In my published list for my textbook, lesson 12 (http://www.skritter.com/vocab/listsect?list=agVza3JpdHIWCxINVm9jYWJMaXN0SW5mbxiN6ocFDA&sect=11) I had the word 著急. Practicing today, it suddenly turned up as 着急. Fair enough, just another of the trad2simpl bugs. But what was strange was that looking at the published list in the vocabulary section, the entry said: "著急 unavailable :("!
What gives?

Because 着 and 著 are both used in simplified and in traditional, it breaks the abstraction we use where a simplified character is the base and it can have multiple traditional variants.

Here's what Wenlin says, after listing zhe, zháo, zhuó, and zhāo for 着 (with all of those plus zhù for 著):

"Originally 着 was just a different way of writing the character 著. Now 著 is mostly written only for the pronunciation zhù, and 着 is written for the other pronunciations; but sometimes 著 is still used rather than 着 among full form characters, regardless of the pronunciation.

"着 seems to have more pronunciations and meanings than any other Chinese character. Don't be discouraged. Even Chinese people can't always get it straight, especially the distinction between 着 zháo and 着 zhuó. For example, a friend of mine says 着陆 as zháolù though the dictionaries say zhuólù. The dictionaries disagree on whether 着 in 不着边际 ('not to the point') should be zháo or zhuó. On the other hand, the distinction between 着 zhe and 着 zháo really is important."

Complicated! We can correctly support the simplified usage by decoupling 着 and 著 and treating them as unrelated characters. If 著 were to be a traditional variant of 着, as we originally had it, then 著 couldn't be a simplified character on its own, and that'd mess up words with 著 as zhù (like 著作, 著名, and 显著).

It does break traditional usage, because words that use 著 in traditional but 着 in simplified will be converted to using 着. But as students of traditional are in a small minority, it's the best alternative. There are only a handful of these, where one traditional character maps to more than one simplified, and we probably saved months by not trying to support them.

Does this explanation make sense? Basically, I clobbered your word for the good of the land.

It seems to me based on searching tradition Chinese web pages (thanks, google) that in traditional Chinese
these are strictly variants. See http://dict.variants.moe.edu.tw/yitia/fra/fra03506.htm
for less common examples along with this one. I think that for once the simplification of Chinese characters in this case created a distinction that previously did not exist.

Variants is a complex topic, and one that has convinced me that the true nature of a Chinese character is not all that different than Roman characters (think of 'g', 'a', and 'y' font styles, 's' in colonial America). The most important thing is that others in modern context can recognize it.

Thank you for providing this rather complex description of the problem and your approach to a solution.

Although, I sincerely do not understand that you have chosen to have a single simplified character map to multiple traditional characters in your internal abstractions. Surely, it would be better to use the traditional characters as a base, as they provide a better distinction, and then just treat the fact that multiple trads get converted to the same simpl as an historical accident?

I even think that the seperate meanings of the same characters should also be represented individually. In my textbook, in the context of a particular dialogue, word or grammatical structure, only one of the multiple meanings are relevant, and frankly, the other meanings just confuse the issue. When I construct a vocabulary list for a particual chapter in my book, it would be great to just pick one of the meanings, and also maybe customize it. And, I should be able to construct any legal combination of characters in my own list, without one of those being "unavailable :("

I appreciate the fact that you have choosen a pragmatic solution to the problem of representing the multitude of characters and meanings, and I applaude the comprehensiveness of the site and the speed at which you have been able to implement it. But I do feel that this fundamental design issue may unduly "simplify" the complexities of the language, and ultimately restrict constructing a fully adequate learning experience.

Yes, it's an evolved system. Using simplified instead of traditional as a base goes back a long way--to a messed-up pair of fonts we had, actually.

But if we were starting over, would we design it with traditional characters as the base? Maybe, maybe not. It would allow more precision, but it would also be more expensive (in both programmer and CPU time) to do. It would also probably complicate things slightly for simplified users, in a similar way to how things are slightly complicated for traditional users. And it would not work as well for people learning both. Then again, I have torn out a lot of hair over supporting traditional variants, and perhaps we would have found better abstractions for some of those complications.

We originally had vocab-list-specific definitions, the way you describe. It turned out to be a big mistake that took a while to recover from. It hugely complicated the making of vocab lists, the storage of data, the loading of words, and the relation of words from different vocab sources. ChinesePod can do this because they have one source of content and tons of editors, but they don't get to leverage any relationships between lessons.

We are planning to let you have custom definitions for your words, because that'll help with definition practice (you can type part of what the word means to you, rather than what we have it defined as). They won't be per-lesson or per-list, but perhaps that will help?