-
-
Notifications
You must be signed in to change notification settings - Fork 62
1. FAQ
This section contains questions that are frequently asked by users.
No. jidoujisho is powered by Flutter and is written as cross-platform Dart code. This means that with some effort, it is technically possible to port the app to iOS. However, there are blockers to iOS development that would prevent a port from meeting feature parity with the Android app.
These include but are not limited to:
- Platform differences with dependencies: With the app being mainly developed for Android and largely untested for iOS, it is very likely that there are many minor platform differences that will need specific fixes on iOS. This includes the app itself, and its dependencies, such as its video player and WebView dependencies.
- File system permissions: This is a blocker for loading files relative to another file automatically, such as same name subtitles or Mokuro images.
- Third-party services: The application scrapes data from third-party sources, which is likely to violate App Store guidelines.
- Store submission, cost and compliance: Developing for iOS is a great inconvenience for hobbyists (not to mention professionals). Distribution and development on Android incurs relatively no cost and difficulty.
The project is free and open source, and developers are encouraged to make an attempt if they so wish. However, with the very high likelihood that efforts to an iOS port that is fully compliant would result in many compromises and feature cuts, this main project instead aims towards building the best application possible on Android devices.
Similar to the above, jidoujisho is developed largely as a hobby project aiming to give both users and developers an unconstrained experience. Google Play's content policy makes it very unlikely that certain features that use third-party services such as YouTube would be included.
Rather than split up and confuse users with versions that are missing key features, and make it more difficult for developers to maintain separate branches for a stripped down Google Play version and a complete free and open source version, it is easier and more convenient for all to keep things as is.
Please refer to the Installation section.
jidoujisho has been tested on multiple emulators and real world devices running Android 7.0 and above, but extensive development and personal use has been performed on a OnePlus 6 and a Google Pixel 6 Pro.
These do not serve strictly as minimum and recommended requirements by any means, but testing has been done to target a smooth experience on these devices. You may use the specifications of these devices to compare and evaluate your choice of device.
Please refer to the Getting Dictionaries section.
jidoujisho is provided as-is, similar to any other video player, e-book reader or web browser. The app is capable of loading media from external sources, such as YouTube, but does not host its own content. Users are to provide any other media by their own means.
At the moment, Japanese and English are officially supported. It is possible to use the application with a supported dictionary format but with an unsupported language, though ultimately it will depend on whether the search algorithm for the user's selected target language works well with the unsupported language.
If you want jidoujisho to officially support your language, you may make a pull request with a proposed implementation or a discussion with the following:
- A specification detailing your language's needs that compassionately explains to non-learners of that language any key concepts necessary for the implementation
- An appropriate common dictionary format that exists for the language
- Suggested implementation for its search algorithm
You can hide the dictionary pop-up by swiping it away horizontally, tapping highlighted text again or if reading novels or manga, using the volume buttons.
Please see refer to this section on Recursive Look-ups.
Installing the app adds a context menu option (the menu where you may Copy, Paste, Select All and so on) which you may access anywhere that Android or other applications allow it to be used. Using the context menu option will launch the app and search the selected text.
Additionally, you may share text or images to the app in order to open the Card Creator. Sharing a YouTube video link or stream URL will also launch the app. You may also use the app as an external player for Jellyfin to stream media from your PC.
Please refer to the page for Workflow Setup.
While the project tries to cater to features that it finds are invaluable to learners, there are too many card types and workflows in the wild, and the project cannot be expected to meet feature parity with all possible user customizations.
You may try to make an issue or discussion that proposes your desired feature, but if you really want it implemented, you can do that best by perusing the Development section on how to implement it for yourself. The project will happily accept pull requests for whatever ingenious feature you find missing!
I have this idea for supporting a dictionary format, language, media source, user enhancement and so on. Can you support it?
The jidoujisho project will try with its best effort to reasonably support common language learning community standards and types of media.
If you strongly wish to see your suggestion realized, see the Development section to learn what you can do to help.
Early in jidoujisho's development, it was decided that a key philosophy of the project would be to prevent users from performing what are certainly some pitfall behaviors that would hinder their progress with their target language by reasonably omitting certain features that they may rely on as a crutch.
The sooner that users drop using translations, the sooner they will come to get used to their target language. As such, this is one such feature that will not be supported.
Nothing is stopping users from creating dual language subtitles in a single subtitle file, but the hope is that the effort it would take to source or create such subtitles will encourage users to attempt to get out of their comfort zone with native subtitles instead.
Users have access to the share context menu option for any text in the application, which they may use to launch a different application. However, as above, the project is designed with the intent of making the user conscious of this repeated effort on their part, and instead push users out of their comfort zone.
The app supports Mokuro for reading manga, which is a format that already has processed text. This approach is preferred to on-board optical character recognition, the existing solutions of which would either be slow, inaccurate or both.
Though Tesseract and Google ML Kit exist and are reasonably good solutions and have historically been included in the past with chisa, the time and performance it takes, the disruption it causes to a user waiting for text to process and the chance for errors make it difficult for the project to reconsider these solutions.
Users are encouraged to be able to read text from images without the use of optical character recognition rather than relying on them, and even better are encouraged to perceive these as a potential exercise to learn how to write these characters with their handwriting IME instead.
As with the above, the aim of the app is to present media or information as a native would see or perceive themselves, with no extra supplements or frills. Language learners will have no special highlight vision in the real world when their tools are taken away from them.
In essence, users may use the app as a standard video player or reader, and it will appear as so until they explicitly instruct the app that they need help by way of looking words up.
Without explicit user intervention, the app gives no passive comprehension assistance to the user. When the user leaves the app and consumes native media by a way that does not involve language learning software, it will appear to them as they have gotten used to in the app this whole time, only that they cannot press on any text.
The goal of the app is to slowly ease users into the point where they are consuming media and are unconscious that they have not required look-ups at all, and eventually not need this application at all. Hence, omitting this feature is in line with the intent to force users out of their comfort zone as with the previous questions above.
jidoujisho or 自動辞書 is itself not a real Japanese term, but is a 四字熟語 portmanteau of 自動 which means automatic and 辞書 which means dictionary. The app is not meant to be an automatic dictionary in the literal sense of the word, but rather represents the spirit of the project, aiming to automate the setup of a language learning dictionary look-up tool and workflow.
In the beginning, the name was definitely conflicting in that Chinese and Korean learners would pronounce the characters very differently, but the project has definitely grown to befit the name and its roots with the Japanese language learning community, not to mention that the romanization is quite nice to look at!
The logo is both by suzy and Aaron Marbella, please support their awesome work if you can!
jidoujisho started naming its separate project repositories with separate names to differentiate it from the very first legacy codebase. For example, 1.x/chisa, and the current project labelled 2.x/yuuna. 0.x does not have a name and is simply called legacy. As for telling you why those names were chosen: それは禁則事項です！