4 ms·

Show HN: Hanly – Learning Chinese characters reimagined

1 points by tutubuilds 2y ago

Learning Chinese characters is unreasonably hard and is arguably the main reason why most people give up on the language. The existing apps mostly just show you a character and ask you to “remember” these random-looking complex shapes. Frustrated with this during my own learning journey, I decided to build Hanly. The core idea: there is deep tree-like structure to Chinese characters, and Hanly teaches them by making you climb up this hierarchy, so that at every step you either learn a very basic building block or combine already known building blocks. This approach makes learning feel more like playing with LEGOs rather than memorizing random shapes. I left a link for iOS but the app is also available on Android (currently completely free on both platforms): https://play.google.com/store/apps/details?id=app.hanly.chinese https://play.google.com/store/apps/details?id=app.hanly.chin... It’s the first app I ever built and any feedback is very appreciated!

yoniploni 2y ago
Nice app! The interface is smooth and attractive and shows a lot of good interconnected information about the characters, their components, and the words they’re used in. Suggestions after a short time exploring the app: 1. More clearly convey when character components/radicals are “meaning” or “sound” components (or both, or neither). The best resource for this I know of is the etymological dictionary from Outlier Linguistics (which you can get as a Pleco add-on). The mnemonic “stories” you provide associating each character with its components do help generate mental images to aid memorization, but they may also cause a false implicit understanding that all character components are “meaning” components and/or a false implicit understanding of character etymologies. For instance, the story for 讲 (jiǎng) notes the meaning of component 井 (jǐng), when this component actually serves as a sound component and not a meaning component in this character. 2. Include all common meanings and pronunciations for characters with multiple different meanings/pronunciations on the “learning” page (or you could create multiple different learning pages). Each character “learning” page seems to cover only a single meaning and pronunciation. For example, basic characters 着, 待, 当, 了 have multiple common pronunciations and meanings. Another example: the learning page for character 壳 shows pronunciation qiào, when alternate pronunciation ké is more common. 3. Note which characters and words from the HSK 2.0 and 3.0 levels are not (yet) included in the app’s vocabulary. It seems only 910 characters and 2000 words composed of these characters are included in the app, even though HSK levels 1-6 are identified. From some quick searching, it seems HSK 2.0 levels 1-6 include 5000 words with 2663 unique characters; HSK 3.0 levels 1-6 include 5456 words with 1800 unique characters (and obviously many more words and characters for HSK 3.0 levels 7-9). 4. Include a quiz to determine exactly which characters the user already knows, rather than just having the user estimate the number of characters.
- tutubuilds 2y ago
  Thanks for checking out the app, this is helpful feedback! To briefly respond: 1. Adding explicit info on which components are phonetic vs semantic is def smth we need to do! 2. Not sure if you have discovered the back of the card where we show all definitions and pronunciations. The purpose of the front card is to introduce a new learner to the character, so unless it is absolutely necessary we avoid enumerating to many meanings & pronunciations there, because it is intimidating and in my opinion unnecessary for initial learning. 3. App's vocabulary includes all words and characters. However the learning sequence includes only some characters (around 1000), which is probably what you are referring to. (In the Characters tab you can switch between main learning sequence and all HSK characters via a switch on the bottom) 4. Avoiding a quiz is a deliberate choice. It's very easy for a learner to mark which characters they know (just hold them down in Characters tab or anywhere else in the app), so making a person take a quiz before starting using the app felt like unnecessary delay.
sherdil2022 2y ago
Very nicely done. How are you animating the strokes for the characters?
- tutubuilds 2y ago
  Thank you! Yup, we do animate the strokes.
  - sherdil2022 2y ago
    How though? Have you recreated the glyhs since animating the stroke with current OpenTypeFonts is not that easy?
    - tutubuilds 1y ago
      Oops sorry misread your original question. I am using this:https://github.com/skishore/makemeahanzi https://github.com/skishore/makemeahanzi
      - sherdil2022 1y ago
        Thank you. This is very helpful. Let me know if you are looking for collaborators.

1 points by tutubuilds 2y ago

yoniploni 2y ago

Nice app! The interface is smooth and attractive and shows a lot of good interconnected information about the characters, their components, and the words they’re used in. Suggestions after a short time exploring the app: 1. More clearly convey when character components/radicals are “meaning” or “sound” components (or both, or neither). The best resource for this I know of is the etymological dictionary from Outlier Linguistics (which you can get as a Pleco add-on). The mnemonic “stories” you provide associating each character with its components do help generate mental images to aid memorization, but they may also cause a false implicit understanding that all character components are “meaning” components and/or a false implicit understanding of character etymologies. For instance, the story for 讲 (jiǎng) notes the meaning of component 井 (jǐng), when this component actually serves as a sound component and not a meaning component in this character. 2. Include all common meanings and pronunciations for characters with multiple different meanings/pronunciations on the “learning” page (or you could create multiple different learning pages). Each character “learning” page seems to cover only a single meaning and pronunciation. For example, basic characters 着, 待, 当, 了 have multiple common pronunciations and meanings. Another example: the learning page for character 壳 shows pronunciation qiào, when alternate pronunciation ké is more common. 3. Note which characters and words from the HSK 2.0 and 3.0 levels are not (yet) included in the app’s vocabulary. It seems only 910 characters and 2000 words composed of these characters are included in the app, even though HSK levels 1-6 are identified. From some quick searching, it seems HSK 2.0 levels 1-6 include 5000 words with 2663 unique characters; HSK 3.0 levels 1-6 include 5456 words with 1800 unique characters (and obviously many more words and characters for HSK 3.0 levels 7-9). 4. Include a quiz to determine exactly which characters the user already knows, rather than just having the user estimate the number of characters.

tutubuilds 2y ago
Thanks for checking out the app, this is helpful feedback! To briefly respond: 1. Adding explicit info on which components are phonetic vs semantic is def smth we need to do! 2. Not sure if you have discovered the back of the card where we show all definitions and pronunciations. The purpose of the front card is to introduce a new learner to the character, so unless it is absolutely necessary we avoid enumerating to many meanings & pronunciations there, because it is intimidating and in my opinion unnecessary for initial learning. 3. App's vocabulary includes all words and characters. However the learning sequence includes only some characters (around 1000), which is probably what you are referring to. (In the Characters tab you can switch between main learning sequence and all HSK characters via a switch on the bottom) 4. Avoiding a quiz is a deliberate choice. It's very easy for a learner to mark which characters they know (just hold them down in Characters tab or anywhere else in the app), so making a person take a quiz before starting using the app felt like unnecessary delay.

sherdil2022 2y ago

Very nicely done. How are you animating the strokes for the characters?

tutubuilds 2y ago
Thank you! Yup, we do animate the strokes.
- sherdil2022 2y ago
  How though? Have you recreated the glyhs since animating the stroke with current OpenTypeFonts is not that easy?
  - tutubuilds 1y ago
    Oops sorry misread your original question. I am using this:https://github.com/skishore/makemeahanzi https://github.com/skishore/makemeahanzi
    - sherdil2022 1y ago
      Thank you. This is very helpful. Let me know if you are looking for collaborators.