(Archived) Skritter Mobile Beta 3.0


  • Added basic list management
  • Fixed issue where dot would remain on canvas
  • Minor improvements to prompt study stability
  • Single stroke tab reveals count against grading

Welcome to the new Skritter!
This week and early next week we’ll be fully rolling out the new Skritter Mobile for Android and iOS beta testers. There are major changes to nearly every aspect of the application and it is not 100% complete yet, but this beta has been in development internally and we’re to the point where we need user feedback and many of the bugs being reported in previous betas (2.3.5) have been fixed. So help us find bugs in the new mobile, please!

There are so many use cases that we just can’t account for, as everyone uses Skritter in their own way. It’s one of the best things about Skritter, actually, but it does make it very hard to test. So… apologies in advance to anyone who wasn’t expecting this big of a change. We might have broken some things for you, and we’re sorry! We are working hard to deploy daily updates that address the largest issues first. This post will serve as our official notes for the short-term future, as replying every bit of feedback takes away precious time from making changes and deploying fixes.

If you’re seeing a bug or an issue that isn’t on this list we certainly wanna hear about it! You can use the contact button in the app or open a topic on the forum. Steps to reproduce really help, and so do screenshots!

If you’re curious about how things work or just want a comprehensive overview of the application we have a walkthrough guide in a separate post here. we will make another version of this guide after things have stabilized a bit and some of the features we’re working on have been deployed.


  • Swipe up to erase
    Swipe up on writing canvas with one finger to erase and start over. You might miss the button at first, but we’re going to save you a lot of time in the long run.

  • Tap hold to reveal the character
    Double tap never worked 100% correct. Tap-hold might take some getting used to at first, but we promise it works better. After the character is revealed it will start to fade as you write the character and the card will automatically be marked wrong.

  • Tap once for the neutral tone (5th tone)
    Drawing circles was a bit silly, and it didn’t work 100% of the time. Just tap on the screen if it is a neutral tone prompt and it’ll register as 5th tone moving forward.


  • Progress screen
  • List management screen
  • Character decomps
  • Word breakdowns
  • Example sentences
  • Hanping/Pleco support
  • Automatic teaching mode support


  • Occasional writing and tone card freezes
    We’re getting reports of writing and tone prompt freezes sometimes, especially on multi-character words. If you encounter this issue you should be able to return to the dash and then start studying again. With that said, we’re still not able to produce this bug on command. We suspect it has to do with timing events firing out of place on occasion. I hope we can pool our collective powers to sort this one out soon!



  • Added auto advance speed setting for anything graded correct
  • Added button to all prompts for viewing mnemonic and sentence
  • Added grading sound effects for beta testing
  • Added option to auto reveal hidden properties upon prompt completation
  • Display top mnemonic if user mnemonic not selected
  • Fixed simplified and traditional display issues when viewing lists
  • Fixed some issues relating to navigation buttons
  • Layout fixes for both Android and iOS
  • Preserve definition line breaks in some locations


  • Added always visible navigation buttons to all prompts
  • Fixed definitions not updating immediately word list after editing
  • Fixed navigating back to writing prompt character rendering issue
  • Fixed tap hold grading bug when auto advance enabled
  • Load contained characters from server when missing from cache


  • Fixed study time display issue caused by timezones
  • More improvements to general prompt stability


  • Fixed manual item adding regression
  • Fixed tone prompts displaying v instead of ü
  • Set auto adding limit to 10 (setting coming soon)


  • Added character font to contained and decomp area
  • Apply grading color to raw squigs when character written
  • Fixed bug preventing Japanese word screen from displaying words
  • Fixed word drawer sometimes crashing when loading definition
  • Improved general prompt stability while studying


  • Changed writing brush style due to performance issues
  • Fixed reading prompts from auto advancing on first tap


  • Fixed major bug causing accounts to crash (Japanese only)
  • Fixed timer formatting when over an hour
  • Fixed word count on the list section screen
  • Improved auto advance flow based on user feedback (detail notes and discussion here)
  • Improved display of words based on account study settings
  • Removed kana only writings when studying kana is disabled


  • Added support for Pleco and Hanping dictionary links
  • Added contained characters and decomps to info drawer
  • Cleaned up text when auto advance enabled on definition and reading prompts
  • Fixed writing and tone prompt orientation change issues


  • Fixed tone prompts failing to load when character data missing
  • Fixed words failing to download on initial login
  • Improved recognition of all last stroke in 乾
  • Moved list due count and adding percent from dashboard to list screen


  • Added heisig keywords when setting enabled
  • Changed grading colors to match official website palette
  • Changed tone prompt styling for improved clarity
  • Fixed regression causing some tone and writing prompts to stop working


  • Added styling to example sentence on info drawer
  • Fixed more bugs relating to simplified and traditional Chinese
  • Improved study performance by keeping components alive
  • Improved neutral tone tap detection


  • Added swipe up to erase tone prompts
  • Fixed back button with auto advance on definition and reading prompts
  • Fixed typo in progress screen redirect link
  • Improved dashboard list due count calculation accuracy
  • Improved initial account download performance and stability


  • Added in support for navigating backwards (check the prompt card toolbar)
  • Added workaround links for progress and list management
  • Added study filter option to disable automatic audio playback
  • Tap hold (2 fingers) to enable teaching mode
  • Fixed several bugs relating to simplified and traditional Chinese
  • Fixed issue with first kana writing needing to be written even when study kana was disabled



iOS users! If you’re already using 2.3.5 on iOS we have your email and will be sending out the first TestFlight build later today. We will also have a new iOS sign up form up sometime soon, too.


What bug reporting method is preferred/better for your team - sending from the contact link in the app or opening a new topic (?) in the forum? Also, if we send using the app, is the version and platform automatically sent or do we need to add that?

I’ll defer to @josh for preferred method. But submitted reports in app does give a lot of helpful info about version and what not.

Using the contact method form within the app is preferred for the reasons @rpetersn mentioned. It’ll automatically pass along things like device model, version and which language you’re studying.


I haven’t seen the iOS beta on TestFlight yet. I just see the older 3.0.4 version. What is the status of the iOS beta?

If you’re already part of the testflight beta then stay tuned and you’ll be getting an email in a day or two about the update. Right now we’ve only released 3.0 to Android and have been making a few tweaks before releasing to iOS.


Maybe there is a way I have not found, but does the forum not show an update when this post is updated with new release notes, since the original post gets edited? Maybe post a comment each time saying it has been updated? Otherwise I miss the fact that a new beta is released. TestFlight is not notifying me either, so I have to check manually.

1 Like

I am seeing the same thing. I have all the Test Flight notifications turned on on my iPad, but never received any notifications on the Skritter updates for 3.0.6 nor 3.0.7.
Had I not been reading this forum, I wouldn’t have known the updates were ready.

I love the sentences with audio playback in 3.0.7. Nice work! I reported a bug in-app that sometimes the new sections are not showing up.

I would love to try the new android Beta and I’m signed up as a beta tester through google play, but the only version of skritter it gives me is 2.1.2. Is there some way to force it to update to the lastest beta version? It was updating fine until 3.0.3. After that version it wouldn’t update so I uninstalled and reinstalled skritter and it took me back to 2.1.2. Then I tried leaving the beta and rejoining and got the same thing.

Is there a way to mark a word as incorrect, even if you knew all the characters? The old Skritter iOS differentiated that, but I don’t see a way to flag the word instead of, say, the last character of a multi-character word.

I have the same question. Related, what did it mean in the system to be able to mark each character as correct but the word as “forgot”, for writing? Did that actually do anything? I did use it in the last occasionally, but I was never sure what it meant in that case.

@rpetersn. Since list entries have different line items for words and characters, I always assumed the tracking statistics from an SRS viewpoint treated them as different “entities”.

But you know what the say happens when one assumes.

Right, but that is less meaningful for character writing than eg definition. What does it mean to be able to write each character correctly individually but not when combined?

My observations have been that the “grade” for a word writing prompt always is the lowest score of the individual characters.

So, a 3 character word with individual scores of: a green, a green, & a red = RED.

For the “item” in SRS its stored score would be RED.

You can then manually override the calculated score with your own choice and Skritter stores that score instead.

This description is just based on how I see the behavior appearing to function. The Devs would have to comment on how it really functions under the hood.

Most of the time words with multiple characters will get graded as either right or wrong, but there are cases where longer words might get graded differently. We programmatically represent things as wrong (1), so-so (2), correct (3) and easy (4). I’m including the code calculation below.

if (totalCount === 2 && totalWrong === 1) {
score = 1
} else if (totalWrong > 1) {
score = 1
} else {
score = Math.floor(totalScore / totalCount)

So a three character prompt graded 3, 3, 1 would be Math.floor(7/3) which is 2. For things that are two characters they will always be marked right or wrong, so-so grading only comes into play with 3+ characters.

1 Like

Oddly, I received the TestFlight notifications for Skritter 3.0.7…today.

Just a few days late…surprisingly so.

@apoxomis We ran into a hiccup with sending them, but the invites did out today. Sorry for the delay!