It is indeed becoming quite the longwinded thread. That being said, let me make another coffee and add some more to it 
First of all, let me express my gratitude for your engagement - not only in this thread but in others where I’ve seen you post aswell. I remember you posting about the STEP coloring app you’ve developed (and unfortunately not getting too much (positive) feedback about it if I recall correctly), but anyone developing their own solution to their own problems gets a big thumbs up in my book.
Also your ‘philosophy’ of helping out others not by giving them the entire ‘solution’ but by giving them just enough information to get nudged in the right direction and with a little brain engagement and commitment it’s possible for them to solve their problems themselves rather than getting spoon-fed a solution drew my attention and made you a noteworthy community member in my book 
That being said, we have slightly different problems and solutions to them.
Your app allows you to easily and quickly adjust the colors of already existing models. As you’ve demonstrated, your solution works great on a small scale. My goal however is to generate a ton of models with as much automation as is reasonably possible and directly getting my desired outcome without too much manual tinkering during modeling or afterwards. That is why I am so hesitant to manually color faces of models, as it doesn’t work on my scale.
To give you an impression, over the weekend I’ve created 730 (yes, seven hundred and thirty) models of surface mount resistors (a ‘mere’ 130MB in total). Only one manufacturer, one series of parts, 10 sizes, and I currently settled only for E12 values, although scaling up to E192 wouldn’t be much more effort and putting a bit of effort into the whole ordeal once and for all then reaping the benefits for the rest of my life is valuable in and of its own, but I figured I wouldn’t use 99% of the values in my lifetime anyway. Why did I go through this trouble? Because I don’t want to settle for ‘generic’ models. I really want the markings on all the models. I’ll spare you most of the reasoning but I really value having an accurate representation of the model. Knowing I need to solder a 22k resistor somewhere, and having a physical part with marking “223” on it is easy to link together, but I like to know that a part having a “01A” marking is a 10R resistor and vice versa. A similar reasoning applies for colored resistor rings for THT parts.
What I’ve done is modeled a single resistor model using the Spreadsheet workbench. With a bit of copy-pasting from my database of part parameters (painstakingly created by manually extracting all the relevent parameters from the datasheet), I can easily adjust the dimensions of the resistor, aswell as the marking on it, and have the correct part number embedded in the model. I think you can agree that your method of coloring the individual faces just doesn’t scale up to my level of requirements (which are, I must admit, are of ‘over 9000’ level OCD 
). That’s why I ‘insist’ on coloring the entire Body of the part marking in one go. One parametric base model → Hundreds of accurate parts with minimal (i.e. no) post-processing.
To give you an impression, I present to you all what I consider the most needlessly accurate representation of surface mount resistors you’ll ever see 
Royal Ohm, 0603 imperial size, only 10R - 82R. I now have 10R-10M in case sizes 2512 all the way down to 01005. The image shows the 100R resistors in all available sizes. I’m not satisfied with the marking being a little bit too wide for my taste on the 1206 and 1210 models, but that’s something for another day.
Royal Ohm - General Purpose Chip Resistors - 0603 - 10R-82R.FCStd (1.0 MB)
Now, the original goal of this thread was more to help the developers of KiCad and/or FreeCad, as I genuinely thought (and still think) there is a bug in either piece of software due to the exception that I described earlier. Yes, it’s a trivial edge-case, but I feel because the low likelyhood of someone catching these kinds of bugs, they are only more worthful to report. This does require, however, more effort on my end to write a very detailed report, as the devil is in the details here I’m afraid. Nonetheless, I agree with you that especially with FreeCad and KiCad, many roads lead to Rome. I was just wondering why at some intersection of FreeCad’s road and KiCad’s road to Rome, the path crumbles to pieces 
That’s enough for now on my end, your replies and feedback are much appreciated!
