New Benchmark Shows iPhones Throttle So Hard They Lose Their Edge Over Android (hothardware.com) 133
MojoKid writes: Apple has repeatedly asserted its dominance in terms of performance versus competitive mobile platforms. And it has been historically true that, in cross-platform benchmarks, iPhones generally can beat out Android phones in both CPU and GPU (graphics) performance. However, a new benchmark recently released from trusted benchmark suite developer UL Benchmarks sheds light on what could be the iPhone's Achilles' Heel in terms of performance, or more specifically, performance over extended duration.
The new benchmark, 3DMark WildLife, employs Apple's Metal API for rendering and Vulkan on Android devices. In testing at HotHardware, for basic single-run tests, again iPhones trounce anything Android, including flagship devices like the Samsung Galaxy S20 Ultra, ASUS ROG Phone 3 and OnePlus 8. However, in the extended duration WildLife Stress Test, which loops the single test over and over for 20 minutes, the current flagship iPhone 11 Pro and A13 Bionic's performance craters essentially to Snapdragon 865/865+ performance levels, while Android phones like the OnePlus 8 maintain 99+% of their performance. Though this is just one gaming benchmark test that employs the latest graphics technologies and APIs, it's interesting to see that perhaps Apple's focus on tuning for quick bursty workloads (and maybe benchmark optimization too?) falls flat if the current class of top-end iPhone is pushed continuously.
The new benchmark, 3DMark WildLife, employs Apple's Metal API for rendering and Vulkan on Android devices. In testing at HotHardware, for basic single-run tests, again iPhones trounce anything Android, including flagship devices like the Samsung Galaxy S20 Ultra, ASUS ROG Phone 3 and OnePlus 8. However, in the extended duration WildLife Stress Test, which loops the single test over and over for 20 minutes, the current flagship iPhone 11 Pro and A13 Bionic's performance craters essentially to Snapdragon 865/865+ performance levels, while Android phones like the OnePlus 8 maintain 99+% of their performance. Though this is just one gaming benchmark test that employs the latest graphics technologies and APIs, it's interesting to see that perhaps Apple's focus on tuning for quick bursty workloads (and maybe benchmark optimization too?) falls flat if the current class of top-end iPhone is pushed continuously.
What edge? (Score:5, Funny)
iPhones are so restrictive, that the only thing they do efficiently is bilk you for money through microtransactions, and keep you inside a walled garden economy.
If you want to do basically anything "Smart" with the smart phone, you have to stick with programs designed with the above model in mind, and pay 5$+ for every bit of arbitrary software. You can completely forget about doing anything actually cool or interesting with the technology, because the hood is both figuratively and literally bolted down to prevent your access.
Android on the other hand, is so open that the notion of security is a farce.
Re: (Score:2, Interesting)
Re:What edge? (Score:5, Informative)
i have to agree, i could not even upload my own ringtone to my iphone, i even re-encoded a 2 second mp3 file to the file format that iphones use for ringtons and it still would not allow me to use it despite emailing the file to my phone because iphones wont allow me to use a thumbdrive or bluetooth for file transfer, iphones are fairly decent hardware but the software is so restrictive that i wont buy another iphone, when the iphone i have now becomes unusable i will toss it and stick to android,
You certainly can make and install your own ringtones on iOS - you just have to (gasp!) follow the directions.
I can't speak to Catalina, but on all previous versions of OS X / macOS you uploaded ringtones via iTunes. I've kept some old ringtones I like, from a very old Motorola phone, which I converted and uploaded to my iPhone.
Re: (Score:1)
"directory", "pc" - the answers to your question are in the question.
Most iPhone users (and probably a chunk of Android users) do not want to know what a directory is, or to navigate a tree of them, or even know that you organise directories in trees.
This is not a criticism of them - why should phone users need to know about these sort of structures?
Re: (Score:1)
Re: What edge? (Score:1)
Re: What edge? (Score:2)
Did you try using the middle mouse button? (Score:2)
Your rant sounds just like the absurd ones people kept repeating about macs in Ye Olde Days.
Re: Did you try using the middle mouse button? (Score:1)
All the proposed workarounds, including running iTunes in a VM, are crazy by comparison.
Do yourself a favour - buy an IPad - you know you want one.
Re: Did you try using the middle mouse button? (Score:5, Interesting)
I don't know about the guy who can't easily get ringtones on--
I do not own apple products. (I openly admit my prejudice. If other people want that crap, that is on them, but I do not like, love, nor really want anything to do with apple products. I do not approve of the direction they are trying to push computing-- which is away from the end user's control, and more toward a "Mother knows best" approach, complete with duct tape and mittens. Their products practically come with a complimentary plastic bubble.)
That said, I have had to fix/do tech stuff/assist co-workers that own them.
Here's a fun story:
A co-worker needed to get a series of text messages (that were harassing/had criminal intent behind them) out of their iphone, in order to provide them to their attorney.
I *WAS* ultimately successful. After 5 hours or work, including dredging up countless forum posts, and even trying to search fucking Reddit, looking for free tools.
Why? Because the **ONLY** way to get the text messages out of the damn phone, is to do an iTunes whole phone backup-- THEN-- Use a BUYWARE APP, DESIGNED FOR MAC, to strip out the text messages from the whole phone backup image iTunes creates-- and I'll be mother fucking god-dammed, if I am going to set up a mac virtual machine, *AND SPEND MONEY*, on something that is doable with fucking ADB on android.
(Seriously, I can do an ADB Pull on the text message database file, and then do whatever the fuck I want with it, with free tools, because it is just a SQL database file. Doesn't cost a dime.)
What I wound up doing, was use the trial version of the buyware app, after finding a "We dont really recommend it, but we also have a crippled windows version" on their website, then skating around their "We only let you extract a tiny number of messages, because this is a trial!!" bullshit, and then assembling the conversation she needed for her attorney by hand.
Every time I have needed to do anything even remotely technical to an iPhone, I have run into endless barriers of "SPEND MONEY! DO IT NAOW!", which is just plain bullshit.
Never been happier I do not own an iPhone.
Re: Did you try using the middle mouse button? (Score:2)
Or you could have done it the easy way - take screenshots and email them. >p> Even Judge Judy knows that one.
Or just go into the Ismail app , click on the arrow at the bottom, and select "Forward."
Never forwarded email before?
Re: (Score:2)
The conversation was months long.
It was over 1000 messages.
Do you really want to take that many individual screen captures, just to sound cute Barbara?
Re: Did you try using the middle mouse button? (Score:2)
Isn't that what you would do on any other device?
Re: Did you try using the middle mouse button? (Score:1)
Re: (Score:2)
Re: Did you try using the middle mouse button? (Score:1)
Re: (Score:2)
No.
No Macs. I used my windows based laptop.
What part of "I dont use apple products" did not sink in?
Re: What edge? (Score:2)
i think think teh problem is you are not gud at computers, but maybe when u are growned up u will be gud
Re: What edge? (Score:2)
Re: (Score:1)
Except you can totally use your own ringtones on an iPhone, it's just that involves owning a Mac and using special Apple software to create the file (namely GarageBand, which comes free with Macs but still, gotta get that Mac). They're specially encoded Apple Audio Codec files or something with special metadata, and without that metadata, the iPhone will just ignore the file.
So you can, but it requires owning a Mac, and it's definitely not encouraged. Instead you're supposed to buy ringtones off the iTunes
Re: (Score:2)
Of course. Because uploading a MP3 is for peasants.
Re: What edge? (Score:5, Informative)
Not true. An iPhone ringtone is just an .m4a file renamed to .m4r and it can be uploaded from an computer running a compatible version of iTunes. I encode the audio on Linux and upload to phone using iTunes running on a Windows VM on the same Linux host.
Re: (Score:1)
You have to install iTunes in a VM just to upload a bloody ringtone? Can't you just email it to yourself or something?
Re: What edge? (Score:2)
gee, if only there was an app in the Apple App Store that you could install on your iPhone that would let you make your own ringtones.
Something like, I don't know, GarageBand â¦. Oh wait , there is.
And a bunch of other ringtone apps ⦠but you wouldn't do the obvious because that deprives you of the ability to gripe about how HARD it is to do anything on an iPhone.
Re: (Score:2, Interesting)
Why can't I just email the file though? Why do I need an app?
Re: What edge? (Score:2)
Re: (Score:2)
I kind of feel like you're simultaneously very informative, and making the OP's point.
Re: What edge? (Score:1)
Re: (Score:2)
involves owning a Mac and using special Apple software to create the file (namely GarageBand, which comes free with Macs but still, gotta get that Mac).
GP is clearly talking about GarageBand on iOS ("long-tap" gives it away). No Mac involved.
Re: (Score:1)
Restricting ringtones to proprietary file format is a perfect example of why anyone who spends a cent on an Apple product has brain damage.
Re: What edge? (Score:2)
it's just that involves owning a Mac and using special Apple software to create the file (namely GarageBand, which comes free with Macs but still, gotta get that Mac).
Bzzt! Wrong! Thanks for playing...
GarageBand for iOS is part of the standard build for every iPhone and iPad (or is a free App Store Install). And you can easily make Ringtones from there.
https://osxdaily.com/2020/09/1... [osxdaily.com]
It's a shame Android was chosen by Google (Score:2, Interesting)
Apple shenanigans aside, it's a shame Android was chosen by Google back in the day. No matter how fast a JVM is, it is still a JVM. The best smartphone I've ever used was the Maemo/Meego Nokia N9: it had a single core 1GHz Cortex A8, and yet it was as smooth as any iPhone, only it could also do much more because the OS (a full linux distro) was not limiting. Due to it being abandoned (long story, but a MS shill got in as Nokia's CEO and buried Maemo to switch to Windows Mobile), I had to switch to Android t
Re:It's a shame Android was chosen by Google (Score:4, Informative)
No matter how fast a JVM is, it is still a JVM.
Isn’t Android using ART mainly?
Re: (Score:1)
ART is faster than Dalvik at startup as it has no JIT, but it still that same bytecode which runs much slower compiled than using a "native" language. For example take a look at this paper [diva-portal.org] where researchers tried to get FFT performance out of ART. They had to go through some hoops and optimisations to get performance similar (for some cases, not all) to what the native C code can give you without spending any extra effort.
The same developer spending the same effort will get an app running faster on iOS than
Re: It's a shame Android was chosen by Google (Score:2)
Re: (Score:3)
JVM was ditched years ago. These days they have ART and it's basically compiling the app at install time, optimizing for your CPU.
Re: It's a shame Android was chosen by Google (Score:2)
Quick Bursty (Score:1)
who cares (Score:1)
seriously
Re: who cares (Score:2)
Seriously? Nobody. People aren't going to make a buying decision based on this article.
It's just an excuse to Apple bash. Look at the people saying how hard it is to install a ringtone on an iPhone, because they have to set up a vim running iTunes, because they never thought to look in the App Store for an iOS app to do that right on the phone.
Or saying it took them hours to extract some emails from someone else's iPhone when they could have either just forwarded the email or a screenshot of the email.
"Craters" (Score:1)
Re: (Score:2)
Yeah the rent cratering story yesterday was a good one.
Your comment looks too much like ascii art? (Score:2)
I occastionally post anonymouly for varioius reasons. Apparently all anonymous comments now "look too much like ascii art" and are impossible to post.
Re: (Score:2)
I'm getting this too, on comments that have nothing to do with ASCII art. This must be the latest Filter Of The Week.
Power (Score:2, Insightful)
99.9+% of users care more about battery than 3D gaming. This leads to inherent thermal and power design trade-offs.
iPhone people tend to do very basic tasks on their devices so one can see why an engineer would trade off 3D performance for other design goals.
The OnePlus must have excellent cooling and is probably thicker (fact check). It's good to know that whatever sub-sub population cares about 3D gaming complexity on mobile has a go-to device and the OnePlus engineering achievement should be undersold
Re: (Score:2)
OnePlus most have excellent cooling. They can charge at 65W.
Meanwhile... all over the world (Score:2)
Lawyers are ordering legal pads and extra supplies of pencils for the deluge of class actions that will be filed in the next few weeks.
If you can do. If you can't, you become a lawyer and sue Apple.
Maybe it's just power management? (Score:3)
Re: (Score:2)
This is also one specific benchmark. On many Android phones using the Snapdragon 865+ handing beat the iPhone even when not throttling.
Re: Maybe it's just power management? (Score:2)
You run a benchmark shit gets hot chips start to throttle its not that fucking hard
Re: (Score:2)
Re: (Score:2)
yea shithead cause the android phones are not sold and marketed as 125% full load
news flash ... apple lies, constantly their hyper tech nonsense cant run 20 min under full load before scaling back to 75% so its only "faster" in short bursts
and that's been the situation ever since the PPC G5 and their ultra thin laptop's and all in ones ... its faster ... for a grand total of 5 min of full load, which is fine if you only use a "computer device" static media consumption machine
Re: (Score:2)
Eh (Score:2)
My iPhones have been solid over the years. Iâ(TM)m happy.
Scandalous (Score:2)
So, it's just a matter of software. The iPhone's speed could change with a software change,
but Snapdragon has nowhere else to go. The benchmarkers must be so proud.
Besides, wouldn't a more real-world test for Android be how fast it uploads all of your
personal data to Google?
Re: (Score:2)
Thermal management is all done in hardware?
iphone throttled still faster than android (Score:4, Informative)
The article only showed one test results from a oneplus8, and it showed that the slowest iphone loop was still faster than the fastest loop from the oneplus8.
According to the article, The edges goes to iphone since even at its slowest, its faster than android, and at its fastest, its way faster than android. Would like to see more comprehensive tests though.
iPhone SE (Score:3)
For a while the cheapest ($350) iPhone SE was faster than the $1000 Android flagship phones
Re: iPhone SE (Score:1)
arm mac pro cube will do the same? at high cost! (Score:2)
arm mac pro cube will do the same? at high cost!
With NO PCI-E slots
With apple over priced storage
TB with no Video cards supported
TB with no boot hdd / ssd
Build in GPU can't drive more then 2 screens at 8K
Apple only ram slots
All at an price that you can build an system with better storage, Good video cards , and not locked to mac os only
Re: arm mac pro cube will do the same? at high cos (Score:2)
Maybe they want to run programs without futzing with Wine?
Maybe they don't like all the shitty programs that come with Linux distros and are willing to pay good money for good software that does what they want?
Maybe they need easy to use text to speech that just works and doesn't sound like Stephen Hawking if it works at all instead of the shit that doesn't even work any more with Linux?
For pe
Re: (Score:2)
Mac os for any hardware will be good.
Apple goes way to far with hard to repair systems, over priced ram / cpu / storage upgrades. And systems to thin for work use.
Why not have an fat imac?
Why not have have an mid mac (does need to have slots but just an mini but bigger with an fan)
Why not have an 15' laptop that is not so thin?
Uhm. Ok (Score:1)
So it took somebody to make a synergetic benchmark to point out that iPhones suck.
Good thing I use my devices in the real world, and it tromps on Android day in and day out.
The real aspect of the misleading title. (Score:2)
Call me surprised that the conversation hasn't focused on the primary aspect of the misleading nature of the title (even if someone else has already used the phrase "misleading title"):
"So Hard They Lose Their Edge Over Android"
I think the data shows that the tested iPhones show diminished performance, considerable, but even with that performance loss, all three models bested every other phone, even those with the newest Snap 865+, in spite of being 1-2 generations older (or, put simpler, being 6-18 months
Apple does it right most of the time (Score:2)
As much as it pains me to say, I think this shows Apple made the right decision in the trade off of performance vs battery life. Much of the phone's UI experience is benefited from very short burst of rapid GPU performance. This gives silky smooth performance and excellent battery life. Android doesn't really come close in either smoothness or battery life. Of course Apple does pay a penalty for their obsession with being thin. When you do need performance over longer periods of time, it clearly falls d
apple is doing the right thing here (Score:2)
the common use for a phone, is to use it in short bursts, requesting info, and then taking some time to think it over. there is no reason to have an app do calculations for 20mn straight on a phone.
I suspect the thermal limits will be much different on the apple silicon macs, due to larger thermal mass and surface.
Re:Misleading title (Score:5, Insightful)
they are talking about gaming performance, and its a 20 minute test.
if you think people dont play games for 20mins at a time you're wrong.
Re: (Score:1, Informative)
It's a 3D stress test, not just a game.
What percentage of people play games with the performance equivalence of stress tests on their iPhones? None in my family.
Re:Misleading title (Score:5, Insightful)
What percentage of people play games with the performance equivalence of stress tests on their iPhones? None in my family.
If it's a CPU/GPU-bound game and not IO-bound then it's the performance equivalent of a stress test already. Lots of people play complex 3d games on phones. Remember all those stories about Apple and Epic? Those weren't about a sliding tile game, or sudoku.
Re: Misleading title (Score:2)
Why has nobody thought that the throttling might be to prevent battery overheating from excessive current draw over extended periods.
Lowering your screen brightness (half way is readable indoors) will reduce power drain - it's where I keep mine because it's easily readable indoors.
So after a few hours of constant use, it's still cool to the touch and plenty of battery left.
Unless they did this comparison, you cannot say for certain if it was the you or the battery heat dissipation that was the reason
Re: (Score:2)
"If games are "CPU/GPU bound and not IO-bound", that would imply that they have to be tailored to every different Apple processor, or provide a different user experience on each, including unpredictable behavior on faster processors."
There are lots of ways to tune automatically by targeting frame rates. This is already a thing for games on other platforms.
Re: Misleading title (Score:2)
Itâ(TM)s a synthetic test looping. The system probably saw that it was taking a lot of resources while doing nothing useful.
Re: (Score:2)
Itâ(TM)s a synthetic test looping. The system probably saw that it was taking a lot of resources while doing nothing useful.
Like a game?
Re: (Score:2)
While I don't doubt they would love this kind of control, I don't see it as being practical to implement.
'We are sorry, but to run this app at full speed that will be $1 per 10 minutes of use rounded up'. On second thought maybe it would be worth their time to implement this, and yo
Re: (Score:2)
The number of iPhones subject to stress tests regularly, or stress-test-like conditions, for long periods of time, is almost certainly miniscule.
Yeah, 3D games on iOS are practically unheard-of.
Re: (Score:1)
Yeah, 3D games on iOS are practically unheard-of.
Yeah, all 3D games on iOS are equivalent to a stress-test.
Re: (Score:1)
This article is literally about how performance takes a massive nosedive, by 50%, after 20 minutes of stress test.
If any of the idiocy mob's claims about "GAME = STRESS TEST" were true, that means that game performance in ordinary games would take a massive nosedive, by 50%, after 20 minutes of gaming. Or there abouts, plus minus. Either the framerate should take a massive nosedive, or the graphics quality, or some other major dimension.
Which games act like this? Not a single one that I have ever seen.
This
Re: (Score:2)
Just like gaming on the Mac. You have, like, Myst, and ... that puzzle game with the Apple logo! /s
Re: (Score:2)
That was long ago.
Re: (Score:2)
Re: (Score:1)
miniscule? So you are saying almost no one uses their iphone for mobile gaming? wow that will be news for millions.
You are conflating "mobile gaming" with "3D gaming with the performance equivalent of a stress test". I don't know if that was deliberate or accidental.
Re: (Score:2)
You are conflating "mobile gaming" with "3D gaming with the performance equivalent of a stress test". I don't know if that was deliberate or accidental.
No, you are trying to draw a distinction that doesn't exist. Any heavy duty game is effectively a stress test for your phone. And there are suprisingly many of those, and they have literally millions of players. I've played a number of high-complexity games on phones.
Re: (Score:1)
Ah, Drinkypoo the serially scatological troll.
1) No, a "heavy-duty" game is not the same as a "stress test" or "effectively a stress test". This is elementary. It's as absurd as claiming that any high usage is effectively equivalent to deliberate and continuous maximum usage.
2) It is absolutely possible to distinguish between highly demanding games, and games in general. This is also elementary.
3) You pretty much contradict yourself in your post here, where you declare that games that are "CPU/GPU-bound ... [slashdot.org]
Re: Misleading title (Score:3)
I am guessing that you do not game on iOS and dont really know what you are talking about from experience. Itâ(TM)s pretty rare that gamesâ"even big 3D gamesâ" really run an iOS device flat-out. Unless someone shows some evidence that this stress test mimicks any particular game, I am skeptical. And it seems that the worst-case scenario is that the device gets as slow as a flagship Android device...
Re: (Score:1)
Not to mention the rather obvious problem that, if games acted like stress tests, according to the very article people are quoting, the games would see slowdowns of 50%, or a reduction in graphics quality of 50%, or some other massive slowdown. This does not happen.
iPhones slow down after a short period of stress tests -- this is very relevant for games, because, as we know, many games act just like stress tests -- this means that games will take a huge performance hit after a short period of gaming -- exam
Re: (Score:2)
Running a game for 20 minutes, is not an unusual situation. The tests claiming higher scores with shorter runs are the ones cheating.
Re: (Score:2)
So what you're saying is that some people doing speed runs are cheating to get higher scores?
Re:Misleading title (Score:5, Interesting)
It's misleading, but not for the reasons you think it is. It applies specifically to GPU performance. Apparently the iPhone will start to throttle the GPU very quickly after it reaches max performance. (It's unclear how long until it throttles by their chart because it measures in "loops" and it basically halves after the first loop.)
Which is weird, because it's not like a GPU will typically need "burst performance" for any reason. Either you're doing a task that requires a lot of GPU and will continue to (like a game) or you're not. (Remember that the iPhone has special ML and image processing co-processors, the GPU in the iPhone is pretty much just for flinging pixels at the screen.)
It is worth mentioning that this might not practically matter for gaming, as the test involved essentially uncapped the framerate to run as fast as it could. Normal games probably won't run into that limit.
But it does really sound like Apple is basically "cooking" their benchmarks by allowing their GPUs to run really fast for the duration of a benchmark and then heavily throttling them after that.
Maybe there's a good reason for this??? (Score:2)
Perhaps the iphones are trying to not overheat, abuse the battery lifetime, not fry your hands, toast seals.... It's plausible that other things degrade as the phone heats up.
I'm not privy to know but I would imagine the SNR of the telephone signal degrades in a hot receiver. Perhaps their choices on things like Ram or Processor lintogrpahy trace width favor fast or low power or multi-core or optimized parallel gpu's over sustained temperature rises.
It's plausible that iphones trade using overclock
Re: (Score:2)
None of that alters the fact that if you run for a significant time period, the performance falls off until it is no longer faster than an Android phone in the same class.
It reminds me of the rednecks in the '80s that would performance tune a Chevy Nova but inevitably forget the cooling system. They loved to challenge people to race. You could beat them every time in a beat up old Toyota as long as you made the race longer than 2 miles. Just drive normal legal speed and after about a mile you'd find them at
Re: (Score:1)
This is obviously not just armchar theoretical speculation and mud-flinging at Apple -- because games are just like stress tests -- and we can list out all the games that take a 50% drop in framerate or graphics quality after 10-30 minutes of gaming. Here, I'll start:
You did not read the article I think (Score:2)
Yeah, but look at the actual article. It says that when the iphone thorttles it drops to a lower speed. indeed. But what speed does it drop to? Well it's slowest speed is faster than ALL of the androids. And notably this is comparing last years iphones to this years Snapdragons.
So even on it's worst day the apple is faster than the android on it's best day.
Citation please (Score:2)
I hate it when people say citation please but your wild ass accusation is so unsupported by reality you need to have something
Re: (Score:2)
What’s also misleading is that they’re comparing last year”s iPhone against this year’s Android phones. The new SoC in this year’s model has 40% more transistors thanks to a shift to a smaller process size. We have every reason to expect it to run cooler for the same performance, or perform better at the same temp.
Benchmarks like these posted on the eve of an announced product launch are rather disingenuous.
Re: Misleading title (Score:1)
Re: Misleading title (Score:2)
I don't see why people go haha over new cameras - I don't need 100 megapixels to show people what my dogs look like.
It's like going from HD to 4K to 8k to 16k tv - after a certain point, your eyes won't see the difference anyway.
Re: (Score:3)
Most of the new camera features are in software, not hardware, and they’re making a huge visual difference. Low light photography has gotten a LOT better in the last few years across the industry. Likewise, taking HDR photos (which isn’t to be confused with the HDR feature present in most 4K TVs) is more or less a standard feature and is making a huge difference. Better image stabilization, denoising, etc. have all come a long way as well.
I very much agree that our eyes can’t appreciate mo
Re: Misleading title (Score:2)
Re: (Score:2)
Agreed, MP for their own sake aren’t particularly useful. At a certain point, you start butting up against the physics of light and those individual pixels become little more than noise because each one is getting so little light. But at the same time, cameras are more than their sensors, and that was the point I was trying to make, hence the emphasis on software. The image signal processors in these new phones are way beyond what they used to be, allowing for photography that wasn’t possible be
Re: So? (Score:2)
At least with IOS you can uninstall gmail, chrome, and the rest of the Google spyware ecosystem.
You might want to try it some time. It's liberating.