Slashdot Mirror
Consumer Reports Now Recommends MacBook Pros (macrumors.com)
Consumer Reports has updated their report on the 2016 MacBook Pros, and is now recommending Apple's latest notebooks. MacRumors reports: In the new test, conducted running a beta version of macOS that fixes the Safari-related bug that caused erratic battery life in the original test, all three MacBook Pro models "performed well." The 13-inch model without a Touch Bar had an average battery life of 18.75 hours, the 13-inch model with a Touch Bar lasted for 15.25 hours on average, and the 15-inch MacBook Pro with Touch Bar had an average battery life of 17.25 hours. "Now that we've factored in the new battery-life measurements, the laptops' overall scores have risen, and all three machines now fall well within the recommended range in Consumer Reports ratings," reports Consumer Reports. Consumer Reports originally denied the 2016 MacBook Pro a purchase recommendation in late December due to extreme battery life variance that didn't match up with Apple's 10 hour battery life claim. Apple worked with Consumer Reports to figure out why the magazine encountered battery life issues, which led to the discovery of an obscure Safari caching bug. Consumer Reports used a developer setting to turn off Safari caching, triggering an "obscure and intermittent bug reloading icons" that drained excessive battery. The bug, fixed by Apple in macOS Sierra 10.12.3 beta 3, is not one the average user will encounter as most people don't turn off the Safari caching option, but it's something done in all Consumer Reports tests to ensure uniform testing conditions. A fix for the issue will be available to the general public when macOS Sierra 10.12.3 is released, but users can get it now by signing up for Apple's beta testing program.
Cha Ching
If they had waited for the update to come out, I'd say it's great they're willing to keep their reviews up to date (even if it was due to pressure from Apple).
But if you're going to recommend a product based on a might-be-working-when-released beta, you lose some credibility to me.
Consumer Reports now admits they were holding it wrong.
I knew it wouldn't take Apple long to fix the bug in the RDF.
defense
They did not re-enable caching. There was a bug that only occurred when caching was turned off (which is why consumer reports got wildly different results than most people). Apple fixed the bug, and then consumer reports re-ran their tests using the beta and got decent results.
In other news, reviewers found that MacBook Pros don't need battery power at all!
Instead, after rigious testing, Consumer Reports found MacBook Pros can run off Apple's cash supply indefinitely by plugging the MacBook into the wall or using wireless charging and simplying manipulating it's review report, given enough cash. Yay!
From the article, it doesn't sound like the fix was to turn on page caching. There was some bug caused by turning off caching that was fixed. What that is....idk. I can only imagine some guy coded a spinloop wrong.
I know it requires RTFA, but they did not change their tests. They applied the update (a bug fix) and re-ran their tests as-is. They didn't change their test to enable caching, it's still disabled, but with the bug fix they're getting acceptable battery life.
In fact, IT has made it clear that the new MacBook Pros are too expensive to be considered for purchase. It's leading to a rush to try and get the old 2015 MBPs while they're still available, since a whole lot of developers that use OS X because it's a Unix that IT supports are rapidly having to find new solutions. (And, no, Linux is not an answer because it involves a waiver process before you can access the network.)
If Microsoft gets their Linux layer working, Apple is toast. The new MBP is too expensive for specs that are too poor. The MacBook line isn't good enough for development work. If Microsoft can turn Windows 10 into an acceptable dev environment, OS X is toast. IT would love to find a reason to axe OS X support.
So the fact that a CONSUMER organization says they're OK for consumers is worthless. I still can't use it.
When we discussed this matter earlier, it was pointed out that this could very well be a software problem (with added emphasis):
This submission's summary now confirms that it was due to "an obscure Safari caching bug".
How was that Slashdot comment, which turned out to be right, modded? -1.
The same happened to several other comments that, it's now obvious, were correct. They're at -1, while a bunch of junk comments were modded up.
It turns out that ddtmm was wrong. reanjr was wrong. fuzzyfuzzyfungus was wrong. Shane_Optima was wrong. lucm was wrong.
All of those smug commenters turned out to be wrong.
Now that we know what happened in this case, I think it would be appropriate for the Slashdot admins to go back and fix up the atrocious moderating that happened in that other submission. Mod up the comments that were at -1 to +5, because they turned out to be right. Mod down the users who were wrong, sending their comments to -1.
Anybody who was responsible for such awful moderating should never moderate again. Strip them of their moderating privilege permanently. And since they screwed up so badly in this case, we should assume they screwed up every other moderating they ever did. Invert all of those moddings. Mod down anything they modded up, and mod up anything they modded down.
It really makes Slashdot look worse than it already looks when such awful moderating goes uncorrected.
Consumer Reports does testing for....consumers. To see what the consumer experience with a product is going to be. As the bug would never affect a normal user, it should not have been used to give the program a poor review.
Apparently, you need to shame Apple publicly in a magazine that quite a lot of people read and seriously consider before making a purchase. Apple still is very much a hardware company, and if CR says their hardware isn't good, their sales will go down.
So, if you find a bug in one of their many software products, don't report the bug as usual. Instead, make an argument that the bug causes unacceptable battery performance and report it to Consumer Reports instead. Then, maybe Apple will start fixing software bugs.
This is sketchy. Tests typically involve disabling the cache and reloading pages to measure the overall impact of visiting lots of different pages.
You should be happy to know that caching remains off. All that was identified is that caching triggered a bug, it was the bug that was fixed and the test was then repeated in the exact same conditions as they always have done.
I can no longer recommend anyone gives any mod points to ACs.
It's still very expensive due to the new touchbar that hasn't proven its worth as yet. The build quality is still there, certainly, but it still fails for not being the latest Intel hardware (yes, I know the issues with low TDP Kaby Lake packages at the time of launch) and a poor AMD graphics chipset. It's the iPad 3 of the Macbook Pro line, and hopefully will be replaced by August with a Kaby Lake version. I don't hold much hope for a Nvidia 1060 version, but those two changes would be the only reason for me to buy it (even less hope for Apple to concede and get into the eGPU game). In the meantime, I've spend $1k less on a brand new laptop that I'll be living in Linux-land for the forthcoming cycle.
This is what Apple doesn't get. The more pissed developers leave their ecosystem, the worse apps they'll get, which means the worse customer experience, the fewer devices sold, and their eventual decline. Of all the times to double down on catering to the smaller but more valuable element of their customer base, it's when you're rolling in infinite money. Instead, they're trying to make infinity*2 money.
I find the whole thing a terrible example of a once respectable online outlet jumping into the clickbait world: CR used to have near print-level standards - where at least the dollar-centric parts were constrained away and didn't govern the copy (Remember what 80% of your print PC mags used to be?)
The reality was that the testing method was *evidently* at fault: they couldn't reproduce the issue reliably, they couldn't reproduce it with Chrome (which is the dominant browser on most platforms) - but rushed to print before Christmas because of clicks. This should always have been a non-story, and it's worth nothing that the *exact* same software that has been fixed has been present on every other MacBook for years - it's not like this was a touchbar related issue. Odds are some OSX update introduced the issue that screwed with the testing outcomes, and it's affected every one of the models they'd already approved previously too. Thats the trouble with static recommendations like this.
Economically though, the clicks meant that despite breaking the rules of professionalism and common sense that they probably made enough from the click rate that jumping to the press like this with half a story has won for them - David vs Goliath always sells. That means next time they've got a dubious outcome they'll not learn anything from this.
Really, because at they state:
"With the updated software, the three MacBook Pros in our labs all performed well, with one model running 18.75 hours on a charge. We tested each model multiple times using the new software, following the same protocol we apply to hundreds of laptops every year."
"Same protocol" implies to me that they did not reenable caching, and they do not mention changing the test at all except for running the updates OS. But you know that's just the official statement from consumer reports, not the fantasy one you are talking about where they claim to have altered their testing procedure...
I guess this means I'm not getting a replacement for a battery that actually lasts 10 hours one day.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
Oh my god people on the internet were wrong! The horror...
Only the State obtains its revenue by coercion. - Murray Rothbard
I've got 20 of these things seeing daily use as testing machines within our company, with the potential to purchase anywhere between 4K and 20K machines should they be deemed acceptable.
Right now, they are not.
So far, every single user that has a testing unit has complained about dongles at least once. Nearly half of them have complained about accidentally activating the touch bar, simply because their fingers accidentally brushed across it. We've also seen three laptops outright fail- the touchbar would randomly hang up, severely reducing the usability of the machine (apparently it's controlled by an embedded computer running it's own OS, rather than being accessible as a second display of pixels through the host OS).
However, the current show stopper is the battery life.
If you use these machines lightly, they're great. We get anywhere from 10 to 15 hours of use. If you load the machine down in any significant way- be it causing the GPU to kick in, or all the CPU cores to fire up- battery life drops to a measly 4-5 hours, sometimes as low as 3.
It's pretty clear that Apple has implemented some very aggressive power saving features, but at the same time this appears to be something they had to do to get any kind of reasonable battery life out of these machines- it's not something they did to extend an already excellent battery life, because if you're actually using the hardware then the runtime isn't that stellar. This is in line with what CR originally found, regardless of any Safari bugs. The machine simply isn't capable of lasting that long if you're using the hardware, regardless of what that use is caused by (be it a stray daemon sucking up 100% of a CPU core, or a Safari bug doing something similar, or anything else).
As it stands right now, these are going to be the first Apple laptops our company won't purchase for any of our employees. The hardware just isn't capable of consistently meeting our requirements. I get the feeling that Apple would have loved to have crammed netbook quality hardware inside this thing for thinness, but they knew that was marketing suicide so they came up with this machine instead, where you've got impressive specs and it works well as,long as you don't try to use them.
It's a shame, because if they'd made the machine a bit bigger with a higher capacity battery, we wouldn't have any complaints. 7-8 hours of life under moderate to heavy use is great, but that's nowhere near what we're seeing with these units.
I'm sorry, but without a scientific study being involved, your statement is invalid.
Consumer Reports was benchmarking a debug mode. That should not be a legitimate config to represent typical usage.
Shane_Optima was wrong.
I most certainly was not wrong. I said that if it was a software bug in Safari (as alleged) that it was obviously still Apple's fault. I didn't address the possibility of CR screwing up one way or another. And guess what? According to TFS, Apple *did* screw up.
Apple is responsible for Safari bugs. That was my assertion then, and it's my assertion now.
How was that Slashdot comment, which turned out to be right, modded? -1.
Wow. So you're complaining that an Anonymous Coward (you?) speculating baselessly (yes baselessly, because no preliminary observations or experiments were mentioned) about the possible cause of the poor test result and then implying that Apple should be let off the hook if it's a Safari bug received a single -1 downmod instead of being modded up to +5, Nostradamus?
No one is going to have their mod privileges revoked. Instead, try re-working your tone to sound less like a perpetually whining fanboy.
I mean, for many years I liked Google (still do, in some ways) but I don't flip the fuck out when people criticize, for example, their decision to drop microSD card slots from their devices. That was a horrible anti-consumer decision and I made sure to mention it any time I talked to someone who was thinking about buying a Nexus device. There's a reason why Apple fanboys have the reputation that they do. No other tech company on Earth inspires this kind of rabid and unthinking loyalty.
Incidentally, if you register for an account people are around here will be less likely to assume you're a blithering fool or astroturfer.
The real moderation tragedy is that your comment here is currently modded up to +4. "Admins, go back and fix the moderation and mod everyone else down! My speculative Apple apologia turned out to be correct in fact [just not in conclusion]!", Jesus fucking Christ...
And just to quickly get this out there, in the name of everything that's holy of course I'm not saying that all Apple users everywhere do this, or that they're in any way bad or stupid for using Apple products. If anything they make suits your requirements perfectly in features and performance and price, great. If you want to mention their advantages in a public forum, great. (I reserve the right to mention their disadvantages.)
And if you want to gently mention how this was an obscure bug that wasn't likely to affect average users, a bug that has now been fixed, that's great too. I agree. This bug obviously wasn't a huge deal.
It's just that this frothing at the mouth, "how dare you blaspheme against the mighty Apple by claiming they're responsible for their own bugs?!" segment of the Apple userbase... is getting really, really tiresome.
Reread the article then as you are wrong...
Consumer Reports was benchmarking a debug mode. That should not be a legitimate config to represent typical usage.
The long version: Safari caches downloads to safe battery life, as they should, and as any decent browser would. This is nice for users. It is not nice when you try to measure battery life by cycling through 20 pages because very quickly everything is loaded because the battery lasts forever.
So they decided to turn a very hard to find developer setting to disable caching downloads. The app now uses a lot more energy and battery life goes down. The same would happen if a user with download cache enabled doesn't visit 20 sites, but thousand sites.
Unfortunately, there was a bug that only happened when the download cache was enabled; that bug did _not_ happen for users visiting thousands of sites. And that bug killed battery life _but only for people who intentionally turned the cache off, which a normal user would not do_.
So two things happened: One, battery life went down because they turned off a feature that is intended to improve battery life. Two, battery life went down a lot because of a bug that only happened when the battery saving feature was turned off. So, entirely irrelevant to 99% of users.
So Apple released a closed-source patch for this issue and Consumer Reports is supposedly taking them on their word that the new software is legit. Is there any way to verify that the fixed software is definitely not using cached data so that the test accurately reflects the browsing habits of a user visiting many different sites over the course of the battery cycle? If not, then how do we know that Apple isn't faking the test by using cached data in a similar fashion to the way VW faked emissions tests in their diesel vehicles?
Damnit. You mean that the first result set WASN'T released just to prompt Apple to give them a bunch of cash for a BETTER result? It was really a bug? Shucks. The economy isn't working like it used to anymore. /sarcasm
I'm just going to C&P this on over:
Methodologies should assume that features do what they say they do and nothing more. If I press the hazard lights on my car and (somehow) the muffler falls off and I report that the muffler fell off while I was driving (not realizing that pressing the hazard lights was what triggered it), my methodology isn't flawed. Clearly, the car company is at fault for creating a flawed product. "But most people never even use the hazard lights! You should have tested it without touching the hazard light button!" is not a reasonable response. This entire tone and focus is steeped in the rankest of apologia via bass-ackwards thinking.
On another note: And it's entirely conceivable and reasonable that they would be using debug mode to either examine more closely how it's performing or to configure it to more closely match their other tested notebooks (perhaps they disable caching in all browsers, for instance.)
It was irrelevant for those users but it is still an Apple-created problem, not a problem with CR's methodology. And given MBP's market penetration, at least some users would have been affected by the bug.
People who are testing laptop performance should be able to use any and all existing functionality of the product without random stuff breaking. I gave a car analogy in my other reply.
What this whole story is really about is that a bunch of Apple haters got a reason to gloat, and are now feeling ripped off when their reason to gloat was taken away from them. The actual technical reasons behind the original decision and the decision to reverse the decision are beside the point.
What going to MS after using OS/X? are you fucking nuts? Maybe Linux, I wont go from a rock to a hard place.
You should be happy to know that caching remains off. All that was identified is that caching triggered a bug, it was the bug that was fixed and the test was then repeated in the exact same conditions as they always have done.
Caching didn't trigger the bug. Not caching didn't trigger the bug either. Using the "disable cache" setting triggered the bug, it didn't have actually anything to do with caching or not caching.
Seems CR is developing a propensity for trashing high profile products - Tesla and now Apple. I think they are doing this to generate "controversy" and publicity for their product.
snydely
I guess it still takes up to 10 business days for checks to clear.
The holidays must have delayed it's processing.