Over the last few weeks the stroke research page that I posted has attracted a lot of attention. And as a result, I have spent a considerable amount of time re-investigating stroke order decisions made more than two years previous. The results have convinced me that we need to change how we accept a number of radical strokes orders and I intend to do a run through of the database shortly to accomplish that change. Because the radicals under consideration range from obscure to very common, I wanted to present them to the community prior to making the database changes to ensure that I am not being rash in changing some of these. The radicals that are up for alteration are as follows:

门 - in hindsight I'm not sure why we thought that the first two strokes (the vertical and the dot) needed to be interchangeable. Upon looking at this again, I was unable to find evidence to support that decision. Arch Chinese, eStroke, MDBG, Nciku, Wenlin, and the official standard published by the central party in Chinese (现代汉语通用字笔顺规范, which I refer to as the "Paper dictionary" on the stroke research page) all agree that the dot comes first.

女 - when the 2 & 3 strokes are not combined, this has a unanimous strokes order: shù zhé, wān, héng (according to Arch Chinese, eStroke, Nciku, the Taiwanese MoE, Wenlin, and the Paper dictionary). However, this gets a little tricky because depending on the style of the character, this can either be written with the wān and héng as one stroke, or they can be written separately. I don't propose we change anything for characters that have the two strokes merged (or optionally merged), but for those characters intended to be written unambiguously with three strokes, this needs to follow the following order: shù zhé, wān, héng.

戈 - somehow I've overlooked codifying a stroke order for this, but in the push to make the stroke research page more accurate, I've done so. All of the sources save the Taiwanese MoE agree that the dot should come last. However, since the two strokes in question are adjacent and the Taiwanese MoE has clout, we are going to support both stroke orders, giving preference to the mainland standard of the dot coming last when animating the strokes.

忄 - since time immemorial, Skritter has permitted either the vertical or the second dot as correct in this radical. However, according to all the usual suspects (Arch Chinese, eStroke, MDBG, Nciku, and Wenlin), this radical has only one stroke order and that puts the vertical stroke last.

母 - there's confusion regarding the order of the interior strokes of this character. Ordinarily, we could rule out contradictory evidence based on the merit of the source, but here the sources are completely divided as to whether the interior dots or horizontal come first. My proposal is to make all the interior strokes interchangeable.

飛 - despite common sense, I spent a while digging around and found that the center vertical for this one should always come last. Since this radical is so rarely used, I figured codifying that new order wouldn't be a big deal.

黽 - at issue here are the strokes 7-10 and particularly, whether we allow stroke 8 to be a single L-shaped stroke, or whether we accept it only as two strokes (one vertical, one horizontal). It was difficult to find sources for this one because it's more obscure, but of the sources I found (eStroke, MDBG and Wenlin), all agreed that the lower left-hand box on this radical should consist of 4 strokes, not three.

If you have strong evidence that these conclusions are incorrect, please post so that I can 1) avoid the work of changing the stroke order for thousands of characters and 2) avoid committing mistaken research to the database.

Can't comment on the others (all makes sense) but the change to make the interior of 母 interchangeable will be welcome, as I find the logic for all orders there somewhat subjective and often get warnings when writing it myself.

Well that's nice to hear, I had rather expected a firestorm of dislike for the changes!

George, I will take your word on it. I just trust that you wouldn't make these changes unless you felt they were necessary and anything that makes my writing practice the best it can be I am happy with.

I bow to superior knowledge. It will take a while to get used to the 门 change but it won't kill me.

As for 女, do you have any examples of it being written merged and not merged? I'm confused about when merging is allowed.

As for 母, I agree with icebear, I've never been able to figure out what Skritter wants me to do with it.

James

Hi James, sorry for the delay, these stroke order changes have taken way longer than expected. >whew<

I don't have concrete examples for merging on 女, generally I let the font guide me, and there does seem to be a noticeable difference in characters that support that shortcut. If it's ambiguous, I put the shortcut in since it's so much easier to actually write. Geez, writing 女 as three strokes is such a pain.

Unfortunately 母 has been a mess for a long time, hopefully when I get done going through all of these characters it will be easier to tell what Skritter expects!

I don't know if George is having the followings in mind. But in the cursive style, 女 is written in two strokes as follows:
好 http://9610.com/zidian/view.asp?id=20826
娃 http://9610.com/zidian/view.asp?id=20173
奴 http://9610.com/zidian/view.asp?id=6600

Also, the Japanese kana ぬ and め are derived from 奴 and 女 respectively.

They all follow the order shù zhé, wān, héng.

Wow, that's some serious cursive, pts. So it looks as though they follow the order I've proposed when not drawn in that style, and to some extent the shortcut version that Skritter supports is moving in that direction, but past that I don't think there's much we can do to support the cursive style owing to it's nebulousity.