This review seems to indicate that the 2015 MacBook has similar issues to the original 2008-era MacBook air: lots of potential, but still worth waiting for a 2nd generation product to work out some of the issues.
Personally, though, I'm just excited that Apple finally offers a laptop in multiple colors. (And without a price premium, for anyone who remembers the "black tax"[1])
True. But in this specific case the whole concept doesn't seem "fully baked" yet. The USB-C was a great idea, but just a single USB port on the entire machine that worse still will be in-use by the charger? And $80 adapters if you want a second usable port?
This line of machine has potential. But it needed minimum a second USB port for mice/keyboards/hubs/USB keys/etc. The Surface Pro 3 only has a single USB port, but in that case that single port isn't also used for charging so it is "workable" (and it has a second Display Port also, which this also lacks).
Battery life is certainly part of it; RAM that isn’t being used still draws power.
But isn’t the whole idea behind an ultra-portable that you make sacrifices for greater portability? The Pro isn’t getting dropped from the lineup…
Augh why are we still having discussions about the specs on low-end/ultraportable laptops? Why do people think that every single laptop model should be tailored for their personal needs?
It depends what your priorities are. If lightness and thinnes and battery life are top priorities, and making the RAM upgradeable or increasing the space on the board to provide that as an option compromises those, then a machine that makes that compromise is hobbled. The whole point of having more than one laptop design in your lineup is that they make different design tradeoffs.
However, enough RAM prevents additional disk operations, which draw substantially more power, be it an SSD or spinning disk.
Also, RAM typically draws very little power compared to other components (screen, etc, etc).
The reason why we are still having discussions about the specs on low-end / ultraportable laptops is that they are so often almost useful, but take things one step too far.
I had a similar problem with the MBA I bought (2012?), in order to get the maximum amount of RAM (4GB), I had to max out the expensive SDD too. To get 50 USD worth of RAM, I had to spend closer to 300+USD extra. And the biggest problem I have with it today is still lack of memory - it's fast enough for what I use it for, as a travel laptop (I run Windows 7 on it, mind).
It's worth noting that, for some reason, Intel's specifications are almost always wrong when it comes to how much RAM you can use with a given processor.
I've had a number of computers (one Apple, one non-Apple) where all the docs claimed that the max RAM was $X GB, when in reality it would work fine with $2X GB.
I suspect Intel's listed max RAM specifications are based on some other engineering considerations than what is actually possible.
Developers often use much more than 8GBs of RAM, especially if they are running VMs, databases, compiling code, running tests, IDEs and emulating smart phones.
More and more I'm seeing a distinction in requirements between "user machines" and "engineer machines".
My 4GB MacBook Air 11" is slow with Parallels, postgres, apache, emacs and the usual programs (mail, chat, multiple browsers etc.) -but usuable.
Even my 3-4 years old MacBook Pro with 8GB was a little slow, but processor speed was the culprit.
My year old MacBook Pro (2,4 i7) with 16GB is fast, but I have more memory then I actually need...
I've even used it with three simultaneous Parallels instances (FreeBSD Server, FreeBSD CI and Windows 7 client) and it runs fine.
8GB memory is not a dealbreaker for me, I'd buy the new MacBook any day :)
I dunno--even as a developer (JVM, some Ruby) I honestly don't use that much RAM anymore now that I don't work for a company that thinks it is a grand idea to load eight gigs of crap into memory just to start the application server. I think we're moving towards much more composable systems that have much less need for fat memory specs.
Even when working on my game in Windows I only have 4GB allocated to the VM (VS2015, C#).
Differentiation. The MacBook is designed to be the thinnest and lightest possible laptop that provides adequate performance for the majority of users. Users who need more ports or more performance are expected to opt for the 13" MacBook Pro Retina, which provides a far better price-performance ratio.
Not unreasonably, Apple assume that most people who need large amount of RAM will also need a fast processor; Adding the option for 16GB on the MacBook would increase the size of the logic board for the benefit of a very small niche of users.
Jony Ive has always been ruthless about avoiding feature creep, going back to the original G3 iMac and the highly contentious decision to drop legacy ports and the floppy drive.
I wonder if people realize Apple could've offered the same product for probably no more than $800 at most using its own next-gen A9 processor with similar performance. All it needed to do is focus earlier on on ARM portability for its Mac apps.
switching architectures ain't something you do just so you can get a temporary price reduction in one product! it's kind of a big deal†. (& apple knows exactly what's involved, since they've done it twice already. (68k => ppc => intel.))
i don't think we'll see arm mac's unless/until intel definitively becomes uncompetitive.
†in an ecosystem where distributing arch-specific binaries to end-users is still a big deal
It looks like a perfect laptop for programming (tinkering).
Most of the time I just ssh to a remote server or keep a window with persistent mosh connection. All my stuff on the server is only tmux attach -d away!
I suppose I could compile small to mid sized programs on the Macbook itself. Benchmarks indicate it has enough power for that. Compiling something like boost or linux kernel might take a while but I don't do that very often.
In the end, I don't know. I'm keen to give it a go. I switched to 13" rMBP from Air for the retina screen and I still remember how clunky it felt.
Probably not enough to be my only laptop but wouldn't mind swapping my 13" rMBP for 12" Macbook + 15" rMBP (when quad core Broadwell or Skylake comes out).
From the article you're notionally commenting on: Ultimately I don’t find the MacBook’s keyboard to be any different to type on as far as key size and feel goes (even with my large fingers), but your mileage may vary.
I spent some time with the keyboard. It's not terrible. It's very different in feel from the Air/Pro, especially when typing fast. It takes some getting used to but I would not complain if this keyboard made its way to the MacBook Pro.
The loss of travel is the sticking point. The only reason to lose it on the Pro would be to standardize the feel across product lines and improve manufacturing efficiency.
I use an HHKB2 Pro and the IBM Model M on desktop computers, which are lightyears different in travel compared to scissor switches. But I've also become pretty comfortable touch typing on a full size iPad, and the new MacBook keyboard feels somewhere in between typing on an Apple scissor switch keyboard and an iPad.
Interesting you say this, since pretty much everyone agrees that this laptop is nothing more than a "facebook" machine. Has a slow processor, low amount of ram, awful keyboard, etc.
My web browsing seems to consume obscene quantities of ram (there was an interesting article a little while ago about how Adblock can eat 400mb _per iframe_) -but them I am one of those tab-hoarding types.
Ironically (incidentally) Facebook seems to be one of the worst offenders. If I sign into facebook and then forget to sleep my desktop, I can get back from a weekend away to find firefox has eaten 15gb of ram and a painful amount of swap. Chromium seemed to be just as bad last I tried it.
Considering that nowadays every tab is basically an independent application (in Chrome, literally), if you have 20 tabs open is equivalent of having 20 separate applications open, it's kind of pointless complaining about RAM usage of the "web browser".
I just think it's odd to say because my development/gaming desktop was just upgraded to 8GB of RAM from 6GB a few months ago and I never close tabs or shut anything down until something crashes. So saying 8GB of RAM is only good enough for Facebook... you should see my Steam library that likes 8GB of RAM just fine.
Maybe the problem isn't that Facebook takes 15GB of RAM, it's that Firefox or Chrome will take literally as much RAM as it can get. I should try surfing Facebook on the 128GB servers we have at work and report back my findings...
It has the same amount of RAM as most ultrabooks (the base model it has more than the base Air models or Dell XPS ones, actually).
The keyboard is shallow, but I haven't seen it described as "awful" or even "bad" anywhere.
Besides the fact that it's probably good enough for things that are not compiling complex applications/gaming/encoding/simulations (heavy loads sustained in time), it is more than enough for the uses that the parent wrote about (remote ssh/compiling small programs).
The CPU mark scores seem to be pretty good. (I wonder whether the turbo boost improves benchmark scores.) For something bursty like compilation, it should be perfectly usable, particularly when combined with the SSD.
Not sure about the RAM though... RAM requirements are climbing slowly enough that 8GBytes is probably fine, but it's still a bit stingy.
i had this happen on a macbook i used some weird cleaning agent on the screen with; this was maybe 3 or 4 macbook's ago -- since then i'm really picky about what i use to clean the screen, and i haven't had the problem.
obviously it's not definitive, but i suspect cleaning products may be to blame, or blame in a large number of cases.
Yes this is what Apple says. The fact is that all screens have serious structured patterns imprinted from the keys - independent on the cleaning method.
Having worked my way through an engineering degree as a lab manager for an optical lab, this looks like what happens to an A/R (anti-reflective) coating when harsh cleaning chemicals are used (Windex, etc...).
On mine, I would just use a water-dampened, clean, soft towel if I need to clean my screen. And by clean towel, I mean clean. I don't even set it on the counter as it can pick up particulate matter. And by soft, I mean never a paper towel. Ever.
Its true, this issue needs to be discussed more openly and a lot more people need to know about the problem - because it really does degrade the value, and overall utility, of what could/should be a premium hardware product. I just paid eu2500 for my new rMBP 4 months ago and it already looks like its going moldy ..
My best guess (typing this on a screen that has those exact issues) is that the screen is reacting to salt and oil from hands. My previous white macbook had the same (but worse) problem. I think any glossy screen had this problem but maybe I'm completely wrong?
That very first picture looks like an iPad with a keyboard attached. I wonder if they'll ever put the logic board in the screen housing and make the bottom piece all battery.
I liked the Windows performance page. Very useful. I'd love a Windows battery life page.
Macs running OS X get very solid battery life (8-12 hrs). Running Parallels they get "aright" battery life. Running bootcamp with any version of Windows tanks them into the ground (2-3 hrs) due to Apple's poorly optimised Windows drivers and bugs (e.g. running the cooling fan at max constantly).
Microsoft has done a lot to improve Windows' battery consumption (with the Surface Pro 2 & 3 getting around 8-9 hrs depending on who you believe), so Apple cannot continue to claim that Windows itself is the problem. Dell, Lenovo, and others have all been able to produce Windows-based laptops and tablets with decent battery life, why can't Apple?
Why would they? They are making hardware to work with their operating system. Do you blame PC makers who have bad battery life on Linux because there are bad open source drivers? Or PC makers that don't run well as a Hackintosh?
> Do you blame PC makers who have bad battery life on Linux because there are bad open source drivers?
In a word, yes. I do.
But the reality on the ground is that a lot of those hardware manufacturers aren't advertising Linux support as a "feature." Apple does advertise Windows support as one.
Ever since the switch to Intel chips, Apple has billed all of their Macs as being able to run Windows and that is a big selling point for them. So, I don't think it's out of line to ask them to make proper drivers for running an operating system that they do support.
IIRC a bug causes the CPU to sit at 100% constantly, that's why the fan runs at max. A sleep-wake cycle after boot temporarily fixes that problem and the battery life increases.
Apple could have moved the Magsafe mechanism an inch or two down the line from the USB-C connector, making it a 2-piece cable much like the safety breakaway mechanism on wired XBox controllers.
New thing here: NVMe SSD!
I've read a bunch of other reviews and none mentioned that. Kudos to AnandTech for thinking about that detail. Thinking about that in conjunction with the storage benchmarks is interesting, faster in most tests but a little slower in random-write.
Now I want that in the next -Pro version to be my next development machine.
69 comments
[ 3.2 ms ] story [ 140 ms ] threadPersonally, though, I'm just excited that Apple finally offers a laptop in multiple colors. (And without a price premium, for anyone who remembers the "black tax"[1])
[1] http://www.512pixels.net/blog/2015/2/future-classics-the-bla...
This line of machine has potential. But it needed minimum a second USB port for mice/keyboards/hubs/USB keys/etc. The Surface Pro 3 only has a single USB port, but in that case that single port isn't also used for charging so it is "workable" (and it has a second Display Port also, which this also lacks).
But isn’t the whole idea behind an ultra-portable that you make sacrifices for greater portability? The Pro isn’t getting dropped from the lineup…
Augh why are we still having discussions about the specs on low-end/ultraportable laptops? Why do people think that every single laptop model should be tailored for their personal needs?
There's compromises, and then there's hobbling.
Also, RAM typically draws very little power compared to other components (screen, etc, etc).
The reason why we are still having discussions about the specs on low-end / ultraportable laptops is that they are so often almost useful, but take things one step too far.
Can't find the reference for it on Intel's website but it was discussed at length on the Accidental Tech Podcast (atp.fm) by John Siracusa.
I've had a number of computers (one Apple, one non-Apple) where all the docs claimed that the max RAM was $X GB, when in reality it would work fine with $2X GB.
I suspect Intel's listed max RAM specifications are based on some other engineering considerations than what is actually possible.
I own a 3 year old MacBook Air 11" with 4GB RAM and that works surprisingly well. It even runs Photoshop fairly good.
I'd say 8GB is ok unless you do extensive image or video editing -and then the MacBook (or the Air) is not the right machine any way.
More and more I'm seeing a distinction in requirements between "user machines" and "engineer machines".
Even my 3-4 years old MacBook Pro with 8GB was a little slow, but processor speed was the culprit.
My year old MacBook Pro (2,4 i7) with 16GB is fast, but I have more memory then I actually need... I've even used it with three simultaneous Parallels instances (FreeBSD Server, FreeBSD CI and Windows 7 client) and it runs fine.
8GB memory is not a dealbreaker for me, I'd buy the new MacBook any day :)
Even when working on my game in Windows I only have 4GB allocated to the VM (VS2015, C#).
Not unreasonably, Apple assume that most people who need large amount of RAM will also need a fast processor; Adding the option for 16GB on the MacBook would increase the size of the logic board for the benefit of a very small niche of users.
Jony Ive has always been ruthless about avoiding feature creep, going back to the original G3 iMac and the highly contentious decision to drop legacy ports and the floppy drive.
Just like Microsoft did with Windows RT?
i don't think we'll see arm mac's unless/until intel definitively becomes uncompetitive.
†in an ecosystem where distributing arch-specific binaries to end-users is still a big deal
Most of the time I just ssh to a remote server or keep a window with persistent mosh connection. All my stuff on the server is only tmux attach -d away!
I suppose I could compile small to mid sized programs on the Macbook itself. Benchmarks indicate it has enough power for that. Compiling something like boost or linux kernel might take a while but I don't do that very often.
In the end, I don't know. I'm keen to give it a go. I switched to 13" rMBP from Air for the retina screen and I still remember how clunky it felt.
Probably not enough to be my only laptop but wouldn't mind swapping my 13" rMBP for 12" Macbook + 15" rMBP (when quad core Broadwell or Skylake comes out).
I use an HHKB2 Pro and the IBM Model M on desktop computers, which are lightyears different in travel compared to scissor switches. But I've also become pretty comfortable touch typing on a full size iPad, and the new MacBook keyboard feels somewhere in between typing on an Apple scissor switch keyboard and an iPad.
Ironically (incidentally) Facebook seems to be one of the worst offenders. If I sign into facebook and then forget to sleep my desktop, I can get back from a weekend away to find firefox has eaten 15gb of ram and a painful amount of swap. Chromium seemed to be just as bad last I tried it.
Maybe the problem isn't that Facebook takes 15GB of RAM, it's that Firefox or Chrome will take literally as much RAM as it can get. I should try surfing Facebook on the 128GB servers we have at work and report back my findings...
It has the same amount of RAM as most ultrabooks (the base model it has more than the base Air models or Dell XPS ones, actually).
The keyboard is shallow, but I haven't seen it described as "awful" or even "bad" anywhere.
Besides the fact that it's probably good enough for things that are not compiling complex applications/gaming/encoding/simulations (heavy loads sustained in time), it is more than enough for the uses that the parent wrote about (remote ssh/compiling small programs).
Not sure about the RAM though... RAM requirements are climbing slowly enough that 8GBytes is probably fine, but it's still a bit stingy.
obviously it's not definitive, but i suspect cleaning products may be to blame, or blame in a large number of cases.
edit: knock on wood.
On mine, I would just use a water-dampened, clean, soft towel if I need to clean my screen. And by clean towel, I mean clean. I don't even set it on the counter as it can pick up particulate matter. And by soft, I mean never a paper towel. Ever.
Buyer beware!
Macs running OS X get very solid battery life (8-12 hrs). Running Parallels they get "aright" battery life. Running bootcamp with any version of Windows tanks them into the ground (2-3 hrs) due to Apple's poorly optimised Windows drivers and bugs (e.g. running the cooling fan at max constantly).
Microsoft has done a lot to improve Windows' battery consumption (with the Surface Pro 2 & 3 getting around 8-9 hrs depending on who you believe), so Apple cannot continue to claim that Windows itself is the problem. Dell, Lenovo, and others have all been able to produce Windows-based laptops and tablets with decent battery life, why can't Apple?
Then why provide Bootcamp at all?
> Do you blame PC makers who have bad battery life on Linux because there are bad open source drivers?
In a word, yes. I do.
But the reality on the ground is that a lot of those hardware manufacturers aren't advertising Linux support as a "feature." Apple does advertise Windows support as one.
http://arstechnica.com/apple/2015/04/using-the-retina-macboo...
Instead of calling it a 'dang thing', which seems incredibly mean, why not just say the keyboard wasn't right for you?