starred/terminal
1
0
Fork 0
mirror of https://github.com/microsoft/terminal.git synced 2026-02-03 21:25:34 +00:00
Code Issues 19.0k Packages Projects Releases 20 Wiki Activity

WT resets palette on input language change #23706

New Issue
Closed
opened 2026-01-31 08:49:53 +00:00 by claunia · 6 comments
claunia commented 2026-01-31 08:49:53 +00:00
Owner
Copy Link

Originally created by @alabuzhev on GitHub (Oct 15, 2025).

Windows Terminal version

1.25.251014001-llm

Windows build number

10.0.19045.6332

Other Software

No response

Steps to reproduce

  1. Add any extra language or keyboard layout to your system.
  2. Rename the attached file (pal.cmd.txt) to *.cmd or *.bat
  3. Run it in WT
  4. You should see the default palette (e.g. campbell)
  5. Press any key
  6. You should see the reversed vintage palette
  7. Switch the input language or the keyboard layout (Ctrl+Shift or Alt+Shift or Win+Space etc.)

Expected Behavior

Nothing. Why would anything happen, right?
Except for the language or the layout change obviously.

Actual Behavior

The palette resets back to the default (e.g. campbell)
What the... How the... Why?

https://github.com/user-attachments/assets/dd322885-6a9b-4c59-8b13-d52787cb11ff

It does not happen in openconsole.

Also, it is interesting that in openconsole changing the palette affects the whole buffer:

Image

But it does not happen in WT. No idea if that is also a bug or by design.

Originally created by @alabuzhev on GitHub (Oct 15, 2025). ### Windows Terminal version 1.25.251014001-llm ### Windows build number 10.0.19045.6332 ### Other Software _No response_ ### Steps to reproduce 0. Add any extra language or keyboard layout to your system. 1. Rename the attached file ([pal.cmd.txt](https://github.com/user-attachments/files/22933930/pal.cmd.txt)) to *.cmd or *.bat 2. Run it in WT 3. You should see the default palette (e.g. campbell) 4. Press any key 5. You should see the reversed vintage palette 6. Switch the input language or the keyboard layout (Ctrl+Shift or Alt+Shift or Win+Space etc.) ### Expected Behavior Nothing. Why would anything happen, right? Except for the language or the layout change obviously. ### Actual Behavior The palette resets back to the default (e.g. campbell) What the... How the... Why? https://github.com/user-attachments/assets/dd322885-6a9b-4c59-8b13-d52787cb11ff It does not happen in openconsole. Also, it is interesting that in openconsole changing the palette affects the whole buffer: <img width="653" height="245" alt="Image" src="https://github.com/user-attachments/assets/dc2a1b0a-ed84-4a91-b8fc-6a235384f1d6" /> But it does not happen in WT. No idea if that is also a bug or by design.
claunia added the Needs-TriageIssue-Bug labels 2026-01-31 08:49:54 +00:00
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@lhecker commented on GitHub (Oct 16, 2025):

Why?

For what it's worth, we reload the settings because keybindings are layout specific. This then triggers a reset of the selected theme. (It should not be doing that, to be clear.)

@lhecker commented on GitHub (Oct 16, 2025): > Why? For what it's worth, we reload the settings because keybindings are layout specific. This then triggers a reset of the selected theme. (It should not be doing that, to be clear.)
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@alabuzhev commented on GitHub (Oct 16, 2025):

Thanks for the explanaiton.

we reload the settings because keybindings are layout specific

Oh that's interesting. May I ask why?
Indeed, a key like / can come from VK_OEM_2 or Shift+VK_OEM_5 or who knows what else depending on the layout, but who cares?
Or are you trying to resolve that back to the physical key combos in the current layout that would have produced that key if pressed and show that in the UI? It feels like it might have funny consequences and is not bug-free by definition, but I'll probably describe that in a different issue.

Also, keybindings are layout specific, but the config isn't, why do you need to reload it?
The actions look like trivial key-value pairs in json, why not just keep those config-specific key names like ctrl+comma in memory? It's not like the users will have millions of actions for it to have any measurable impact.

In some situations the user might need to switch the layout every several seconds (imagine writing a piece of HTML in a non-Latin language, but with Latin tags etc.) Attaching disk IO, config parsing and UI updates to such events looks a bit excessive.

@alabuzhev commented on GitHub (Oct 16, 2025): Thanks for the explanaiton. > we reload the settings because keybindings are layout specific Oh that's interesting. May I ask why? Indeed, a key like `/` can come from VK_OEM_2 or Shift+VK_OEM_5 or who knows what else depending on the layout, but who cares? Or are you trying to resolve that back to the physical key combos in the current layout that would have produced that key if pressed and show _that_ in the UI? It feels like it might have funny consequences and is not bug-free by definition, but I'll probably describe that in a different issue. Also, keybindings are layout specific, but the config isn't, why do you need to reload it? The actions look like trivial key-value pairs in json, why not just keep those config-specific key names like `ctrl+comma` in memory? It's not like the users will have millions of actions for it to have any measurable impact. In some situations the user might need to switch the layout every several seconds (imagine writing a piece of HTML in a non-Latin language, but with Latin tags etc.) Attaching disk IO, config parsing and UI updates to such events looks a bit excessive.
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@lhecker commented on GitHub (Oct 16, 2025):

Oh that's interesting. May I ask why?

Initially because of the Quake-mode keybinding. It's bound to a scancode (whatever key is below Esc) but the RegisterHotKey API takes a vkey. So, we need to map the scancode to a vkey after loading settings and update it whenever the layout changes.

Also, keybindings are layout specific, but the config isn't, why do you need to reload it?

It was the easiest solution. 🙈 Keybindings with their assigned actions get put into various hashmaps and there's a lot of spaghetti code between the settings load and those hashmaps. The intention has been for a while to fix that and do exactly what you suggest.

@lhecker commented on GitHub (Oct 16, 2025): > Oh that's interesting. May I ask why? Initially because of the Quake-mode keybinding. It's bound to a scancode (whatever key is below Esc) but the `RegisterHotKey` API takes a vkey. So, we need to map the scancode to a vkey after loading settings and update it whenever the layout changes. > Also, keybindings are layout specific, but the config isn't, why do you need to reload it? It was the easiest solution. 🙈 Keybindings with their assigned actions get put into various hashmaps and there's a lot of spaghetti code between the settings load and those hashmaps. The intention has been for a while to fix that and do exactly what you suggest.
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@alabuzhev commented on GitHub (Oct 16, 2025):

It feels like it might have funny consequences and is not bug-free by definition, but I'll probably describe that in a different issue.

It can get funny indeed. Described in #19464.

@alabuzhev commented on GitHub (Oct 16, 2025): > It feels like it might have funny consequences and is not bug-free by definition, but I'll probably describe that in a different issue. It can get funny indeed. Described in #19464.
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@j4james commented on GitHub (Oct 16, 2025):

FYI, this is a duplicate of #11522.

Also, it is interesting that in openconsole changing the palette affects the whole buffer:

@alabuzhev This is because the openconsole background color is mapped to palette entry 0 by default, so when you update that entry in your script, it affects the entire buffer. In Windows Terminal the default background color is a separate palette entry outside the basic ANSI range, and your script doesn't update that entry.

If you look at the Terminal tab in the openconsole settings properties dialog, there's a checkbox labelled Use Separate Background. If you check that option it should work the same way as Windows Terminal.

@j4james commented on GitHub (Oct 16, 2025): FYI, this is a duplicate of #11522. > Also, it is interesting that in openconsole changing the palette affects the whole buffer: @alabuzhev This is because the openconsole background color is mapped to palette entry 0 by default, so when you update that entry in your script, it affects the entire buffer. In Windows Terminal the default background color is a separate palette entry outside the basic ANSI range, and your script doesn't update that entry. If you look at the _Terminal_ tab in the openconsole settings properties dialog, there's a checkbox labelled _Use Separate Background_. If you check that option it should work the same way as Windows Terminal.
claunia commented 2026-01-31 08:49:55 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Oct 22, 2025):

Indeed! I'm gonna merge this one over there. Sorry/thanks!

@DHowett commented on GitHub (Oct 22, 2025): Indeed! I'm gonna merge this one over there. Sorry/thanks!
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
⛺ Reserved
A11yCO
A11yMAS
A11ySev1
A11ySev2
A11ySev3
A11yTTValidated
A11yUsable
A11yVoiceAccess
A11yWCAG
Area-Accessibility
Area-AtlasEngine
Area-AzureShell
Area-Build
Area-Build
Area-Chat
Area-CmdPal
Area-CodeHealth
Area-Commandline
Area-CookedRead
Area-DefApp
Area-Extensibility
Area-Fonts
Area-GroupPolicy
Area-i18n
Area-Input
Area-Interaction
Area-Interop
Area-Localization
Area-Output
Area-Performance
Area-Portable
Area-Quality
Area-Remoting
Area-Rendering
Area-Schema
Area-Server
Area-Settings
Area-SettingsUI
Area-ShellExtension
Area-ShellExtension
Area-ShellExtension
Area-Suggestions
Area-Suggestions
Area-TerminalConnection
Area-TerminalControl
Area-Theming
Area-UserInterface
Area-VT
Area-Windowing
Area-WPFControl
AutoMerge
Blocking-Ingestion
Culprit-Centennial
Culprit-WinUI
Disability-All
Disability-Blind
Disability-LowVision
Disability-Mobility
External-Blocked-WinUI3
Fixed
Gathering-Data
good first issue
HCL-E+D
HCL-WindowsTerminal
Help Wanted
Impact-Compatibility
Impact-Compliance
Impact-Correctness
Impact-Visual
In-PR
InclusionBacklog
InclusionBacklog-Windows TerminalWin32
InclusionCommitted-202206
Issue-Bug
Issue-Docs
Issue-Feature
Issue-Feature
Issue-Question
Issue-Samples
Issue-Scenario
Issue-Task
Needs-Attention
Needs-Author-Feedback
Needs-Bisect
Needs-Discussion
Needs-Repro
Needs-Tag-Fix
Needs-Tag-Fix
Needs-Triage
No-Recent-Activity
Priority-0
Priority-1
Priority-2
Priority-3
Product-Cmd.exe
Product-Colortool
Product-Colortool
Product-Colortool
Product-Conhost
Product-Conpty
Product-Meta
Product-Powershell
Product-Terminal
Product-WSL
pull-request
Mirrored from GitHub Pull Request
 Resolution-Answered
Resolution-By-Design
Resolution-Duplicate
Resolution-External
Resolution-Fix-Available
Resolution-Fix-Committed
Resolution-No-Repro
Resolution-Won't-Fix
Severity-Blocking
Severity-Crash
Severity-DataLoss
spam
this-will-be-a-breaking-change
Tracking-External
WindowsTerminal_Win32
Work-Item
zAskModeBug
zInbox-Bug
No Label Issue-Bug Needs-Triage
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
claunia
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: starred/terminal#23706
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?