The likes of Microsoft, IBM, Google, and more publish amazing research every week. Apple rolls in with one fucking paper and the tech blogs are all covering it like breathless teenagers. Sometimes I really hate my industry.
True that. On IBM Bluemix you can get face detection and train Watson for image classification, and it's free for up to 250 images per day (after that it's about $0.01 per 5 images).
IBM is not only years ahead of Apple on this, they already commoditized visual recognition and they're making money with it.
And if Apple did the same thing, you'd be the first to call them Greedy.
If I get the new MBP, then I would be running the new version of Safari, naturally, so my MBP would not be providing the advertised battery life If I have this issue!!
It doesn't really matter if the underlying reason is Hardware or Software problem, As long as there exists a problem, then the Unit does not provide the battery life promised, and CR's test results would still be valid, if they show the issue.
Yes, it matters. A LOT. You're an idiot if you don't understand why; but here goes...
If it is a software issue, then that should be fairly straightforward to diagnose and fix, and the fix can be instantly rolled-out to all existing users, and included in the software build of all machines being built from this time forward. Problem solved! And meanwhile, the "workaround" could be as simple as using Chrome instead of Safari until the fix is available. Most reasonable people would understand. But I'm sure you would act like using another Browser for a couple of weeks would be a crime against humanity...
Conversely, if it is a Hardware issue, the fix may be difficult, could temporarily halt production while updated hardware is designed, qualified, sent though agency approvals, etc, and would most likely result in a recall of all units already sold.
No, but what it does do is give an absolute worst case scenario that intelligent, knowledgeable people (the ones buying 'pro' models of products) can use to intuit what they might expect for their own usage. It's just as valid a metric as 'typical LCD use case.' They measure different things for different people.
But show me a single laptop OEM that tests that way....
My 2012 Macbook Air has a non-user replaceable battery. When the battery was no longer holding adequate charge, I took it to an Apple store and they replaced it for me. I forget what the cost was, a little over a hundred I believe. Expensive, but reasonable if I can get a few years out of it.
Bottom line: The battery IS replaceable. You just need to take it to Apple to do it.
Or simply get out your $30 Harbor Freight heat-gun and DIY...
But honestly, by the time you make sure you aren't getting some bootleg battery offa Amazon, go to Harbor Freight for the Heat Gun, get any other tools you might not have, then wrestle your way to the battery and get it out, that $100 or so to an Apple Tech seems pretty reasonable, indeed...
Battery replacement is. After a few years, battery life will be half of whatever it started with. At that point, the MBP and its irreplaceable batteries can never stray very far from the charger. Users might accept that, as many people don't depend on the battery all that much. The ultimate deal breaker is soldered SSD. When that fails (and it will), the computer is junk.
If Apple offered a MacBook Pro with HALF of the current battery life, HALF of the memory, and HALF of the storage capacity, but made the components replaceable, they would sell a lot more of them, even if they were TWICE as thick.
Sorry to burst your fantasy-bubble; but my 2012 MacBook Pro gets indistiguishably-differernt battery life from when it was new. If it is "down" at all, it is only by a few minutes, and nothing I could ever put a finger-on, given varying workload.
As far as soldered SSD, I'm more with you on that one. But I will bet you will find that less than 1% of MBP owners suffer a catastrophic SSD-failure before they are ready to replace their machines anyway. I believe that this is essentially what Apple is using.
Also, the iFixit teardown uncovered a "connector to nowhere", that appears to be a way for at least Apple techs to access the SSD of an otherwise dead MBP directly, for data-recovery purposes.
As far as the batteries go, they are glued-down as usual; but are reasonably replaceable, considering you might have to do it once before the laptop is due for replacement.
Does this mean bribing them to write more favourable reviews, or threatening them with lawsuits if they don't?
It means Apple is trying to determine the best way to spin this into a story about how CR's test doesn't represent reality. They've got their top spindoctors working around the clock to construct a narrative that'll appeal to Apple fans.
Or maybe, just maybe, they have their top Engineers working on it around-the-clock to determine the cause of these clearly anomalous test results.
Apple says the devices should operate for up to 10 hours between charges
up to 10 hours includes lasting only 10 minutes.
That's true but historically, Apple has been quite honest about how long the battery will last. Up to the point where you'd have a decent chance of doing better.
However with the current "Late 2016" generation, there's people with 15" MacBook Pros who report not getting even half of that. So it's really a break from the past.
And in every single anecdote I have read online regarding this, the poster proclaims "I was only using Safari..."
Problem's with Safari, or an underlying Framework, folks. The problem is not in the MBP itself.
Safari hasn't remained constant over the years. In fact, it sees a lot of change. If it is a software problem affecting Safari, for example, then it could have been introduced with recent versions of Safari.
You mean like the Custom, Touch-Bar-Aware version of Safari released specifically for this product?
CR is far more trustworthy than Apple. There is little reason to trust Apple on this. CR has been doing this shit for decades on myriad models. To suggest Apple knows better is to suggest Apple game the system.
Let's leave "trustworthiness" out of this speculation, shall we?
CR has posted some numbers which vary wildly in both directions from Apple's published figures.
Apple is asking politely if they can see if they can understand how CR got these results.
Nothing to see here. Sounds like it is an issue with the newly-released Touch Bar version of Safari. Which is actually quite good news.
This is not new news. People have been complaining of bad battery life well before Consumer Reports: http://www.macrumors.com/2016/....
The bottom line is it shouldn't have such drastically different test results with the same repeated tests. Whether it's the batteriesSafari or otherwise, it should be at least consistant.
http://www.macrumors.com/2016/...
But one VERY possible reason it wouldn't be was if Safari exhibited an intermittent bug, where SOMETIMES, one or more Runaway Processes run at 100% CPU and/or GPU for as long as it is Launched (and maybe even after it is Quit).
Why do people keep ignoring the 800 lb Gorilla; that being, that the problem seemed to "Follow" Safari. Safari = Wildly Differing Battery-Life; Chrome = Battery Life per Apple's Published spec.
If "extra processing" causes a decrease in battery life (it does!) that's not a software glitch, it's exposing the bogus inflated battery life figures. Intel power management doesn't give battery manufacturers a free ride to more grandiose claims, but that is what they have been doing.
Ya know, there's a reason why everyone's battery tests are conducted in "real-world" simulations, rather than by simply staring-up some sort of "battery-stress-test" software that would peg the CPU/GPU/Hard Drive/Display Brightness until the machine stopped running: And that reason is, it doesn't inform the potential buyer about what they would expect to achieve in the REAL WORLD.
So, if (as I suspect) this is a fault of the often-launched-for-hours-at-a-time Safari Browser (or some Framework thereunder), not just doing "some extra processing", but rather leaving one or more Processes running full-tilt on one or more CPU Cores and even possibly the dGPU for that same hours-at-a-time, it is quite natural to expect that the battery life would be quite affected in any laptop.
I was also reading a poster on MacRumors that owns both a 2015 and 2016 MBP, and he claims he is getting virtually identical run-times on the two systems, doing the same things. And that the results are in-line with Apple's published figures. And I am fairly certain that that is as least as scientific a test as what CR conducted.
How can you be so sure that there isn't some sort of yet-undiscovered flaw with the testing methodology that was being used? How can you be so sure that it is a problem with the batteries?
It could very well be a problem with the testing methodology, and not the batteries. In such a case, there isn't even anything Apple could do to "fix the battery issues", as these issues to "fix" wouldn't even exist to begin with!
The article says that the testing involved Safari repeatedly loading certain web pages. How are you so sure that this type of testing doesn't, for example, sporadically cause Safari to do extra processing, perhaps due to a bug in Safari? If something like that is happening, and it isn't consistent, then it could very well be a purely software problem being mistaken for a hardware problem. Anyone who has done any real world software development will know very well how unpredictable and random some bugs can be.
The most sensible thing to do in a situation like this is look at the testing methodology first. Apple is doing the right thing, by trying to get to the bottom of this in a systematic, controlled, intelligent manner. What you're proposing is the opposite of that: it's all about random, uncontrolled, amateurish guessing.
The above post is spot-on.
And the fact that CR found that the problem seemed to go away by using Chrome instead of Safari, points rather directly to an issue with the Touch Bar version of Safari, or some Framework used thereunder, that is leaving one or more Processes spinning at 100% CPU for hours-on-end, rather than some weakness in the MBP's batteries.
They are trying to prove them wrong, they are not working on fixing the battery issues. Within a few days we will all learn on Slashdot that Consumer Reports was doing it wrong.
And, even to me, it's obvious if Phil Shiller is the point person on this, Apple is looking for a PR "solution" to this battery life issue.
This is antenna-gate all over again.
I think that is jumping to conclusions.
Phil is definitely "on" this, even reaching-out to individual Users who have Responded to his Tweet on CR, asking them to send him certain specific Logs to help Apple diagnose this problem.
The difference is, Uncle Phil is doing this on a one-on-one, direct basis, rather than issuing some "Press Release" explaining why MBP Users are just "charging it wrong", or something.
I honestly think that, while it is true that Schiller wears a Marketing hat, he is high-enough up in the Organization that if he wants to take this on as a "Personal Project", nobody inside Apple is going to question his sincerity; and if you look at his bio, he has enough "tech" chops to understand and appreciate the possible software (and likely some of the hardware) causes for all this.
Behind the scenes, I am also sure he has someone(s) on the Mac Dev. Team working over Christmas on this, even if he is engaging in a little direct communication with individuals, CR, and possibly the media in general.
ORLY?
If Apple released an iPhone version with this many defects, you would literally never hear the end of bitching. And yet, you dismiss this with a hand-wave.
Smartphones are complex. A few examples of any given model will have problems. iPhone, Galaxy, Pixel, all of them. Does anyone know what typical problem rates are? Are the problem rates for Pixel phones close to typical? If so, there's nothing notable to see here.
Not this many problems in one unit, sorry.
Google rushed this out for the holiday season, and witness the result.
Plus, they essentially are pulling a slow-motion "Plays for Sure" trick with their fork of Android and their OEM Android Licensees.
And now CyanogenMod is dead.
Stick a fork in Android. It was horrible while it lasted...
They're literally the same thing. Apple's "secure enclave" uses the ARM A7's TrustZone/SecurCore tech, which is basically an independent CPU in the same die. There's nothing magical about it - Android's been supporting it for eons now: https://source.android.com/sec...
Sorry, still incorrect. Although Android is (gradually) getting better, slowly, it still is hit and miss, depending on which SoC your device OEM decides to use.
The only thing I was incorrect about is that I was under the impression that the SE was a separate component, rather than being implemented on-die in the Axx SoC. Other than that, my assertion that iOS' Security is more robust than Android's, stands.
Tomorrow may be different; but we're living in today...
It sounds like that is half of the system that Apple has implemented. Apple's Axx SoC's have a TEE, but they also have the "Secure Enclave" (SE) CHIP. What is still missing in Android's implementation is the SEPERATE Secure Enclave chip.
I'm not a security expert, but from what I understand, the SE holds the key (literally) to decryption, and there is no direct or even indirect path to that key. That provides an important additional level of abstraction and security that no "on chip" solution can provide. And, according to those who ARE Security Experts, that is a significant advantage.
While I agree it would be significantly stronger than not having encryption enabled, how much stronger it is entirely dependent on the particular device's hardware.
Wrong. Android uses AES via dm-crypt, which is the same cipher used by iOS - the difference is that not all devices have hardware crypto support. IIRC this is one of the reasons Android doesn't enable it by default; it comes with a performance hit.
I was thinking about the Secure Enclave chip that Apple uses. How many Android devices have that sort of thing?
Ever notice how the FBI never goes after Google/Android/Samsung/etc. for access to their phones? Apple has a much smaller market share and the feds are all over them?
Because Android is like a screen door. The feds don't need to sue, they just walk in.
No, it is because Android doesn't do encryption by default - the "why" beats me. Enable it and it is as strong as any Apple offering.
While I agree it would be significantly stronger than not having encryption enabled, how much stronger it is entirely dependent on the particular device's hardware.
And in any event, it is not likely to be "as strong as any Apple offering", sorry.
Slashdot Mirror
The likes of Microsoft, IBM, Google, and more publish amazing research every week. Apple rolls in with one fucking paper and the tech blogs are all covering it like breathless teenagers. Sometimes I really hate my industry.
True that. On IBM Bluemix you can get face detection and train Watson for image classification, and it's free for up to 250 images per day (after that it's about $0.01 per 5 images).
IBM is not only years ahead of Apple on this, they already commoditized visual recognition and they're making money with it.
And if Apple did the same thing, you'd be the first to call them Greedy.
If I get the new MBP, then I would be running the new version of Safari, naturally, so my MBP would not be providing the advertised battery life If I have this issue!!
It doesn't really matter if the underlying reason is Hardware or Software problem, As long as there exists a problem, then the Unit does not provide the battery life promised, and CR's test results would still be valid, if they show the issue.
Yes, it matters. A LOT. You're an idiot if you don't understand why; but here goes...
If it is a software issue, then that should be fairly straightforward to diagnose and fix, and the fix can be instantly rolled-out to all existing users, and included in the software build of all machines being built from this time forward. Problem solved! And meanwhile, the "workaround" could be as simple as using Chrome instead of Safari until the fix is available. Most reasonable people would understand. But I'm sure you would act like using another Browser for a couple of weeks would be a crime against humanity...
Conversely, if it is a Hardware issue, the fix may be difficult, could temporarily halt production while updated hardware is designed, qualified, sent though agency approvals, etc, and would most likely result in a recall of all units already sold.
BIG DIFFERENCE!
No, but what it does do is give an absolute worst case scenario that intelligent, knowledgeable people (the ones buying 'pro' models of products) can use to intuit what they might expect for their own usage. It's just as valid a metric as 'typical LCD use case.' They measure different things for different people.
But show me a single laptop OEM that tests that way....
My 2012 Macbook Air has a non-user replaceable battery. When the battery was no longer holding adequate charge, I took it to an Apple store and they replaced it for me. I forget what the cost was, a little over a hundred I believe. Expensive, but reasonable if I can get a few years out of it.
Bottom line: The battery IS replaceable. You just need to take it to Apple to do it.
Or simply get out your $30 Harbor Freight heat-gun and DIY...
But honestly, by the time you make sure you aren't getting some bootleg battery offa Amazon, go to Harbor Freight for the Heat Gun, get any other tools you might not have, then wrestle your way to the battery and get it out, that $100 or so to an Apple Tech seems pretty reasonable, indeed...
Battery replacement is. After a few years, battery life will be half of whatever it started with. At that point, the MBP and its irreplaceable batteries can never stray very far from the charger. Users might accept that, as many people don't depend on the battery all that much. The ultimate deal breaker is soldered SSD. When that fails (and it will), the computer is junk.
If Apple offered a MacBook Pro with HALF of the current battery life, HALF of the memory, and HALF of the storage capacity, but made the components replaceable, they would sell a lot more of them, even if they were TWICE as thick.
Sorry to burst your fantasy-bubble; but my 2012 MacBook Pro gets indistiguishably-differernt battery life from when it was new. If it is "down" at all, it is only by a few minutes, and nothing I could ever put a finger-on, given varying workload.
As far as soldered SSD, I'm more with you on that one. But I will bet you will find that less than 1% of MBP owners suffer a catastrophic SSD-failure before they are ready to replace their machines anyway. I believe that this is essentially what Apple is using.
Also, the iFixit teardown uncovered a "connector to nowhere", that appears to be a way for at least Apple techs to access the SSD of an otherwise dead MBP directly, for data-recovery purposes.
As far as the batteries go, they are glued-down as usual; but are reasonably replaceable, considering you might have to do it once before the laptop is due for replacement.
Does this mean bribing them to write more favourable reviews, or threatening them with lawsuits if they don't? It means Apple is trying to determine the best way to spin this into a story about how CR's test doesn't represent reality. They've got their top spindoctors working around the clock to construct a narrative that'll appeal to Apple fans.
Or maybe, just maybe, they have their top Engineers working on it around-the-clock to determine the cause of these clearly anomalous test results.
Damn. People are completely over-the-top cynical.
What a vague phrase - "working with".
Does this mean bribing them to write more favourable reviews, or threatening them with lawsuits if they don't?
How about "Working With" meaning exactly that.
up to 10 hours includes lasting only 10 minutes.
That's true but historically, Apple has been quite honest about how long the battery will last. Up to the point where you'd have a decent chance of doing better.
However with the current "Late 2016" generation, there's people with 15" MacBook Pros who report not getting even half of that. So it's really a break from the past.
And in every single anecdote I have read online regarding this, the poster proclaims "I was only using Safari..."
Problem's with Safari, or an underlying Framework, folks. The problem is not in the MBP itself.
Safari hasn't remained constant over the years. In fact, it sees a lot of change. If it is a software problem affecting Safari, for example, then it could have been introduced with recent versions of Safari.
You mean like the Custom, Touch-Bar-Aware version of Safari released specifically for this product?
CR is far more trustworthy than Apple. There is little reason to trust Apple on this. CR has been doing this shit for decades on myriad models. To suggest Apple knows better is to suggest Apple game the system.
Let's leave "trustworthiness" out of this speculation, shall we?
CR has posted some numbers which vary wildly in both directions from Apple's published figures.
Apple is asking politely if they can see if they can understand how CR got these results.
Nothing to see here. Sounds like it is an issue with the newly-released Touch Bar version of Safari. Which is actually quite good news.
This is not new news. People have been complaining of bad battery life well before Consumer Reports: http://www.macrumors.com/2016/.... The bottom line is it shouldn't have such drastically different test results with the same repeated tests. Whether it's the batteriesSafari or otherwise, it should be at least consistant. http://www.macrumors.com/2016/...
But one VERY possible reason it wouldn't be was if Safari exhibited an intermittent bug, where SOMETIMES, one or more Runaway Processes run at 100% CPU and/or GPU for as long as it is Launched (and maybe even after it is Quit).
Why do people keep ignoring the 800 lb Gorilla; that being, that the problem seemed to "Follow" Safari. Safari = Wildly Differing Battery-Life; Chrome = Battery Life per Apple's Published spec.
DUH!!!
If "extra processing" causes a decrease in battery life (it does!) that's not a software glitch, it's exposing the bogus inflated battery life figures. Intel power management doesn't give battery manufacturers a free ride to more grandiose claims, but that is what they have been doing.
Ya know, there's a reason why everyone's battery tests are conducted in "real-world" simulations, rather than by simply staring-up some sort of "battery-stress-test" software that would peg the CPU/GPU/Hard Drive/Display Brightness until the machine stopped running: And that reason is, it doesn't inform the potential buyer about what they would expect to achieve in the REAL WORLD.
So, if (as I suspect) this is a fault of the often-launched-for-hours-at-a-time Safari Browser (or some Framework thereunder), not just doing "some extra processing", but rather leaving one or more Processes running full-tilt on one or more CPU Cores and even possibly the dGPU for that same hours-at-a-time, it is quite natural to expect that the battery life would be quite affected in any laptop.
I was also reading a poster on MacRumors that owns both a 2015 and 2016 MBP, and he claims he is getting virtually identical run-times on the two systems, doing the same things. And that the results are in-line with Apple's published figures. And I am fairly certain that that is as least as scientific a test as what CR conducted.
How can you be so sure that there isn't some sort of yet-undiscovered flaw with the testing methodology that was being used? How can you be so sure that it is a problem with the batteries?
It could very well be a problem with the testing methodology, and not the batteries. In such a case, there isn't even anything Apple could do to "fix the battery issues", as these issues to "fix" wouldn't even exist to begin with!
The article says that the testing involved Safari repeatedly loading certain web pages. How are you so sure that this type of testing doesn't, for example, sporadically cause Safari to do extra processing, perhaps due to a bug in Safari? If something like that is happening, and it isn't consistent, then it could very well be a purely software problem being mistaken for a hardware problem. Anyone who has done any real world software development will know very well how unpredictable and random some bugs can be.
The most sensible thing to do in a situation like this is look at the testing methodology first. Apple is doing the right thing, by trying to get to the bottom of this in a systematic, controlled, intelligent manner. What you're proposing is the opposite of that: it's all about random, uncontrolled, amateurish guessing.
The above post is spot-on.
And the fact that CR found that the problem seemed to go away by using Chrome instead of Safari, points rather directly to an issue with the Touch Bar version of Safari, or some Framework used thereunder, that is leaving one or more Processes spinning at 100% CPU for hours-on-end, rather than some weakness in the MBP's batteries.
They are trying to prove them wrong, they are not working on fixing the battery issues. Within a few days we will all learn on Slashdot that Consumer Reports was doing it wrong.
And you know this, how?
And, even to me, it's obvious if Phil Shiller is the point person on this, Apple is looking for a PR "solution" to this battery life issue.
This is antenna-gate all over again.
I think that is jumping to conclusions.
Phil is definitely "on" this, even reaching-out to individual Users who have Responded to his Tweet on CR, asking them to send him certain specific Logs to help Apple diagnose this problem.
The difference is, Uncle Phil is doing this on a one-on-one, direct basis, rather than issuing some "Press Release" explaining why MBP Users are just "charging it wrong", or something.
I honestly think that, while it is true that Schiller wears a Marketing hat, he is high-enough up in the Organization that if he wants to take this on as a "Personal Project", nobody inside Apple is going to question his sincerity; and if you look at his bio, he has enough "tech" chops to understand and appreciate the possible software (and likely some of the hardware) causes for all this.
Behind the scenes, I am also sure he has someone(s) on the Mac Dev. Team working over Christmas on this, even if he is engaging in a little direct communication with individuals, CR, and possibly the media in general.
the 1 month old phone retard
ORLY? If Apple released an iPhone version with this many defects, you would literally never hear the end of bitching. And yet, you dismiss this with a hand-wave.
Typical Fandroid Slashtard.
Smartphones are complex. A few examples of any given model will have problems. iPhone, Galaxy, Pixel, all of them. Does anyone know what typical problem rates are? Are the problem rates for Pixel phones close to typical? If so, there's nothing notable to see here.
Not this many problems in one unit, sorry.
Google rushed this out for the holiday season, and witness the result.
Plus, they essentially are pulling a slow-motion "Plays for Sure" trick with their fork of Android and their OEM Android Licensees.
And now CyanogenMod is dead.
Stick a fork in Android. It was horrible while it lasted...
early adopter tax
Bullshit.
HOW long has Google been Developing Android?!?
HOW long has Google been Designing Android Phones?!?
HOW long has HTC been building Android Phones?!?
So exactly WHAT is "EARLY"???
An Indian company seems to he done better https://en.wikipedia.org/wiki/...
And an American Company has done even better.
They're literally the same thing. Apple's "secure enclave" uses the ARM A7's TrustZone/SecurCore tech, which is basically an independent CPU in the same die. There's nothing magical about it - Android's been supporting it for eons now: https://source.android.com/sec...
Sorry, still incorrect. Although Android is (gradually) getting better, slowly, it still is hit and miss, depending on which SoC your device OEM decides to use.
The only thing I was incorrect about is that I was under the impression that the SE was a separate component, rather than being implemented on-die in the Axx SoC. Other than that, my assertion that iOS' Security is more robust than Android's, stands.
Tomorrow may be different; but we're living in today...
https://en.wikipedia.org/wiki/... . Any device using Android 6.0 onwards.
It sounds like that is half of the system that Apple has implemented. Apple's Axx SoC's have a TEE, but they also have the "Secure Enclave" (SE) CHIP. What is still missing in Android's implementation is the SEPERATE Secure Enclave chip.
I'm not a security expert, but from what I understand, the SE holds the key (literally) to decryption, and there is no direct or even indirect path to that key. That provides an important additional level of abstraction and security that no "on chip" solution can provide. And, according to those who ARE Security Experts, that is a significant advantage.
While I agree it would be significantly stronger than not having encryption enabled, how much stronger it is entirely dependent on the particular device's hardware.
Wrong. Android uses AES via dm-crypt, which is the same cipher used by iOS - the difference is that not all devices have hardware crypto support. IIRC this is one of the reasons Android doesn't enable it by default; it comes with a performance hit.
I was thinking about the Secure Enclave chip that Apple uses. How many Android devices have that sort of thing?
Well if simply uttering a three-word meme was the only proof of any wrongdoing then it deserved to be "punish-modded" then as it does now.
Why?
People get modded Insightful on herd all the time for repeating ridiculous memes about Apple.
Ever notice how the FBI never goes after Google/Android/Samsung/etc. for access to their phones? Apple has a much smaller market share and the feds are all over them?
Because Android is like a screen door. The feds don't need to sue, they just walk in.
No, it is because Android doesn't do encryption by default - the "why" beats me. Enable it and it is as strong as any Apple offering.
While I agree it would be significantly stronger than not having encryption enabled, how much stronger it is entirely dependent on the particular device's hardware.
And in any event, it is not likely to be "as strong as any Apple offering", sorry.
Now if only there was a Mobile Platform that has a longstanding history of providing Updates for reasonable periods of time...
Oh, wait...