Bug Description
When typing Korean (Hangul) characters in the OpenCode interface input field, the characters appear as broken/garbled text, making it impossible to verify typed content in real-time.
Steps to Reproduce
- Launch OpenCode
- Try typing Korean text in the input field (e.g., 안녕하세요)
- Observe that the characters appear broken/garbled
Expected Behavior
Korean characters should display correctly in the input field as they are being typed.
Actual Behavior
Korean characters appear as broken/garbled characters or empty boxes, preventing users from seeing what they are typing.
Environment
- OS: macOS 26.0.1 (Build 25A362)
- OpenCode Version: 0.15.18
- Terminal: xterm-256color
- Locale: ko_KR.UTF-8
Additional Context
- The AI correctly recognizes and processes Korean input (backend works fine)
- The issue is purely visual/UI rendering related
- After sending the message, AI responds correctly in Korean
- The problem only affects real-time input visibility, not functionality
Impact
This makes it extremely difficult for Korean users to:
- Verify spelling while typing
- Catch typos before sending
- Use OpenCode effectively in Korean
Suggested Fix
Possible solutions:
- Add CJK (Chinese, Japanese, Korean) font support to the input field
- Ensure proper UTF-8 rendering in the UI component
- Test with fonts that include Korean glyphs (e.g., Noto Sans CJK, D2Coding)
Bug Description
When typing Korean (Hangul) characters in the OpenCode interface input field, the characters appear as broken/garbled text, making it impossible to verify typed content in real-time.
Steps to Reproduce
Expected Behavior
Korean characters should display correctly in the input field as they are being typed.
Actual Behavior
Korean characters appear as broken/garbled characters or empty boxes, preventing users from seeing what they are typing.
Environment
Additional Context
Impact
This makes it extremely difficult for Korean users to:
Suggested Fix
Possible solutions: