Awesome find! I was not considering that a promise is executing as soon as it is created, and this is why the implementation still results in parallel execution.
I've gone back to the drawing board, and I've updated my post accordingly with a better solution using reduce. I've also used your demo above as an example and credited you in my article (please let me know if that is not OK, and I will take it down).
As long as the displays are plugged directly into the eGPU gaming shouldn't be an issue; the setback arises if you want to use an internal display. Same goes for iMacs and other computers with built-in high-DPI screens: eGPUs are not allowed to drive them.
It seems like a good idea, but of course its not running the latest version of Android. Most of the problems with Android stem from security updates not reaching devices in a timely manner, and RemixOS is just another example of a skin that is behind. Most devices are only on RemixOS 1.5, which is Android 5.0.2.
It's definitely a tradeoff that has led to some problems. However, I think it's important to remember Bookshelf's flexibility, and that the tradeoffs are more language-level than library-level.
For the UI. Atom is a significant update in user interface, and while there are performance dips (mainly at startup or using complex packages), the overall experience of using Atom has improved my workflow.
Interesting. I'll check it out, but unless I can get the same level of integration, super-fast file loads and searching, and parity with packages... I'm likely to not be a convert.
I've gone back to the drawing board, and I've updated my post accordingly with a better solution using reduce. I've also used your demo above as an example and credited you in my article (please let me know if that is not OK, and I will take it down).
Thanks again!