They have audio. If you aren’t hearing it, make sure your ringer switch isn’t set to silent/vibrate.
Huh, well I perpetually leave mine set that way for… reasons.
Looking at this video:
Looks like the LG V30 is actually quicker in most things over the Iphone 8+ in real world scenarios apart from the loading of games. The difference is pretty immaterial though in either case.
After listening to @wumpus on this issue for so long, that was quite shocking for me to see. WTF wumpus?
Well many have been saying here for a long time that these synthetic benchmarks are ass and make little to no difference in real world use cases, or do so to such a little extent as to make the difference irrelevant for the majority of buyers.
Yes and that is obviously true. I think his point is that performance gains enable new types of applications in the future. And maybe that’s true, and there are Discourse features he can’t build due to Android holding the market back, or it makes his developers’ jobs harder as they’re forced to spend most of their time optimizing code rather than adding new features.
Is it an app launch test? Those should be close by now, they primarily test:
- flash (disk drive) speed
- available memory (particularly on app switching)
Overall CPU speed is a factor, definitely, but Android should be able to close the gap in basic app load and switching. If they have, that’s good. Games might still be pretty bad though, per Tom’s Hardware:
The iPhone 8 also opened more demanding apps faster than its predecessor and the top Android phone right now, although those differences were less dramatic. It took the iPhone 8 11 seconds to fully load the Injustice 2 game, compared to 14.53 seconds for the iPhone 7 Plus and 19 seconds for the Note 8.
19 seconds versus 11 seconds is almost 2x as fast.
Remember too that in theory Android should have a huge advantage in app switching, as even the crappiest Android devices have 3 - 4 GB of RAM and Apple is still miserly as hell with RAM on phones: only 2GB on the regular iPhone and 3GB on the fancier plus models.
I found the most interesting test that of loading up web pages, the Android device came out on top in this test every time.
Weird, per the video it looked like DNS resolution times were off, since it took forever to even get the websites to begin the process of loading – they weren’t loading at all on the Apple device, just many seconds sitting waiting staring at the stalled address bar for the site to resolve. I do know Apple devices are super horny for ipv6, so if his local network had ipv6 issues, or the ipv4 worked but ipv6 did not, you’d see extreme weirdness.
I tend to favor the guys who run the webpagetest service here:
Wumpus, entertain my curiosity. Can you do your own test comparing your one plus 5 with your latest iPhone?
I’d like to see some real world results.
I was thinking that’d be a good idea since I do have both devices! And we could test with this very forum. I will do, but bear in mind we cut the work in half for Android (we load, say, 10 posts instead of 20) due to the historic performance issues.
This is actual code in Discourse
def slow_platform? request.user_agent =~ /Android/ end
Pick some random sites then
There were so many cut aways in that video that I’ll take it with a grain of salt. Especially when the video is made to prove a point. As opposed to the scientific method which is to go where the data leads.
Based on the oneplus5 video I already saw (and linked to above – it’s months old) I do expect Android to catch up on the basic app load and re-load scenario. This is not really contingent on raw CPU speed (it’s flash drive storage speed and RAM) and it is something they can fix in software. It should be close by now.
What do you make of this? @wumpus
For cold launch, this is definitely true. For warm launch (warm caches, that is), it’s more subtle. In that scenario, where the time goes is much more software related. Static initializers and other on-load functions, shared library load and symbol binding, ASLR rebase, etc, all take time. You’re still correct that it’s not raw CPU speed that’s the main factor, of course, but it gets pretty interesting where the time goes. There’s a lot of variance from app to app depending on coding styles, libraries used, etc…
For iOS devices, there’s a fair bit of detail in some WWDC talks from the last few years:
edit: Here’s a good blog post about how to use that sort of data to drive down an app’s startup times.
The Qualcomm-hole strikes back!
Don’t Innovate, Litigate!
Both sides have vast patent portfolios. They had an agreement and Apple hit first, breaking it. Now the lawyers will sink their teeth in.