集 is a bit weird. The upper thing should be pointing to the left instead. This is also according to http://jisho.org/kanji/details/%E9%9B%86

You are correct that from an Japanese perspective this should be written right to left. What happens here i think is that the Radicalpart(隹) in Unicode is mapped to the Chinese version of it(隹) and that is the other way around. http://www.unicode.org/cgi-bin/GetUnihanData.pl?codepoint=隹 . I am currently struggling with the very same when converting 40.000 data files in an Japanese mainframe to IBM 939 code paged mainframe. There are in fact a lot of characters having similar issues. All of them are coming back to the Unicode font. If i use the same map points in an Japanese shift jis or EUC font I can get the correct look but not in Unicode. ( without going to the obsolete sections)

I have previsosly pointed out e.g the bone radical having similar issue. http://www.sp.cis.iwate-u.ac.jp/icampus/u/ihwr.jsp?k=骨 here the box in Japanese should be on the right side but in Chinese (and unicode) it is to the left.

This is in fact not bothering me at all but my customer here in Japan are in fact claiming that they are not able to understand the kanji when it looks like that:-)

How the character looks like actually depends on the font used to display them. In windows, using the unicode msmincho and msgothic fonts can display all those 隹and 骨 radicals properly. Currently, the 教科書体 is the preferred font for printing text books. An example is the 游教科書体M from http://shop.tokyo-shoseki.co.jp/shopap/feature/theme0043/ . There is no need to switch to Japanese shift jis or EUC font to get the correct look.

That is just the point, you need to use a Japanese font or a unicode font with CJK scripts telling what to show. There are no unique points in the Unicode chart for each of those. So, If I study Chinese in skritter I need to use a Chinese font and if I like to switch to study Japanese i need to switch to a Japanese font or I will still see the other "version". Not a big problem but if you like us have to handle CJK data in the same system with embedded fonts it becomes a minor issue.

It is possible to specify in the style sheet the font used to display a character. So, as long as a Skritterer has installed all the required fonts, the proper font will be shown when required. I don't think that this is a technical problem for Skritter. The only problem is that if they are willing and can allocate the resources to do it.

I was talking about the way it looks on the "writing pad" when you practice. So it seems that only the Chinese version shows up.

I think both directions are OK in both languages. Did I miss something?

You can see both variants here:
http://japanesekanji.nobody.jp/animal/bull.htm

http://www.bunka.go.jp/kokugo_nihongo/joho/kijun/naikaku/kanji/joyokanjisakuin/index.html crearly shows that there is only one possible variant.

Yeah, but that's just how the font looks like.

Here is another Japanese source with the dot pointing to the right:

http://kakijun.main.jp/page/furutori08200.html

Don't worry everyone, either way is fine!

I wouldnt have brought this up if not for my teacher who keeps insisting that the character is Chinese when it points to the right and Japanese when it points to the left...

I'm sure you wouldn't have, but I was raised by a Japanese teacher and she never mentioned that. So I asked her just now, and she said she writes that stroke from top left to bottom right.

My Chinese teachers also have contradictory opinions sometimes, so take what they say about such details with a grain of salt ;)

Even if both are correct, what is the most common version you will encounter in real life?

This is our CSS font stack:

*[lang="zh"] { font-family: "SimSun", "Microsoft YaHei", "Arial Unicode MS", "Arial Unicode", "serif"; }

*[lang="ja"] { font-family: "Meiryo", "MS Gothic", "Arial Unicode MS", "Arial Unicode", "serif"; }

*[lang="zh-Latin-pinyin"] { font-family: "Arial Unicode MS", "Arial Unicode", "Lucida Sans Unicode", sans-serif; }

Suggestions for changes to this stack are welcome. Your browser will use the first of these fonts that it finds on your computer, so if you don't have Meiryo or MS Gothic, it will fall back to Arial Unicode MS or Arial Unicode, or even a generic serif font. These would be the fonts which would lack language-specific character rendering.

@lechuan: For "printed" style fonts (Mincho, Gothic), "/" seems to be more common in Japanese than "\".

When I have seen it written (pen or brush) in Japanese it has been /, which is consistent with how　I have seen the radical written in other kanji. I have always written it / and no one has complained. Of course, now I use the computer, so I have to take what it gives me ^_^.

Changing the direction of those strokes wouldn't be too difficult. I agree with changing it to / as well. What's the vote though, all in favor?

I disregard from McGoats comments since he is just yet another uneducated child. For the rest of us, I think that the font today in Japanese mode do show the "correct" form in examples and decomposition etc. but the actual drawing of it do it the Chinese default way. However, the remaining strokes follows the Japanese stroke order rules. I have had several discussions around this and other oddities with my All Japanese team as well as my Japanese teachers and they all seems to agree that there are flexible rules around this. However, they all also tends to say that / is the preferred way. Nick, is it possible to have those "Japanese" rules reflected in the actual drawing of the kanji? I did see that 骨 have been changed for Japanese earlier so I guess it is possible. It is actually not important at all for me, just my usual curiosity that makes me dig in to those tiny issues. I really should spend less time looking in to why for every character i encounter.