Delayed reaction on mouse wheel spin on sliders

[AxelG-DE]

Describe the bug

Background info: I use a Logitech MX Master 2S mouse. <-- has a scrollwheel without ticks=free spinning

When I use this function e.g. mouse-hover on WB slider and spin the wheel,

• currently the slider will stay at the initial place,
• no screenupdate and
• only when the spinning speed got way slower or -- I stop the wheel, the GUI gets updated, -- means the slider jumps on 0sec flat to that new position and you see the WB change on the pic

Expected behavior

smooth GUI reaction and screenupdate as it was

Edit 1 (added not) #4820 might not be related

bisect results I besected from dt-3.1.0+5-gcd01ccf7a I did my best to judge well, but I confess, this bisect result makes no sense to me (in between I was forced to skip two times, as I could not build that version)

d6a9dad030fb06565ebf73b260f8da71fcfc534c is the first bad commit
commit d6a9dad030fb06565ebf73b260f8da71fcfc534c
Author: phweyland <[email protected]>
Date:   Wed Apr 8 10:31:56 2020 -0300

    fix $(EXIF_EXPOSURE) issue on watermark

 src/iop/watermark.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--> is there any file from .git folder might help ?

@phweyland may I ping you here, even with my total noob understanding, I cannot imagine, what that commit above should have to do with it

Platform (please complete the following information):

• Darktable Version:
• OS: Gentoo Linux
• OpenCL: on and off, both show error
• GPU 0 Nvidia GTX1060 OC
• GPU 1 Nvidia RTX2070 Super
• nvidia-drivers-440.82

Additional context

• Are the steps above reproduce with a fresh edit (removing history)? yes (more obvious if expensive modules are activated)
• Did you compile Darktable yourself? If so which compiler was used, with what options? gcc 9.3.0 | -march=native -O3 -mtune=native -pipe
• Is the issue still present using an empty/new config-dir yes

[phweyland]

You mentioned the type of mouse... have you noticed a different behavior with another type ?

[AxelG-DE]

Hihihi I don't have another one, which has this kind of function. With click you will never be able to spin that fast :-)

I can just say,

• it once worked well
• the PC does recognize the spinning speed, as I can rev to max in terminal or firefox
• it really feels, there is a kind of limiter, where the max speed of ticks has a cap value, because once the wheel gets slower the reaction comes back.

...please get me right! I am not playing for fun, how fast I can spin the wheel. It is really the normal workflow, that I kick the wheel for WB adjustment. Just for this issue, I checked a bit harder, how far can I go, in order to describe better :)

[AxelG-DE]

I tried to bisect this again and find only bullshit because it is floatingly getting worse and worse...

When you e.g. try 2.7.0+1525~g21ac343cd, you can see, it is way more fluent and swift compared to todays current master...

I am willing to support in analysis, but need guidance, what to do next.

@phweyland sorry I bothered you @parafin do you have an idea, what I could do to figure out?

I am on and off whether #4820 is related or not...........

Sorry for again being a perf PITA

[phweyland]

sorry I bothered you

You don't. I'm just worried because the use case you have described above works perfectly for me.

Could you share an xmp for which you have this issue especially visible (same for #4820 ) ?

[parafin]

No idea here. sorry.

[github-actions[bot]]

This issue did not get any activity in the past 30 days and will be closed in 365 days if no update occurs. Please check if the master branch has fixed it and report again or close the issue.

[AxelG-DE]

@phweyland sorry somehow I overlooked your wish for an xmp...

I just tested again, the issue is still there, for both caes, this one and #4820

(tooo big to zip in here, as inclued the RAW) https://fromsmash.com/Mouse-delay

The xmp-only is here: 6402.xmp.zip

[AxelG-DE]

@aurelienpierre id that bot a bit over nervous? Issue from May 1st get no-issue-activity already 21st June? (9 days ago today)

[github-actions[bot]]

This issue did not get any activity in the past 60 days and will be closed in 365 days if no update occurs. Please check if the master branch has fixed it and report again or close the issue.

[AxelG-DE]

Cannot test atm :-/

[dterrahe]

I assume this is still an issue for you?

[github-actions[bot]]

This issue has been marked as stale due to inactivity for the last 60 days. It will be automatically closed in 300 days if no update occurs. Please check if the master branch has fixed it and report again or close the issue.

[jenshannoschwalm]

Very likely fixed on 4.4 so closing this for now.

[AxelG-DE]

sorry, that I reply with huge delays. Life is still in my way (and I wonder, when would it stop and things come to normal)

If I go very extreme, I can make color calibration still behave as described here, but that is unrealistic brutal. For normal usage, I can confirm after a few quick tests, that it works as expected now.

[jenshannoschwalm]

Thanks for confirming :-) Haven't seen you here for a long time ?

[AxelG-DE]

@jenshannoschwalm yeah health status of me, my wife and other troubles keep me away unfortunately. I do IRC but also very limited... (sorry for OT)

[jenshannoschwalm]

Ich wünsche euch einfach alles Gute!