speed tests stupid?

[whateverx2019]

alright i just got the p30 pro (6 GB) and I ran my own speed test where I load a bunch of my commonly used apps (starbucks, uber, whatsapp, chrome, etc.)

sure, when i re-open these apps immediately after loading them (like 15 total), they're all still in memory, but I find that over time, they need to be reloaded even tho I'm not really opening new apps. just seems to be like these speed tests are stupid because the OS, over time, closes inactive apps.

yes, I disabled Huawei's aggressive app killing by going to "battery:" -> "app launch" and disabled "manage all automatically"

OR is it because I "only" have 6 GB of RAM? Would getting the s10 Plus with 8 GB of RAM (or p30 pro 8 GB) make a difference? i mention Samsung cause I read soemwhere it has better RAM management than huawei.

 

[ocnbrze]

what speed tests are we talking about? most that i know are for upload and download speeds. i do not see how you can measure how responsive a phone can be.

 

[whateverx2019]

my bad, i;m talking abouit those speed tests on youtube where they run through a bunch of apps and then see if theyre still in memory

 

[ocnbrze]

meh yeah sounds dumb then. i have not seen any of those....just do not see the point in it. as long as your phone runs smoothly........who cares?

6gb of ram is fine. the only thing is if you have like 10 apps open all at once, of course your phone will slow down.

 

[Hadron]

Tests of that sort - if done properly - will measure different things.

If you start from a clean state, all apps closed (and make sure they're closed, not just re-loaded in the background) then run through a series of apps you are testing a mixture of how quickly they load and how quickly they run. If instead you load a bunch of apps up, run through them and then run through them again you are testing the phone's memory management, how many of that set of apps it can keep in memory in the short term (I say memory management because that's frequently more important than the amount of RAM installed: to use a personal example, my Samsung tablet has 1.5 times the RAM of my old HTC phone, but the phone multitasked better than the tablet. Sticking very large amounts of RAM in is usually more a marketing point than a genuine benefit).

Both of those tests are artificial because nobody uses a phone that way, but there's some validity to them. Testing how long apps go without being reloaded is much trickier to control: the phone is doing a lot of stuff of its own, not just the apps you load, so you can't be sure it didn't need that RAM for something else in the meanwhile. Conversely it will learn your usage, so will tend to pre-load apps that you use a lot, and that may also affect which ones it keeps and which is closes. So it's very hard to compare that sort of behaviour between two phones because it will depend on their histories.

Of course if they change memory management settings as you have then it's not a valid test because your average buyer is never going to find that setting.So for someone doing a comparison on YouTube it's a matter of knowing what you want to test, devising the protocol to make that a fair and consistent test, and explaining to the viewer what exactly it is you have done so they can understand what the test means.

But it all sounds more valid than "drop tests", which are just anecdotes done for entertainment value and have very little to say about real world durability. Take a hundred copies of each phone, drop each one in "the same way" and see that fraction survive, and then I'll take those things seriously. Take one phone, drop it one or 3 times (so it's largely a matter of the luck of how exactly it lands), continue dropping it after it's already compromised, that type of nonsense tells you slightly less than using it as the subject of a "will it blend" video (which at least tell you that the blender in question can destroy the phone).