Each derived Screen class had a background and a foreground. Recently a shared foreground was considered for FadeToWhite, and doing so would have meant re-implementing the feature across all screens. This PR moves the background into Base, and adds a foreground as well (of type AllegroFlare::Backgrounds::Base* for now).
(I also considered writing a wrapper class but it would have meant some excess "design" downstream.)
Remove ElementID parent from Screens/Base
From a somewhat old design, Screens/Base had a parent class of ElementID. Within AllegroFlare, that parentage was never being used. Any properties have been much better managed with actual members on the class for a while now. A complete rebuild returned no instance of dependence on the ElementID parent with all tests passing. FadeToWhite was then rebuilt with no issues and all tests passing.
Some things to keep an eye on:
✅ FadeToWhite was rebuilt with no issues
⚠️ Other downstream projects may rely on the ElementID parent, but I would wager there are actually none.
⚠️ Other downstream projects may rely on #includes that were a part of ElementID. An missing symbol for std::cout surfaced, so it's possible most dependencies are trivial.
Add
backgroundandforegroundEach derived Screen class had a
backgroundand aforeground. Recently a shared foreground was considered for FadeToWhite, and doing so would have meant re-implementing the feature across all screens. This PR moves thebackgroundinto Base, and adds aforegroundas well (of typeAllegroFlare::Backgrounds::Base*for now).(I also considered writing a wrapper class but it would have meant some excess "design" downstream.)
Remove
ElementIDparent fromScreens/BaseFrom a somewhat old design,
Screens/Basehad a parent class ofElementID. WithinAllegroFlare, that parentage was never being used. Any properties have been much better managed with actual members on the class for a while now. A complete rebuild returned no instance of dependence on theElementIDparent with all tests passing. FadeToWhite was then rebuilt with no issues and all tests passing.Some things to keep an eye on:
ElementIDparent, but I would wager there are actually none.#includes that were a part ofElementID. An missing symbol forstd::coutsurfaced, so it's possible most dependencies are trivial.