Howdy Y'all,

Nick and I are about to commence our bicentennial painstaking update of the character map, or what the layman has termed, "the differences between a character's simplified form, traditional form, and other variant forms." In an ideal Skritter, any one character should be linked to all of its variants, and no character is linked to any incorrect variants.

We want to make sure that we remain as accurate as possible during this process, and cover as many errors in the system as we can so we don't have to do this again. If you can think of any variant problems that you have not yet reported to us, please post them here. If you have time to provide us with a source, it would be much appreciated, as it would save us quite a bit of time.

With Much Love and Post-turkey Bloatedness,

The Skritter Team

What do you mean by "linked to all of its variants"?
I thought skitter only support two forms, traditional and simplified and if there are variants in each of those then the most common form will be picked. These are also limit by code point in Unicode.

Are you saying that skitter will support more then two variant in the future?

My understandig is as 戴金霸 said.
Remember this thread?

http://www.skritter.com/forum/topic?id=233871398&comments=8

Nick says there can be only one variant per Unicode point. Is that going to change?

If yes, as I said in the thread above, I’d love to see the traditional forms of 鬼, 差, 象 etc. with the extra stroke :)

It’d be great if we could choose the variants for each character!

We're talking about linking up different Unicode points as separate traditional versions of the same simplified characters.

Here are a few examples of what we mean:

鬼: the same character in simplified and traditional. No mapping.
国 -> 國: 国 is simplified, 國 is traditional. One-to-one mapping.
说 -> 說, 説: simplified 说 can either be 說 or the weird variant 説 in traditional.
发 -> 發, 髮: simplified 发 used to represent two distinct traditional characters 發 and 髮.
了 -> 了, 瞭: simplified 了 is also a traditional form of itself, with 瞭 as a separate traditional form.
只 -> 只, 隻, 祇: ... and they get more complicated.
台 -> 台, 檯, 臺, 颱: ... and more complicated.
并 -> 并, 並, 併, 幷, 竝: ... whoa!

So we have a bunch of corrections built up to adjust this map, like:

"迴 has been simplified into 回, so the simplified form of 輪迴 is 轮回."

回 and 迴 are currently not linked to anything, so the correction will be to add "回 -> 迴" to the map.

If there are more corrections like this, where the simplified/traditional mapping of different characters is off, we'd love to know now, since it takes forever to go through this modification process.

How about 着 → 著?

Can't do that one, unfortunately. From http://www.skritter.com/faq#char_zhe :

"These two characters are complicated. Both are used in both simplified and traditional, but in different ways. Because it's too difficult to support a mapping of one traditional character to two simplified characters, we choose to support the correct simplified usage of 著 for zhù and 着 for zhe, zhāo, zháo, and zhuó, even though in traditional, 著 is used for some of these, like zhe. Sorry about this, but sometimes for traditional learners, Skritter will incorrectly display 着 instead of 著."

OK, but I don’t quite get what exactly you guys are gonna change about the mapping system.

I figure you guys would have implemented this if it were feasible, but instead of mapping from character to character, how about mapping from word to word?

We are making individual corrections to the existing mapping system. We are not changing the system itself, which is too complicated by half already.

I am still not quite clear what this mean to the user. Would this mean if we added one of the variants to a list we would automatically get all the variants?

No; you only get the variant that you added, since you can choose which traditional variant you mean when adding a single character to the list (from the ones allowed in our mapping). We still choose which variant goes in which multi-char word, so that's not user-customizable.

For the character, 咸, the traditional definition should be 'all' and not 'salty'.

What if you are originally studying simplified, and then start adding all your lists from the beginning, studying traditional? Will Skritter then choose on its own which variant it want you to study? I'm planning on doing that somewhere down the line.

I always wondered about 为/為/爲, I was once told that the second one was used in Taiwan while the last one is prominent down in Singapore or something like that. So 所為 and 所爲 are both words. Do you still choose which variant all the users are to study, or are both variants possible?

The variant is chosen at the time of list creation, not at the time of studying a list. In other words, whether it's 為 or 爲 depends on the list, not on the user studying the list. We use 為 in all of our standard lists, and in every multi-character word, it has to be 為 (so always 因為 and never 因爲), but you can create your own list with a single 爲 in it.

This seems mainly about Chinese, but getting back to the originial Evan message, there are some issues with some Japanese characters. The main ones are those with the "old bird" radical, such as 潅ぐ、雁がね，贋、罹る、and others; the second stroke on the top of the radical (per Denshi Jisho and Tatoeba and my spy in Tokyo) always goes Rt to Lt, rather than Lt to Rt, as in 住． There was a forum on that a while ago.

Another one is the 骨　radical, which (same sources) as in 滑る、has the little box to the right in the upper part, rather than the left (as it appears here). In the keyboard input dropdown list it is to the right (correct) but it oddly switches to the left when one hits the "enter" key (error in Unicode?). Anyway, in actual, brushed Japanese, the box is to the right (again as in Jisho and Tatoeba).

Another issue with Japanese is that although the bamboo radical ⺮ has the sixth stroke going down toward the right in typed characters (and the way it is in Chinese), it is written with that stroke going down towards the left. Japanese teachers will mark characters including the bamboo radical wrong if they are written with that stroke going down towards the right.

For example,
http://kakijun.main.jp/page/12174200.html

Skritter does not accept the stroke as it is supposed to be in writing, so I have to write it incorrectly for it to be accepted.

To submit corrections for stroke orders, find the character in my words or when studying, open the word popup, and click 'submit correction for x' at the bottom. It's a brand new system for handling all sorts of fixes and it sends them where Jeremy and Evan will be looking for them and handling them regularly. I'm not sure they're reading this forum but I'll make sure these two get seen.

I've made some fixes:

Characters containing 隹 :
潅、雁、贋、罹、隹、隼、准、隻、惟、崔、碓、截、蕉、樵、應、擢、雖、燿、雛、灌、灘、讐、攫、軈

Characters containing ⺮ :
等、竺、竿、笈、笑、笠、笹、笙、符、笛、第、笥、筑、筈、筒、策、筋、筆、答、筐、筏、節、箔、箇、箋、管、算、箒、箆、篇、箸、範、箱、箭、箴、篆、箪、篤、築、篭、篠、櫛、簡、簾、簿、簸、纂、籍、籠、籤

Characters containing 骨 :
猾、髑

If there are any that were missed that strike the mind, you can post them here, or through the submit correction like Scott said.

@ジェレミー 曜

曜 has been fixed too, thanks!

Awesome! Just noticed the correction to the bamboo radical when I was writing. Cheers!

I’d like to add 干貝 (traditional), but Skritter only accepts either 乾貝 or 干贝.

干貝 means ‘scallop’, not ‘dried scallop’.

Also, can you use the same 麼 variant in 不怎麼 as in 怎麼?

Scallop is 扇貝 or 鮮貝. According to the dictionary provided by the Taiwan moe, 干貝 - 以蚌類扇貝上的肉柱晒乾製成的食品。或作「乾貝」。 Also in the same dictionary, one of the definitions of 干 is 經脫水加工製成的乾燥食品。通「乾」。如：「筍干」、「豆腐干」。

Oh really? I always seem to get fresh scallops when I order 干貝, so I asked my teacher about it. Maybe it’s 口語, then?

I sent feedback about this, but I figured I put it in this thread, too.

The traditional variants of 周二、周四、周六 should be 週二、週四、週六. Petty words, I know :p

The words 上周、下周、黃金周、周记 and 每周 also have the same problem. But I dare not say that they are wrong because the mainland government has decreed that 周 should also be used as the traditional form of 周 to mean “week”. So, they are the correct traditional variants when used in the mainland. http://zh.wikisource.org/zh/第一批异体字整理表

Then 週到 and 不週到 are wrong. They should be 周到 and 不周到.

Also, can someone please check the definition of 周天. Although example sentences can be found to show that 周天 means Sunday, the dictionaries do not seem to agree.

Anything new on this topic?

I've fixed the traditional writing for 干貝.

As for 周，this is a tough one to resolve. Right now I'm thinking of making sure each Taiwan MoE listing containing 週 is accounted for, and then having the remaining entries use plain old 周. The only issue is that the MoE's inclusion of either 周 or 週 seems fairly arbitrary, and it would be nice to use just one variant for all week-related stuff, and the other for dynastic stuff, etc.

There's no perfect option, but for now I have all days of the week using 週, following MoE's precedent of 週日, and will get to the specific 週 entries as well.

PTS and Schnabelhund, I'd like to hear what you have to say on the matter before this is finalized.

Thanks, Evan! 周四 doesn't seem to be fixed yet, though.
Also, as pts said, we need to clarify what 周天 means. Maybe there is a distinction between 周天 and 週天 in traditional writing.

Not 100% certain if this is the right place, but:

台幣 is only available as 臺幣. The former is usually used here.

服務臺 and 櫃臺 usually use the simplified version (台), at least in Taiwan.