The point is, there's a pattern that consists of different images. We need to put it on the screen. Using DP doesn't help, because it's starting to get out. He wrote his class, which recycles pictures of picsel, which means he's taking a co-signer for the screen on standard images and sausages and rewriting them. But this approach has a lot of operational memory. Who's gonna tell you what?
I think you're too easy and simple to say goodbye to dp, and this unit of measure needs a little more attention to make it work as you want.
I recommend that we take into account the scattered principles that are used by resource modifiers and this is indeed very convenient:
1. The size of the screen.
2. Ekran's density.
4. screen orientation.
You can open the DashboardAndroids and see what parameters are now more popular and work out of them, pre-emptively taking most of the market.
Moreover, without using this approach, you are distracted by many design principles. Because on different screens, on different devices, the interface has to look identical, but it sometimes takes almost all the elements to move or even add new ones. (phone/tablet/tv/wear) Which is also easy to take into account the modifiers, to choose for you.