if I have some time to play around with it.   What do you like about it compared to other 五笔 input methods?

[Publius]

Well, it's customizable, as in you have to edit a txt file to change the font, size, color, etc, lol. I don't know. Actually I was looking for a Jyutping input method... and it's more like a framework that supports multiple input methods.

[rossg]

Downloaded. Now I am free to type 牛 - arguably the most useful of all characters..  Thanks

[murrayjames]

@imron, do you still use 五笔? How did you memorize the keyboard layout? I want to try to learn it... but it seems daunting and I do not know where to start.

[imron]

Yes I do still use 五笔.

 

For memorising the keyboard layout, I used typing tutor software that unfortunately no longer exists (五笔快打).

 

The other typing tutor program I know of that still exists is 金山打字通, which does have an option for learning 五笔 but I don't know if it's any good.

 

9 hours ago, murrayjames said:

but it seems daunting and I do not know where to start.

Don't be daunted by it, and don't try to memorise keycodes, just try and find a typing tutor program and if you practice 15-20 minutes a day for about a month you should develop muscle memory such that your fingers know where the appropriate 'root' is for the shape that you want.

[hejmeddig]

@imron Out of curiosity, what's your reasoning for disliking Sogou input? Or were you just meaning the Sogou wubi input?

 

I have been using it since I started Chinese and have never had any problems with it, it seems to have a pretty smart recommendation algorithm.

[imron]

Just now, hejmeddig said:

what's your reasoning for disliking Sogou input?

Popup ads and data mining, for their pinyin input method.  The wubi input I also didn't like because the word database it used was missing many common double and triple character shortcuts (I forget what they were now, I just remember it was missing them and it was annoying).

[murrayjames]

Thank you Imron. I think I need to try a tutor program.

I read a long article called "Wubizixing: For Speakers of English." The article explained how 五笔 works, but not how to learn or memorize it. Then I read a few posts about 五笔 on 知乎. While there was some helpful advice, mostly everyone was recommending their preferred 口诀, and the 口诀s were nonsensical (to this foreigner, anyway).

 

Then I printed a picture of the 五笔 keyboard layout. Even with the picture in front of me, though, I often could not figure out how to decompose a character. Probably muscle-memory repetition through a tutor program would help with this.

 

One question about the character 你. The four-character code for 你 is WQIY. It seems that WQI spells the entire character. What is Y doing here?

 

EDIT: I found my answer. For Chinese characters with only three 编码, a fourth 识别码 can be added. The 识别码 is determined by the character’s final stroke. Y represents the downward-right sloping stroke on the right side of 小.

 

For me the attraction of 五笔 is not the increased efficiency, but the opportunity to strengthen my understanding of how Chinese characters are written. There are many characters that I can type using pinyin but cannot write by hand.

[imron]

7 hours ago, murrayjames said:

I read a long article called "Wubizixing: For Speakers of English."

This is a good article for providing an overview and understanding of the system.  It was also the first explanation of Wubi that I saw (all the Chinese people I'd seen using it would just say "it's too complicated for a foreigner to learn").

 

7 hours ago, murrayjames said:

mostly everyone was recommending their preferred 口诀, and the 口诀s were nonsensical

Muscle memory from a typing tutor program is definitely the way to go.

 

7 hours ago, murrayjames said:

The four-character code for 你 is WQIY

Because of the isolation rule.  Each character technically needs 4 keys (although many have shortcuts).  WQI gets you 亻⼓ 小, so what to put for the 4th key?  Well, according to the isolation rule we need to look at the last stroke of the character (to determine which of the 5 'pen' regions to use) and also the character composition to know which key in the pen region to press.  The last stroke of 你 is the 丶, which puts it in the right-falling region (yuiop), and 你 is a left-right character (亻 on the left, 尔 on the right) and the isolation rule states that for left-right characters we should use the first key in that region, which is y (top-bottom characters take the second key, and everything else uses the third key), hence WQIY.

 

That being said, I don't think I have *ever* used WQIY to type 你.  All IMEs I have used will allow you to type this with WQ<space>. 

 

8 hours ago, murrayjames said:

For me the attraction of 五笔 is not the increased efficiency, but the opportunity to strengthen my understanding of how Chinese characters are written

For foreigners, I think this is the primary benefit of learning and using Wubi, and why I strongly recommend people to learn it.  Although sometimes the breakdown of characters requires you to stretch your imagination somewhat (see halfway through this post for some examples).

 

What IME are you using by the way?

[murrayjames]

I added the default 五笔 keyboard in Windows 10. One neat function: if you type in pinyin, the IME will show you the 五笔 codes for common characters with that spelling.

[imron]

Yes, that's a very common feature of modern Wubi IMEs, but be warned, it can become somewhat of a crutch.  Best bet is to practice with a typing program, then turn the feature off (you're learning Wubi to avoid using pinyin remember, and the characters you need pinyin for are the very characters that need the most Wubi attention given to them).

 

I've not used the native Windows 10 五笔 IME, but the one I like to use on windows is 极点五笔。

[murrayjames]

Question: Why is the 字根 “小” in the same 字根区 as 三点水? 三点水 belongs to the “捺” 字根区. But 小 should belong to the “竖钩” 字根区, no?  The strokes are 竖钩, 撇, 捺.

[murrayjames]

A second question. The 五笔 code for 今 is WYNB. How would I know that final hook stroke is N? The stroke does not look like any of the 字根 for 51N.

 

EDIT: And a third question arises. I must be misunderstanding a basic 五笔 principle. 提手旁 is 32R, in the “撇” 字根区. But none of the strokes in 提手旁 are 撇! What am I missing here? (EDIT OF MY EDIT: I may have figured it out. 提手旁 is derived from 手, and the first stroke of 手 is 撇. Is that it?)

[imron]

17 hours ago, murrayjames said:

Why is the 字根 “小” in the same 字根区 as 三点水?

Because.   

 

As mentioned previously (either here or in one of the other threads) wubi takes a bit of leeway with its rules.  Later versions of Wubi (e.g. Wubi 98) were designed to try and clean up some of these inconsistencies but it never took off (and in any event Wubi98 also had inconsistencies of its own).

 

So what you describe is the ideal logic for assigning a shape to a region, but there is also the practical aspect of trying to minimise conflicts between different characters using the same keycodes, so sometimes a shape is moved on to a different key where it 'sort of' fits in.   小 is one example, so is 扌and there are quite a few more that you'll just have to accept as a quirk and move on.

 

16 hours ago, murrayjames said:

How would I know that final hook stroke is N?

Practice.

 

With enough practice you'll get a feel for shapes and your fingers will know where to go for different shapes.  You can try and reason about it all you like in your head, but it will be difficult to reason about because there are inconsistencies.  As you practice, you'll get a feel for the main ideas and logic that make up the Wubi, and also get a feel (and learn to accept) the inconsistencies.

 

16 hours ago, murrayjames said:

What am I missing here?

Wubi is mostly logical, but also has inconsistencies.

 

[murrayjames]

4 hours ago, imron said:

Wubi is mostly logical, but also has inconsistencies.

 

Yes. My mistake was thinking there was some rigorous logic by which I could derive the 五笔 code for every character. Just now I tried typing a few sentences, working out each character one 字根 at a time. By the end of those few sentences, I could guess the appropriate letter for “口” or “马” without looking at my cheat sheet. It was still slow going, though.

 

The 字根 logic also goes deeper than I first thought. Repeating the representative stroke of each 字根区 corresponds with an increase in the 字根位. The four roots 丶冫氵灬  in the “捺” 字根区 is an obvious example of this. Also, the 横，竖，撇，捺 pattern of the 字根区 is repeated at the 字根位 level. For example, most of the 字根 in 11G begin 横 and 横; many 字根 in 12F begin 横 and 竖, and most of the 字根 in 13D begin 横 and 撇. 

 

I have yet another question. The four-letter code for 我 is TRNT. 我 is a character with five roots. Why is the last letter in TRNT T (the penultimate 撇 stroke), and not Y (the final 捺 stroke)?

[imron]

1 hour ago, murrayjames said:

Why is the last letter in TRNT T (the penultimate 撇 stroke), and not Y (the final 捺 stroke)?

Because according to some, the last stroke is 撇 not 捺.

 

In any event, I more or less only ever use Q to type 我, or sometimes TR if I'm typing 我们 (with the full code being TRWU).

[大块头]

I've been having a lot of fun learning this system. A question for the 五笔巫师:

 

Any idea why 属 is NTKY and not NTJM?

 

[imron]

According to the rules of wubi, if a character has more than 4 roots, the code for the character is the first 3 roots + the final root. 
 

属 has more than 4 roots and the last root is on the Y key (as noted by the image you posted), hence NTKY. 

[大块头]

Thank you. When I looked at this character though I thought it would only have four roots: 尸、撇、虫、冂