For multiple years, each wave of new devs has been asking roughly the same question over and over again:
"Why does the real device clear the screen before onUpdate() but not the simulator? This completely breaks my code where I optimize onUpdate() by only redrawing the part of the screen that changed!!!!"
And every time this happens, somebody has to explain that you have to redraw the whole screen in onUpdate(), because some devices clear the screen before onUpdate() and others don't. Furthermore, the system may draw a toast/popover on top of your app, which means you should clear the screen anyway (or at least that's been the conventional wisdom of forum, and I have little reason to believe it's wrong in this case).
Sometimes this leads to pages of arguments and bug reports saying that, actually, the behaviour of Connect IQ on the devices is wrong, that there must be some workaround to avoid this, or that there must be some scenarios where they can get away with not redrawing the whole screen.
I would submit that if the exact same issue comes up over and over again with new devs, then the best way to fix it is to change the behaviour on the sim, rather than doing the same thing and expecting new devs to react differently.
All of this could be avoided by having the simulator model the worst possible scenario for onUpdate(), which is that the screen is cleared beforehand. That way, devs would have no reason to believe they could get away with not redrawing the screen every time onUpdate() is called.
References:
(January 30, 2026) RE: Dc glitches black sometimes, not in simulator
Screen being cleared on onUpdate
Persistence of the Dc content between calls to onUpdate()
I'm sure there's many more that I missed
