First of all, Skritter is an incredible tool. Since I have discovered it last Friday, I have become a veritable addict (10.0 hours in six days, 964 characters learned...)

There are two reasons, why I am so excited about it:

1. The possibility to input the characters on the screen
2. The dynamic review (Difficult characters appear more often) THIS IS IT!!!

Especially the second one is such an incredible boost to learning efficiency. My method until now was: Repeat a character a few times until I remember it a little. If I bump into it soon after that, I am lucky and remember it, deepening my memory of it. If not, well, it sinks into the mud on the ground of the ocean of my memory, lost until it will be re-learned at a later stage. I don't know how many hundreds of times I have looked up a word only to find out that I had learned it in the past! The dynamic system of skritter really „leaves no character behind"...

At the moment, I would urgently encourage to add the following features (In my opinion, these two are a MUST):

Must-have A) Possibility of studying words, not only characters (english translation pops up, characters/pinyin are written in response. There should be a choice, if only characters, only pinyin or both have to be replied) Currently, I guesstimate that I can recognize 3500-4000 characters, but this is not even close to being halfway through learning the language. Studying the vocab, even if it is composed of already known characters is the much harder task now. The dynamic review cycle of Skritter would boost the efficiency for that task enormously.

Must-have B) Tool to create one’s own lists, preferably with the ability to create groups (like „job“, „newspapers“, „literature“, whatever...) - and then add sub-groups to it. Taking me as an example: As I am not using books anymore, my vocab to study comes from different sources, like emails at work, newspapers etc. For example, some of the last few words I needed to memorize were: 过敏症 (Allergy), 码头操作费 (Terminal handling charges) and 苯甲酸 (Benzoic Acid) – Your vocab lists are great, and I use them for a quick revision of my past studies, but at my present level, they are of little help to progress further and to practice the vocab I need everyday in my job.

Further suggestions, which would add to the luxury of this wonderful brain organizer:
1. Teaching handwriting style. Many people who learn characters (including me) at some stage start to write more quickly and then blur their characters, thinking this is handwritten style. But actually, the cursive hand is a technique of its own, many Chinese people’s handwriting is influenced by calligraphy. The most common style studied by many Chinese in order to improve handwriting is 行书(Xingshu) (It's somewhere between the rigid printed style and the crazy 草书(Caoshu), running grass style, which even many Chinese can't figure out). I would recommend to add one set of characters in this style (Just like there is one set for traditional and one set for simplified at the moment). There are different stroke orders and some characters are changed in way that makes them unrecognizeable for people who have only studied the 楷书 (Kaishu) style that skritter uses at the moment. You could choose one style by a certain calligrapher (I bought a very good "instruction" book by 朱涛(Zhu Tao), http://www.bookschina.com.tw/1445409.htm ， just as an example of a really fluent 行书, there are also more rigid ones around). But I think, learning the cursive hand should be one step after mastering the normal printing style (楷书), therefore separated by having it as a different set of characters.

2. Dynamic tone changes : Some characters (一, 不) change their tone according to the one coming after, as you all know. This is not featured in skritter when it asks for tones.

3. Example sentences. It is just better to see the characters in context. In reverse, there could also be a tool to have the example sentence displayed and having to insert the character in question, as another training method.

4. Japanese – Yeah, looking forward to that, I am really keen on getting into Japanese as well.

That should be it, no more ideas pop up at the moment. I am looking forward to seeing how this fine tool develops into that perfect high pressure device pumping tons of Chinese vocab into my brain, haha. Great job so far!

Wow, thanks for all the suggestions! The learning ocean mudground sounds exactly like how I used to learn characters, too.

964 characters added in 10 hours / 6 days: I'm assuming you knew how to write many of them previously? That's 37 seconds per character. Nice.

Must-have A): We're currently building it! It may or may not make it into the rearchitecturization, but it should be soon after if not. We weren't planning to prompt for both pinyin and characters from a definition at the same time, since it would be too ambiguous in many cases, but there'll be several different modes, all of which you'll be able to turn on/off, will mix together intelligently (trying to keep you from switching between keyboard/mouse too often), will be fast and use active recall, and will intelligently space out other reviews of the same thing from each other to keep yourself from short-circuiting the prompts. You're right that the word-level reviews are kind of broken, right now.

Must-have B): Adding personal vocabulary will soon be as easy as dropping a bunch of characters/words into a box and filling in the definitions / pinyins for any we don't have yet. We don't have immediate plans to do subgroups, although we may add tags in the future. I'm not sure when we'll have this, but it's a high priority and shouldn't take too long.

1. It's an interesting suggestion, to teach semi-cursive handwriting. It's currently beyond the scope of our tech, though, and too much work for three guys in an apartment to do well when there's so much else yet to be done. I can see how adding another set of characters could help; that's an interesting suggestion. It would certainly be easier to allow only one style at a time instead of trying to do both together. I think it's still not feasible in the near future, though.

2. Actually, in compound vocab words, I've changed all the tones on 一 and 不 to match their tone sandhi (let me know if I've missed any). That Skritter allows the original tones, too, perhaps needs some rethinking; I'll have to ponder how tone prompts should work in the new system. I've left all third-tones as thirds, though, despite sandhi.

3. We're going to be able to do example sentences really well, probably by picking all characters/words you know except the one being prompted for, etc. Whoa, it'll be cool. It'll be a while before we can implement this, though.

4. Mmm, Japanese...

So, hopefully those must-have features will be soon in coming. I think you will enjoy.

I think all of Xia's suggestions are great.

>3. We're going to be able to do example sentences really well, probably by picking all characters/words you know except the one being prompted for, etc. Whoa, it'll be cool. It'll be a while before we can implement this, though.

Wow, I just got excited reading this, That is seriously cool!!!

Sorry if this was posted already. I try to keep up with forum/blog entries here though admittedly on many of the longer posts my attention span has me skimming and skipping through the bulk.
To the point: Due to the disqus powered comments on the blog the section doesnt incorporate our account details from skritter.com. You may already have this adjusted for in the upcoming super-update, but I thought I would put in a friendly reminder just in case.

VERY COOL! Great job guys!!

I learned about you on the Chinesepod forums. Another suggestion I saw there that I wanted to bring to your attention was "integration" with Chinesepod... namely, making vocab lists for the Chineepod lessons available... interesting idea.

Cheers, keep it up.

I can see how using this will have the additional benefit of tidying up my style -- getting the immediate feedback on stroke positioning is something I really appreciate.

I've got some additional suggestions:

Suggestion 1: Hear tones practicing tool.
I'm a beginner with learning Chinese. One or the things I want to learn is to be able to hear which tone is pronounced.

Skritter has a database full of words and the pronunciation of these words. This can be used to make a tone-practicing-tool: i.e. one that pronounces a word and then the student has to tell which tone it is.
When the students masters this enough, then you can move on to partials and sentences instead of individual words.

Of course this functionality can be combined / extended to learn other things, like:
a. Write the pronounced word in pin yin (including the tone sign).
b. Write the Chinese character of the word that is pronounced.
c. Show the English meaning of the word afterwards (So the student can first say to himself "the pronounced word means ..." and then verify if he was right.) It would be even better if the student can write down the English meaning of the pronounced word and that Skitter verifies if it's right.


Suggestion 2: Pronunciation practicing tool
Another thing that I want to learn is to pronounce Mandarin Chinese correct. The Skitter database with pronunciations is useful for this again.

My second suggestion is to make a pronunciation tool:
i.e. one that lets the student hear a word, then the student has to pronounce the word in the microphone. Then Skritter lets the student hear the two audio tracks at the same time, so the student can compare if he pronounced it the same way.
To make it even better the following functionalities can be added:
a. Show the Chinese character, the pin-yin and the English translation of the word that the student has to pronounce.
b. Show the wave form of the Skritter audio track, the wave form of the students audio track, and the two wave forms overlapped. So the student has an visual comparison of how well he pronounced the word.
c. Let Skritter compare the Skritter audio track and the students audio track and let him know if he pronounced it: bad, reasonable, good, or like a native.

These were my suggestions. Hopefully they will be implemented.

Thnx,

Ward

Hi Ward, thanks for the suggestions!

These are good ideas, but making a separate practice mode for them is out of our reach right now, as will be most practice tools that won't directly integrate into the main practice. When you've got only three guys, focus is key.

However, you can get some of this either now or soon. Specifically, if you want to practice figuring out which tone has been said, you can click on the character's pinyin and the Infant will pronounce it. You can then do the tone input based on that listening discrimination.

We're working on pinyin and definition practice soon, where you'll be shown the character and have to fill in those things; spaced repetition will be done on those like it currently is done on the characters and tones. Depending how we do the interface, you may be able to type in the pinyin based on what you hear, if you so choose.

The pronunciation practice feature, though, I don't think will happen soon. All existing implementations I've seen of this are pretty grody. Perhaps we can make one that's actually really good, but we won't know until we've gotten a lot more of the core features done.

I was playing with the viewer earlier and I wanted to see all those characters that I wouldn't forget in 30 years, but when clicking on "Next" I saw it only organizes in really the next one (which is logical, but I was hoping clicking a 2nd time might organize the other way round).
And then I was looking for a button to go to the end of the viewer, but I could only go see the next 20. Since I have heaps and heaps of characters, it would take many clicks to get to those characters I won't forget in my old age.

Couldn't we have another button "end of the list"? Maybe it's a bit wacky, and after all it is late, but I'd like that :)

I tried to do the same thing, and did lots of clicking. That's why I suggested reverse sorting, I think in the new features thread I started.

Oh, I didn't look in the features thread, sorry. I thought of which thread to put it into, but it ended up in this one cuz I felt it was more of a light tweak than a real new feature. But that's rethorical really.

Reverse sorting would require another datastore index on UserItems, which would make writing the record of each review more expensive, and it's already expensive enough. We can't add a skip-to-end button, either, because we can't know how long the list is in terms of date offsets.

But when we create the data export thingy, you'll be able to get all the info you desire, yes.