starred/terminal
1
0
Fork 0
mirror of https://github.com/microsoft/terminal.git synced 2026-02-04 05:35:20 +00:00
Code Issues 19.0k Packages Projects Releases 20 Wiki Activity

clear screen ansi doesn't prevent previous screen from being pushed to scrollback buffer #23198

New Issue
Closed
opened 2026-01-31 08:35:07 +00:00 by claunia · 10 comments
claunia commented 2026-01-31 08:35:07 +00:00
Owner
Copy Link

Originally created by @ldemailly on GitHub (Apr 24, 2025).

Windows Terminal version

1.21,1.22,1.23

Windows build number

10.0.26100.3775

Other Software

none though real case is fortio brick golang ansipixels demo

Steps to reproduce

Write-Host "$([char]27)[2J$([char]27)[1;1Hone`r`nprevious`r`n$([char]27)[2J$([char]27)[1;1Htwo`r`n"

or when comparing on unix/mac

echo -e "\033[2J\033[1;1Hone\r\nprevious\r\n\033[2J\033[1;1Htwo\r\n"

Expected Behavior

nothing in scroll back

Actual Behavior

one
previous

in scroll back

Originally created by @ldemailly on GitHub (Apr 24, 2025). ### Windows Terminal version 1.21,1.22,1.23 ### Windows build number 10.0.26100.3775 ### Other Software none though real case is fortio brick golang ansipixels demo ### Steps to reproduce ```powershell Write-Host "$([char]27)[2J$([char]27)[1;1Hone`r`nprevious`r`n$([char]27)[2J$([char]27)[1;1Htwo`r`n" ``` or when comparing on unix/mac ```bash echo -e "\033[2J\033[1;1Hone\r\nprevious\r\n\033[2J\033[1;1Htwo\r\n" ``` ### Expected Behavior nothing in scroll back ### Actual Behavior ``` one previous ``` in scroll back
claunia added the Needs-TriageIssue-Bug labels 2026-01-31 08:35:07 +00:00
claunia commented 2026-01-31 08:35:08 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 24, 2025):

Initially mentioned in https://github.com/microsoft/terminal/issues/18825#issuecomment-2822778126 - cc @lhecker

@ldemailly commented on GitHub (Apr 24, 2025): Initially mentioned in https://github.com/microsoft/terminal/issues/18825#issuecomment-2822778126 - cc @lhecker
claunia commented 2026-01-31 08:35:09 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 24, 2025):

note: I'm partially wrong: ghostty doesn't push to the scroll back with the above, but apple terminal does indeed
(interestingly, apple terminal with a black full rest of the screen while windows terminal only has the 2 lines right above)

alacrity does same as windows terminal
kitty does same as ghostty

so 2 against 3 :) ? what should I do in my program to not get stuff pushed and yet not use the alternate screen?

@ldemailly commented on GitHub (Apr 24, 2025): note: I'm partially wrong: `ghostty` doesn't push to the scroll back with the above, but apple terminal does indeed (interestingly, apple terminal with a black full rest of the screen while windows terminal only has the 2 lines right above) alacrity does same as windows terminal kitty does same as ghostty so 2 against 3 :) ? what should I do in my program to not get stuff pushed and yet not use the alternate screen?
claunia commented 2026-01-31 08:35:09 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Apr 24, 2025):

So! You've stumbled onto one of the great things that terminal emulators (and their authors!) disagree on!

Some of them, like gnome-terminal (libvte), Apple Terminal, and Windows Terminal implement CSI 2 J (ED Erase in Display All) as "push the contents of the viewport into scrollback". There is some disagreement on whether it should be only the actual used lines or the whole viewport.

Others (xterm, rxvt) implement it as (I believe) "destroy the current screenful of content."

There's been some other discussion about this:
https://github.com/microsoft/terminal/issues/7894 (summary here: https://github.com/microsoft/terminal/issues/7894#issuecomment-714854915)
https://github.com/microsoft/terminal/issues/2832#issuecomment-533779605

Consensus among application authors is that they should just clear the scrollback on those terminals which require it. If you look at the output of clear, for example, on gnome-terminal: \e[H\e[2J\e[3J.

This is likely why clear also sports an argument to "try not to clear the scrollback", -x.

Image

@DHowett commented on GitHub (Apr 24, 2025): So! You've stumbled onto one of the great things that terminal emulators (and their authors!) disagree on! Some of them, like gnome-terminal (libvte), Apple Terminal, and Windows Terminal implement `CSI 2 J` (ED Erase in Display All) as "push the contents of the viewport into scrollback". There is some disagreement on whether it should be only the actual used lines or the whole viewport. Others (xterm, rxvt) implement it as (I believe) "destroy the current screenful of content." There's been some other discussion about this: https://github.com/microsoft/terminal/issues/7894 (summary here: https://github.com/microsoft/terminal/issues/7894#issuecomment-714854915) https://github.com/microsoft/terminal/issues/2832#issuecomment-533779605 Consensus among application authors is that they should just clear the scrollback on those terminals which require it. If you look at the output of `clear`, for example, on gnome-terminal: `\e[H\e[2J\e[3J`. This is likely why `clear` also sports an argument to "try not to clear the scrollback", `-x`. ![Image](https://github.com/user-attachments/assets/a00811f9-f6f7-45ec-b8d4-972b600c3756)
claunia commented 2026-01-31 08:35:09 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 24, 2025):

Thanks a lot for the detailed response and context.

But using 3J deletes the whole scrollback, that's not what I aim to do, I aim for the users to not lose their previous scrollback and yet see the output of my program when it's done and also not flood their scrollback

That works as I expect already with some terminals; what can I do to achieve ^ with windows terminal?

@ldemailly commented on GitHub (Apr 24, 2025): Thanks a lot for the detailed response and context. But using 3J _deletes_ the whole scrollback, that's not what I aim to do, I aim for the users to not lose their previous scrollback and yet see the output of my program when it's done and also not flood their scrollback That works as I expect already with some terminals; what can I do to achieve ^ with windows terminal?
claunia commented 2026-01-31 08:35:09 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Apr 24, 2025):

Hmmmm....

Maybe if you home the cursor with CSI H and then CSI 0 J to erase below?

@DHowett commented on GitHub (Apr 24, 2025): Hmmmm.... Maybe if you home the cursor with `CSI H` and then `CSI 0 J` to erase _below_?
claunia commented 2026-01-31 08:35:10 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Apr 24, 2025):

(Or if you're redrawing the whole screen, can you just home the cursor with CSI H and just... overdraw everything? No need to clear!)

@DHowett commented on GitHub (Apr 24, 2025): (Or if you're redrawing the whole screen, can you just home the cursor with `CSI H` and just... overdraw everything? No need to clear!)
claunia commented 2026-01-31 08:35:10 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 24, 2025):

Genius:

echo -e "\033[H\033[0Jprevline1\033[2;1Hprevline2"; sleep 3; echo -e "\033[H\033[0Jtwo\r\n"

appears to work!

(Or if you're redrawing the whole screen, can you just home the cursor with CSI H and just... overdraw everything? No need to clear!)

Yeah I do that when most of it is overwritten; eg

curl --compressed https://demo.fortio.org/fire

but for some case where the output is "sparse" it's less bytes to clear the screen and only write where needed, eg

curl --compressed https://demo.fortio.org/life

but if I can rely on H+0J as a replacement for 2J on all terminals, I think I'll switch to that (or rather switch to that after the first clear screen so users don't loose their current screen either)

@ldemailly commented on GitHub (Apr 24, 2025): Genius: ```sh echo -e "\033[H\033[0Jprevline1\033[2;1Hprevline2"; sleep 3; echo -e "\033[H\033[0Jtwo\r\n" ``` appears to work! > (Or if you're redrawing the whole screen, can you just home the cursor with CSI H and just... overdraw everything? No need to clear!) Yeah I do that when most of it is overwritten; eg ```bash curl --compressed https://demo.fortio.org/fire ``` but for some case where the output is "sparse" it's less bytes to clear the screen and only write where needed, eg ```bash curl --compressed https://demo.fortio.org/life ``` but if I can rely on H+0J as a replacement for 2J on all terminals, I think I'll switch to that (or rather switch to that after the first clear screen so users don't loose their current screen either)
claunia commented 2026-01-31 08:35:10 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Apr 24, 2025):

(These demos are amazing! I also really like your fps tool.)

@DHowett commented on GitHub (Apr 24, 2025): (These demos are amazing! I also really like your `fps` tool.)
claunia commented 2026-01-31 08:35:10 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 24, 2025):

Thanks a lot (and thanks for the compliments), will use that strategy in fortio/terminal#103

@ldemailly commented on GitHub (Apr 24, 2025): Thanks a lot (and thanks for the compliments), will use that strategy in [fortio/terminal#103](https://github.com/fortio/terminal/issues/103)
claunia commented 2026-01-31 08:35:10 +00:00
Author
Owner
Copy Link

@ldemailly commented on GitHub (Apr 25, 2025):

ps: deployed that new strategy so now the life curl doesn't push every frame into the scroll back anymore while also pushing the initial previous screen... best of both world (ditto go run fortio.org/terminal/brick@latest)

@ldemailly commented on GitHub (Apr 25, 2025): ps: deployed that new strategy so now the life curl doesn't push every frame into the scroll back anymore while also pushing the initial previous screen... best of both world (ditto `go run fortio.org/terminal/brick@latest`)
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#23198
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?