Feat: native blink feedback

[alex-Arc]

#674

[thedist]

I'd recommend holding off on doing anything like a native blink feedback until after the graphics rework as that may change things.

[Julusian]

Thanks for looking at this, but my immediate thought is that this isn't something I want to tackle before the graphics overhaul is done.

I have 2 immediate concerns:

1. I expect that many users will find the choice to blink the background to black to be too limiting. I suspect that this will result in this needing to be expanded to allow configuring every style property. This concerns me as this patch of ui is already not very nice; I have been holding off adding more style options for a while as adding anything more makes it feel incredibly cluttered.
2. I don't currently see how we can translate this to the new drawing ability in the graphics overhaul. I believe that everything else translates well enough that we should be able to auto-migrate existing buttons. But I think that this blink largely goes against the new drawing ideas and will be very hard to migrate across without making a mess in some form.

The combination of those concerns has me leaning towards waiting until the graphics overhaul is finished (or abandoned). Until then, I hope that either a custom/expression variable or the generic-blink module are good enough (although something native would be nicer)

[alex-Arc]

Thanks both for the comments. I will take a look at the overhaul branch. I understand this might not be soon but will keep my hopes up 🙂

[Julusian]

@alex-Arc this PR got me inspired to have a play. I had an idea earlier of what if there was simply an expression function to do this blinking? That way it can be used in a feedback to change whatever properties the user wants, and that will fit into the graphics overhaul.

I hope you don't mind, but while I was feeling inspired I gave it a try, and I think it works quite nicely. #3831 I am a little worried about the performance cost of piping so many variable updates through the system to facilitate this, so I am thinking it wants to wait for some investigation into that. (Unless @thedist disagrees; you have more experience with large amounts or frequent variables updates)

[alex-Arc]

I am a little worried about the performance cost of piping so many variable updates

Yeah that was also my concern and the reason I moved it down into the render function