Closed saugon closed 1 week ago
I've tested two Android devices with @tkz00 and they still crashed when using two hack ultis at the same time. I will keep on testing with an IOS build as well
I’ll change the PR to draft. Tomorrow, I’ll try to use a phone that still has issues. If I can’t, we can still merge this PR because it also improves the performance in general.
Talked with @saugon how to review this best:
- The haptic feedback for damaging an enemy or crate is completely gone in my android phone (Samsung S10+)
Good catch, i will fix that, thanks!
- Also it didn't get any better in respect to performance, but I don't know if my problem is the one being targeted in this PR.
That specific issue is related to the VFX and being worked on in another other PR.
The code reorganization is a great improvement, much easier to follow, also the logic to avoid running the haptic feedback multiple times works great. That said, I'm not sure I can see a performance difference between the current state of the game and this branch, based on the fps metric on top of the screen, both still have my screen flickering issue, but it has been explained that this PR doesn't aim at solving this issue. What do you think?
The difference could be seen in two ways:
@AngieDutra has already measured the performance improvement (though she hasn’t commented here yet). If this is blocking you due to the devices you have access to, don’t worry about it and let’s continue with the review. This PR is blocking others that rely on changes made in the character prefabs.
These are some comparisons to show how many times the haptic feedback has been executed in the same situation using prints.
@tkz00 Nice test! I already noticed this issue, but we have to wait until Manu’s update on player movement and actions is merged. Once that’s done, we can trigger the feedback only if there has been a change, instead of on every frame.
As @saugon said, I reviewed the performance improvements yesterday in a Redmi Android device. I'll review asap the last additions and use the profiler this time to check further
Motivation
We detected some performance issues on some low-end devices when H4ck's ultimate was triggered. The main issue was an incorrect haptic triggering that was running multiple times. Additionally, I refactored the color overlay implementation. It’s now cleaner, more lightweight, and optimized, though the performance impact isn’t as significant as the haptic fix.
Summary of changes
How has this been tested?
The performance should not be affected by the skill.
Checklist