You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Severely laggy typing in chat input (contenteditable has ~500ms latency) #28162
Typing has ~500ms+ latency per keystroke. Pressing two keys rapidly
swallows the second. UI is unresponsive despite low resource usage.
Likely cause
The input uses <div contenteditable> instead of native <textarea>.
Combined with an unreconciled long conversation DOM tree, each keystroke
stalls the main thread on reflow/render.
Suggestion
Switch to native <textarea> for input
Add virtual scrolling for conversation history
Ps.Likely cause and Suggestion is made by opencode himself not me
Plugins
none
OpenCode version
1.15.4
Steps to reproduce
1.use open code electron app for a whole day
2.typing in conversation box become 不利索,even typing english with out chinese input. method I am using win11 en 23h2
I'm using Linux Fedora 43, I have the same issue.
Based on @halilang70-png suggestion, In production app, I changed input to textarea (via dev tools) and the delay was gone but the @mention styling was also gone as expected.
However, There's no laggy typing when I run the desktop app in my chrome browser (from source code).
I'm using Linux Fedora 43, I have the same issue. Based on @halilang70-png suggestion, In production app, I changed input to textarea (via dev tools) and the delay was gone but the @mention styling was also gone as expected.
However, There's no laggy typing when I run the desktop app in my chrome browser (from source code).
They say it's the git, If a large change was made by ai agent ,the open code trying to scan the git DB so it become laggy,but how can I use an ai code agent without git, maybe this a asynch problem
I'm using Linux Fedora 43, I have the same issue. Based on @halilang70-png suggestion, In production app, I changed input to textarea (via dev tools) and the delay was gone but the @mention styling was also gone as expected.
However, There's no laggy typing when I run the desktop app in my chrome browser (from source code).
They say it's the git, If a large change was made by ai agent ,the open code trying to scan the git DB so it become laggy,but how can I use an ai code agent without git, maybe this a asynch problem
I built an RPM file on my linux, then used this RPM to reinstall opencode desktop app.
Now it's not laggy any longer when I use opencode on the same project which was laggy before.
Give it a try. Solve the issue first, then we'll find the reason :-)
@halilang70-png I added this option to app settings so that I can disable electron hardware acceleration and the input delay decreased significantly.
I think there might b some issues with electron hardware acceleration on some gpus too.
Activity
github-actions commentedon May 18, 2026
This issue might be a duplicate of existing issues. Please check:
If your issue is distinct, feel free to clarify what makes it different.
behnammby commentedon May 19, 2026
I'm using Linux Fedora 43, I have the same issue.
Based on @halilang70-png suggestion, In production app, I changed input to textarea (via dev tools) and the delay was gone but the @mention styling was also gone as expected.
However, There's no laggy typing when I run the desktop app in my chrome browser (from source code).
halilang70-png commentedon May 20, 2026
They say it's the git, If a large change was made by ai agent ,the open code trying to scan the git DB so it become laggy,but how can I use an ai code agent without git, maybe this a asynch problem
behnammby commentedon May 20, 2026
I built an RPM file on my linux, then used this RPM to reinstall opencode desktop app.
Now it's not laggy any longer when I use opencode on the same project which was laggy before.
Give it a try. Solve the issue first, then we'll find the reason :-)
behnammby commentedon May 27, 2026
@halilang70-png I added this option to app settings so that I can disable electron hardware acceleration and the input delay decreased significantly.
I think there might b some issues with electron hardware acceleration on some gpus too.