[Bug]: When maximizing a non-focused block, the block border is set incorrectly.

[ZLATAN628]

Current Behavior

When I maximize a non-focused block, the block border is set incorrectly. I tried debugging the source code to find the answer and discovered that the width and height are first assigned the correct values but are then updated back to their original values.

Expected Behavior

Steps To Reproduce

Clicking the maximize button on the non-focused block

Wave Version

main

Platform

macOS

OS Version/Distribution

macos

Architecture

arm64

Anything else?

No response

Questionnaire

• I'm interested in fixing this myself but don't know where to start
• I would like to fix and I have a solution
• I don't have time to fix this right now, but maybe later

[esimkowitz]

I'm not able to reproduce this, could you record a video? Also how are you magnifying the block, using the context menu item or the button in the block header? You mention that the block is not focused, but either of these methods for magnifying the block should focus it first so a video would really help understand what's happening.

Also is this only in the dev version where you see this or also in the release version? Also do you notice this with all the widgets or just the web widget? Sorry for all the questions 😅

[ZLATAN628]

Hi @esimkowitz ,Please watch this video. I have tried the latest release and dev versions and this problem exists in all widget.

QQ20241227-081306-HD.1.mp4

[esimkowitz]

That's so strange, this does not reproduce on my system...

[ZLATAN628]

I am not very good at front-end. I found that the event of updating the window size is triggered twice. The first one should be triggered by the focus event, so it enters the setInnerRectDebounced delayed update, and the second one is triggered by maximizing the window, but it is directly updated immediately here. So the performance is that after maximizing the window, it is normal at first, and then because of the delayed update, it is updated back to the window size before maximization.

QQ20241227-083529-HD.1.mp4

[esimkowitz]

Ahhh yes that makes sense. We should cancel the denounced update when a newer one comes in

[ZLATAN628]

Hi @esimkowitz , I believe I’ve identified the issue. During the maximize update, clearInnerRectDebounce tries to clear the previously set timeout update event. However, it fails to retrieve the timeout event, so it doesn’t get cleared, which causes the timeout event to still trigger. I have created a PR #1702 that uses a global Map to address this issue.