Skip to content

Conversation

@xiaoditx
Copy link
Contributor

@xiaoditx xiaoditx commented Nov 1, 2025

Hi, this PR introduces Chinese translations for the README and fixes some broken links in the original document.

What's included:

  • Added three variants of Chinese translation: Simplified Chinese, Traditional Chinese (Taiwan), and Traditional Chinese (Hong Kong/Macau).
  • Fixed a number of incorrect links in the original README (though I'm not entirely sure if all corrections are perfect).
  • Added a link to the Chinese version in introduction.md. Since the current translation only covers about 50% of the full English website, I've temporarily pointed this link to my personal site. This allows users to access the most up-to-date version and provides a foundation for future, more comprehensive localization.

Thank you for considering this contribution!

@ArthurSonzogni
Copy link
Owner

Thank you very much @xiaoditx !

  • Broken links I fixed the generation here
  • I've temporarily pointed this link to my personal site => That's okay for me. Especially if this avoids me to maintain it.

I am okay to merge the part adding the external link!

Long term, I believe I would like something with low maintenance cost, mostly correct, improvable, that would scale to multiple languages, not increasing the repository size. Here is an idea:

  • In an orphan branch and/or an external repository:
  • A python script would pull:
    • The content of the orphan branch
    • The content of origin/<latest_release> .
  • A gemini-cli prompt would be asked to translate the Doxygen comments and markdown from origin/<latest_release>. As much as possible, it should copy the previous translatiosn from the orphan branch.
  • Generate the doc and push it to the orphan branch.
  • When FTXUI generates the doc, it can cherry-pick the content of the orphan branchs.

What do you think about this ideas?

@ArthurSonzogni
Copy link
Owner

I tried the AI idea:
https://github.com/ArthurSonzogni/ftxui-translations/tree/zh-CN

It is not complete, and costed me ~35€ already, which is way too much. However calling the script to update translation would be scalable to multiple language and maintainable.
I am going to investigate how to scale down the cost and let you review the python script driving AI to achieve this.

@xiaoditx
Copy link
Contributor Author

I've taken a close look at that AI translation repository, and the overall performance is indeed quite good. However, as you mentioned, the pricing is on the higher side—thirty-five euros, based on the cost of living in China, is roughly equivalent to 4–6 days of daily food expenses for a family of three. I recently looked into Youdao's translation API, which is officially priced at 48 RMB per million characters. It seems much more cost-effective, though I'm not yet sure if it fully meets our functional requirements.

Additionally, regarding the quality of AI translations, I noticed some inaccuracies in certain parts of the translated documentation. It might be a good idea to retain human-translated versions for these specific segments and implement a protection mechanism to prevent them from being automatically overwritten by AI translations, unless the corresponding source content is updated.

By the way, shouldn’t we add a link to the Chinese version of the repository in the README now?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants