Enormous CPU Draw -- Massive Memory Leak about spotube HOT 6 CLOSED

I just tried the Nightly version. I'm listening to it now, and I'm noting significantly less memory draw, on par with the official app. It's much more responsive, and far less laggy. The icon blinks a little bit in my tray at random intervals, but that is a microscopic complaint compared to the previous issue, and likely a problem on my end.

This is a vast improvement! Well done!

I'm actually new to programming. I just started learning at university, and we haven't started learning cache management yet. How do you find problems like this, and what are the steps you take to correct them?

from spotube.

Can you try out the nightly build and check if it was fixed there by any chance? Because there had been major re-writes of the internal cache manager.

Barring that, it may be time to rewrite this from scratch, possibly in a low-level, memory-safe language like Rust, Go, or Java.

I'm pretty sure that will cut off half of the platforms supported by this project. The issue you're facing could be a rare case, because I myself use Linux Mint 21.3 (latest ig), and yet I haven't found any memory leaks.

from spotube.

I'm pretty sure that will cut off half of the platforms supported by this project.

it won't cut off mac, windows, or linux for sure, but Rust and android just seems to be yew.rs or tauri (essentially webview)

from spotube.

I'll give it another try. Thank you. The version causing the most trouble was the Flatpak edition. If you'd like, I can supply logs and screenshots. Tell me what commands you'd like me to display, and I'll give you all the info you need.

from spotube.

Can you try out the nightly version? The cache management was changed significantly like I said before. I want to make sure we're not tracking an already fixed bug in nightly

from spotube.

I'm actually new to programming. I just started learning at university, and we haven't started learning cache management yet. How do you find problems like this, and what are the steps you take to correct them?

Here's the secret, we don't. We simply re-write the entire project /s

But on serious note, user usually complaints and there's usually debuggers (devtools for flutter stack) that show memory usage waterfall or allocation based graphs (scatter graphs usually). From these graphs/reports/dumps we can learn about what instances of what Class and with what dependencies are allocating memory more frequently and not getting Garbage Collected for a long time (singleton classes are ignored in this case). Also, we can see the same for methods/functions for a more granular report.

But the tricky ones are the CPU spikes. They can be hard to find, especially the spikes that are happening at random and depends on ridiculous but frequent scenarios to happen. It's like finding a needle in a haystack.

Don't get confused by all these terms and analogies. Just keep on learning, after sometime, you'll know it way better than ever.

from spotube.