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

Text between box drawing characters is rendered aliased (???) #9009

New Issue
Closed
opened 2026-01-31 01:43:38 +00:00 by claunia · 10 comments
claunia commented 2026-01-31 01:43:38 +00:00
Owner
Copy Link

Originally created by @DHowett on GitHub (Jun 13, 2020).

Originally assigned to: @miniksa on GitHub.

When I run this in PowerShell (7),

"`e[30;47m`u{2500} What `u{2500}`e[m"

the text comes out without antialiasing. It looks like it's being drawn with FillGeometry just like a box drawing glyph.

image

If I flip the colors, it comes out antialiased.

image

To verify my theory that What was becoming part of the box drawing run, I put together a debug tool that turns glyph runs into graphviz documents.

image

Unfortunately, it doesn't look like it's so simple.

Originally created by @DHowett on GitHub (Jun 13, 2020). Originally assigned to: @miniksa on GitHub. When I run this in PowerShell (7), ``` "`e[30;47m`u{2500} What `u{2500}`e[m" ``` the text comes out without antialiasing. It looks like it's being drawn with FillGeometry just like a box drawing glyph. ![image](https://user-images.githubusercontent.com/189190/84561566-4e18a800-ad02-11ea-89b8-fe4cc58127bb.png) If I flip the colors, it comes out antialiased. ![image](https://user-images.githubusercontent.com/189190/84561569-58d33d00-ad02-11ea-83c0-56b763950ae4.png) To verify my theory that `What` was becoming part of the box drawing run, I put together a debug tool that turns glyph runs into graphviz documents. ![image](https://user-images.githubusercontent.com/189190/84561756-ff6c0d80-ad03-11ea-8e76-447cafa1dc83.png) Unfortunately, it doesn't look like it's so simple.
claunia commented 2026-01-31 01:43:39 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Jun 13, 2020):

It doesn't happen with midnight commander, which uses a lot of box drawing glyphs... because however MC is rendering, they get broken into different CustomTextLayouts.

image

image

@DHowett commented on GitHub (Jun 13, 2020): It _doesn't_ happen with midnight commander, which uses a lot of box drawing glyphs... because however MC is rendering, they get broken into _different CustomTextLayouts_. ![image](https://user-images.githubusercontent.com/189190/84561813-68538580-ad04-11ea-84d0-56147a81a335.png) ![image](https://user-images.githubusercontent.com/189190/84561816-6f7a9380-ad04-11ea-8cc4-b047c2c21478.png)
claunia commented 2026-01-31 01:43:39 +00:00
Author
Owner
Copy Link

@miniksa commented on GitHub (Jun 15, 2020):

what the what.

@miniksa commented on GitHub (Jun 15, 2020): what the what.
claunia commented 2026-01-31 01:43:39 +00:00
Author
Owner
Copy Link

@wackoisgod commented on GitHub (Jun 21, 2020):

Should be noted that i have the same issue in #6603 and when I run the same text posted above in 1.0 it renders correctly for me but in 1.1 it renders incorrectly as show by @DHowett

@wackoisgod commented on GitHub (Jun 21, 2020): Should be noted that i have the same issue in #6603 and when I run the same text posted above in 1.0 it renders correctly for me but in 1.1 it renders incorrectly as show by @DHowett
claunia commented 2026-01-31 01:43:40 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Jun 21, 2020):

@skyline75489 looks like this regressed with 94eab6e39, but I can't figure out why (since box drawing detection happens after the analysis code you changed!)

@DHowett commented on GitHub (Jun 21, 2020): @skyline75489 looks like this regressed with 94eab6e39, but I can't figure out why (since box drawing detection happens after the analysis code you changed!)
claunia commented 2026-01-31 01:43:40 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Jun 21, 2020):

Okay, this is because we're not populating _glyphClusters when text is "simple". When we go to split the run later (we split runs for things other than text analysis!), we try to remap the glyph clusters around the split.

Inevitably, we end up falling into DrawGlyphRun with run.glyphCount = 0 because _SplitCurrentRun stomped the glyphCount because there was no glyph cluster mapping.

Somehow magically we end up doing a "get raw geometry, draw" for the entire string which very handily hides the glyph clustering bug!

This fix works but I do not believe it is correct:

diff --git a/src/renderer/dx/CustomTextLayout.cpp b/src/renderer/dx/CustomTextLayout.cpp
index d19d2d0c3..395a58224 100644
--- a/src/renderer/dx/CustomTextLayout.cpp
+++ b/src/renderer/dx/CustomTextLayout.cpp
@@ -336,6 +336,7 @@ CustomTextLayout::CustomTextLayout(gsl::not_null<IDWriteFactory1*> const factory
             for (size_t i = glyphStart; i < _glyphAdvances.size(); i++)
             {
                 _glyphAdvances.at(i) = (float)_glyphDesignUnitAdvances.at(i) / designUnitsPerEm * _format->GetFontSize() * run.fontScale;
+                _glyphClusters.at(i) = static_cast<UINT16>(i);
             }

             run.glyphCount = textLength;

I'm worried about that "magical" rendering fallthrough. With this fix, are we striking the text twice -- once as a geometry and once as a glyph run?

(I commented out the glyph run drawing, and I think the answer is no.)

@DHowett commented on GitHub (Jun 21, 2020): Okay, this is because we're not populating `_glyphClusters` when text is "simple". When we go to split the run later (**we split runs for things other than text analysis!**), we try to remap the glyph clusters around the split. Inevitably, we end up falling into DrawGlyphRun with `run.glyphCount` = 0 because `_SplitCurrentRun` stomped the glyphCount because there was no glyph cluster mapping. Somehow *magically* we end up doing a "get raw geometry, draw" for the entire string which very handily hides the glyph clustering bug! This fix _works_ but I do not believe it is correct: ```diff diff --git a/src/renderer/dx/CustomTextLayout.cpp b/src/renderer/dx/CustomTextLayout.cpp index d19d2d0c3..395a58224 100644 --- a/src/renderer/dx/CustomTextLayout.cpp +++ b/src/renderer/dx/CustomTextLayout.cpp @@ -336,6 +336,7 @@ CustomTextLayout::CustomTextLayout(gsl::not_null<IDWriteFactory1*> const factory for (size_t i = glyphStart; i < _glyphAdvances.size(); i++) { _glyphAdvances.at(i) = (float)_glyphDesignUnitAdvances.at(i) / designUnitsPerEm * _format->GetFontSize() * run.fontScale; + _glyphClusters.at(i) = static_cast<UINT16>(i); } run.glyphCount = textLength; ``` I'm worried about that "magical" rendering fallthrough. With this fix, are we striking the text twice -- once as a geometry and once as a glyph run? (I commented out the glyph run drawing, and I think the answer is _no_.)
claunia commented 2026-01-31 01:43:41 +00:00
Author
Owner
Copy Link

@skyline75489 commented on GitHub (Jun 23, 2020):

I don't think this is correct, either. But right now I can not think of a better fix for this.

I kinda hate the fact that "we split runs for things other than text analysis". Is this a final thing or we are gonna change that in the future.

@skyline75489 commented on GitHub (Jun 23, 2020): I don't think this is correct, either. But right now I can not think of a better fix for this. I kinda hate the fact that "we split runs for things other than text analysis". Is this a final thing or we are gonna change that in the future.
claunia commented 2026-01-31 01:43:42 +00:00
Author
Owner
Copy Link

@skyline75489 commented on GitHub (Jun 23, 2020):

Per communication with @reli-msft I suggest fixing it by considering all runs containing box-drawing chars as complex. It's a quick fix. It should work as expected, with reasonable performance penalty.

@DHowett

@skyline75489 commented on GitHub (Jun 23, 2020): Per communication with @reli-msft I suggest fixing it by considering all runs containing box-drawing chars as complex. It's a quick fix. It should work as expected, with reasonable performance penalty. @DHowett
claunia commented 2026-01-31 01:43:42 +00:00
Author
Owner
Copy Link

@reli-msft commented on GitHub (Jun 23, 2020):

A proper fix should change _isEntireTextSimple into an enum to indicate how complex the text run is, and use that to determine the itemize passes to be enabled and whether we need "glyphing". For now, @skyline75489's fix would work and won't cause a lot of performance lost, since the frequency of box-drawing characters is low.

@reli-msft commented on GitHub (Jun 23, 2020): A proper fix should change `_isEntireTextSimple` into an enum to indicate _how complex_ the text run is, and use that to determine the itemize passes to be enabled and whether we need "glyphing". For now, @skyline75489's fix would work and won't cause a lot of performance lost, since the frequency of box-drawing characters is low.
claunia commented 2026-01-31 01:43:42 +00:00
Author
Owner
Copy Link

@DHowett commented on GitHub (Jun 24, 2020):

So, thinking a bit more about this.

I'm going to revert this in the 1.1 and master, so that we can bring it in correctly later. Folks have reported some crashes, and I'd prefer an expedient patch fix for servicing, and a fully-baked solution later 😄

@DHowett commented on GitHub (Jun 24, 2020): So, thinking a bit more about this. I'm going to revert this in the 1.1 and master, so that we can bring it in correctly later. Folks have reported some crashes, and I'd prefer an expedient patch fix for servicing, and a fully-baked solution later :smile:
claunia commented 2026-01-31 01:43:43 +00:00
Author
Owner
Copy Link

@ghost commented on GitHub (Jun 30, 2020):

:tada:This issue was addressed in #6665, which has now been successfully released as Windows Terminal Preview v1.1.1812.0.🎉

Handy links:

@ghost commented on GitHub (Jun 30, 2020): :tada:This issue was addressed in #6665, which has now been successfully released as `Windows Terminal Preview v1.1.1812.0`.:tada: Handy links: * [Release Notes](https://github.com/microsoft/terminal/releases/tag/v1.1.1812.0) * [Store Download](https://www.microsoft.com/store/apps/9n8g5rfz9xk3?cid=storebadge&ocid=badge)
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 Area-Rendering Issue-Bug Needs-Tag-Fix Priority-2 Product-Terminal Resolution-Fix-Committed
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#9009
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?