Would anybody want me to expand my Numix icon selections?

Recently I've built this monstrosity of an icon suite which while only being ~40MB, gives a total of 108 icon theme selections to choose from. (It also requires the mandatory inclusion of non-themes which could still be chosen, since they reside in the same space for theme inheritance.)

I've been wondering; Should I add more to it? More colours? More styles provided by numix-folders? Or should I just quit while I am ahead and let people use numix-folders to define their own preference as the Numix team intended?

I've ironed out the issues I've encountered previously, so expansion itself isn't the problem, bt it will be a significant time investment as I run the routines in numix-folders hundreds of more times, possibly requiring the creation of a new repository to keep everything organized.

I'm sure any number of people would, because while some people are happy with anything that's roughly the right hue, some are incredibly fussy and will tweak a channel by about 2 points just to get to the "ideal" version they want. :slight_smile:

However, as long as you've got the source layers AS layers, wouldn't it make a lot more sense to simply distribute those along with the script that hue-shifts and composites them? (Or is that what the numix-folders you mention does?). That way you wouldn't have to take requests, and everyone could tweak things to exact shades they want.

numix-folders does just that. Uses imagemagick and other tricks with custom images to adjust up to three colours for most styles. Procedure is fairly simple; Copy base themes a bunch into specific directories, glob-link into ~/.icons a set, run numix-folders, glob-delete links, rinse and repeat for each variant, then for what I've distributed glob-delete light variants, delete duplicate images via script, glob-delete instances of index.theme, rename icon themes via script, mass-copy via script template index.theme files, glob-open in a text editor said templates, replace with colour name, glob-link into ~/.icons all icons to test, debug, archive, then distribute.

All of that can be done in a few hours. What's really important for how I'm doing things is to have common images used across every variant ever as separate themes and call from those, whenever necessary. The reason I ask is because I]ve got nothing better to do, but i want to evaluate interest before I spend eight hours cranking out set after set for specific combinations like what I've done for Materia / MateriaV2 and Mint-Y.

Now I think about it, could be better to just clean up and publish my scripts so other people can do this if they have multiple variants and want it all in a small amount of space.

That's my take on it, yes. Distributed workloads are a very nice thing.

If you're bored and looking for something to do, I'd be eternally grateful if you felt like applying the same techniques to a few other theme pieces, like the "Traditional" controls... :stuck_out_tongue:

Because right now, those don't have the sources needed to actually generate the "hover" PNGs etc, and I haven't been able to find them online anywhere. Those are the last missing piece for being able to restore color choice functionality in themes, which you've probably noticed is a fervent wish of mine for the 20.04 release. I can handle everything else, but I SUCK beyond imaginable levels when it comes to doing image work, and with nothing to composite from those pieces are always going to look out of place even though the rest of the theme will adjust the way users want it to.

@tiox: Yes, please! Go ahead! I feel like this shouldn't be a separate topic though..