Om du inte får den du älskar, får du älska den du får.
My previous MacBook Pro was a mid-2014, 15-inch model, purchased somewhere shortly after the introduction of the ill-fated USB-C MacBook in 2015. In the first few months, I saw the butterfly keyboard essentially crash and burn in the press, the USB-C “dongle life” get ridiculed, and the Force Touch trackpad also land in a MacBook Pro revision.
Then, later on in the year, the first MacBook Pro with a Touch Bar was introduced and the butterfly keyboard and USB-C adopted wholesale. That’s when I got the sinking feeling that, without serious backlash, there’s a significant risk Apple has made up its mind and is going to turn a corner. Very rarely are hardware changes, once introduced, reverted. Floppy drive and pre-USB IO goes, it’s done. Unibody manufacturing, glass trackpads? Kept forever. Occasionally, a whole line is dropped, like the PowerMac G4 Cube, or the rare first-time-around MacBook aluminium is retconned into a MacBook Pro, and a unibody plastic MacBook introduced. Change is possible, but was it likely?
Odds against me, having just gotten a new one, I decided to sit this out. (I’m also sitting out developing for iOS while they only have an App Store, so, you know, I’m that guy.)
I wasn’t the only one to miss the old MacBooks Pro though. Even if the whole keyboard debacle can be elided at this point (and strangely underreported by the non-Mac parts of the web, maybe under the impression that being failed by a keyboard was a mark of the parodically effete, and that surely not even Apple could so botch a keyboard), the picture of the old MacBook Pro generation being a stalwart warrior of comfort and utility earned some cultural purchase and nodding agreement. As much as you may like Thunderbolt 3 or USB-C, no one likes the dongle life per se. And as updates didn’t seem to address much aside from bumping specs (and prices), thermals cast yet more shadows, leaving recent Intel chips to throttle in a sweltering aluminium prison.
Earlier this year, rumors started spreading about another model which would finally clean up some of the weaknesses, and the days thence filled with silent prayer: let this bastard hold until the new one’s out. The rubber feet cracked and peeled off, and the battery bloated to push up against the trackpad, rendering it cumbersome to press and starting its own countdown to a required exchange (which would be the third of its kind for that generation of MacBook Pro for me, just as a reminder that they had their issues too). But my fingers now bounce on a reborn scissor-switch keyboard inside a Space Grey chassis; this is the very first MacBook Pro 16”. Our long national nightmare is over.
I find myself in the fresh embrace of multiple new technologies that aren’t really new to this model. The SSD is now NVMe-based, lapping the previous model by three or four times in both read and write performance. The display is capable of going almost disgustingly bright and covering the DCI P3 color space, although I’ve never discerned going to or from such a display.
Also new-but-not-really are several advances in input and human-computer interaction.
The idea behind the Touch Bar is, in isolation, admirable. F1 through F12 are originally remnants of terminal clients from the 70’s, and we deserve a clearer instrument for what is now often labeled “commanding”, the act of issuing an instruction to an application. Although I disagree with its implementation, and my immediate reaction to the physical Escape key being added back was that they were now ¹⁄₁₃ of the way to a solution, I have tried to give it its due.
To me, there are two major usability flaws. The first is that the commanding is constantly changing and dependent on state. This may sound strange, since this dynamism is the primary observation from which the Touch Bar springboards. But at least my eyes are constantly on the screen. The Touch Bar morphs to offer commands as you make changes or make your way across windows and input states. Knowing that some button is in some place will take repetitive learning.
Combined with the second flaw, that any sense of tactile or haptic feedback is missing, it adds up to an environment hostile to pattern forming. When you look at a tablet or a phone, your eyes are already looking at where your fingers are tapping. There’s no second step needed to acquire a new target and verify that it’s what you thought — you just do. The Touch Bar, at least set to the app-contextual “App Controls” commands, constantly needs these affirmations, landing you in the odd situation of diverting your attention from a ~1120 points high screen to one that is 30 points high, and from a tactile lattice of keys that we all know to operate from muscle memory built up over decades to a shallow strip with no feedback except visual. Since effective commanding is often about being able to do things blindfolded, this gets in the way.
The Touch Bar does open a few doors – it can display colors (and therefore also emoji), it can offer fine motor control without affecting the mouse cursor or requiring precision in two axes. If you can fit your workflow into scrubbing or sliding or panning, you can operate it without needing to look, allowing a new kind of gesture not offered by the trackpad with both precision and agility, and at least local efficiency gains.
I consciously have been keeping track of all F-key presses, and they add up to 0 now. (In sharp contrast to Escape key presses, which I lost count of the first day.) I mostly don’t use the Function keys themselves at all in my daily routine; many people do, especially developers, gamers, professional application users (which this thing is ostensibly for) and double/triple-booting multi-platform users, but not me. What I do use are the global control commands – controlling volume, music playback and brightness (both screen and key backlighting – the new backlighting looks much better). I use enough of the commands that they don’t fit well into the condensed right-edge “Control Strip” partition, which essentially kills the App Controls view for me, since the extra command to show the full Control Strip is to tap the smallest known target in the whole Touch Bar, an expander chevron sitting right next to one and often two commands I don’t wish to activate.
Set to show the Expanded Control Strip at all times, the predictability helps, but at what cost? An opportunity lost – App Controls tossed to the wolves because of how they shadowed the, for me, much more common global controls; the Touch Bar only existing to poorly and lossily reincarnate what was previously there. “So don’t get a Touch Bar” is a valid complaint, and if I could have gotten a model with plain function keys, I certainly would have. Or, better yet, a model with a Touch Bar set in a row beyond the function keys. Certainly in the 16” model, there’s room for another row up top, maybe at the cost of being slightly uncomfortable to reach for.
If you live your computing life inside the auspices of the Touch Bar’s functionality and manage to use it to the hilt, you may well feel fonder of its flexibility than I do, especially if it adds significantly to your personal workflow. For me, it just plain doesn’t – and I miss the utility, tactility, predictability and dependability of what was there before. It is tempting but harsh to say “the reason it’s called a bar is because it makes you want to drink”, but it nevertheless sums up my experience with it even after giving it a chance.
Touch ID exists and works. It works well, it works fast, it’s integrated in many places, it’s available to third-party applications and it’s convenient, even making double-clicking the Apple Watch side button (which also serves as authentication) feel like a comparative struggle. It is the best kind of feature, implemented in the best kind of way; it is also incredibly boring in its perfection.
The main conceit of the Force Touch trackpad — besides having an eyebrow-raising name — is the observation that administering haptic feedback from below (using the Taptic Engine linear actuator) feels, to the finger, indistinguishable from the trackpad button actually depressing, in the way of traditional “diving board” trackpads. For this to work, force has to be sensed, and the click simulated beyond a threshold, and this sensing is also used to provide “deeper” presses with alternate meanings.
As a side note: Being used to a trackpad misbehaving with a swelling battery underneath it, my clicks were forceful/deep enough to naturally trigger deep presses unless I took care to go feather light and feel for the activation point during the setup process, which lead me to activate the “deep press lookup” feature on many links, instead of merely following them to skip steps — a fate I posit is both common and confusing, since the force touch capability is not obvious and not explained along the way.
The trackpad allows configuring a click stiffness, as well as turning off deep presses entirely. In theory I like the idea of deep pressing to trigger lookups, and also to “press harder” to fast forward even faster, but I have switched the lookup feature to use the old three finger tap instead, since it was too easy to activate by accident; similarly, dragging files in the Finder is likely to accidentally trigger renaming. I am hopeful that, having some haptic feedback, there is room for muscle memory to relearn these parameters, unlike with the Touch Bar.
Lastly, the trackbar is also comically huge at first glance. The palm rejection is stellar – I have not given related errant input once in the form of clicks. As a consequence of its virtual nature, it also allows clicking at any point on its surface, which comes in handy but doesn’t yet readily occur to me.
Aside from the Escape key, the previously described features were all present and in full force on the models I resigned to stay away from. Readily absent, however, is the butterfly mechanism. The scissor switch is nicked almost wholesale from the standalone Magic Keyboard, with the notable adjustments of a slightly lower profile key and reintroducing the inverted-T layout of the arrow keys, which serves as a necessary locating feature. It feels different than the previous scissor switch mechanism, but is so close that it will take weeks to game out if there even is a winner in comparison.
San Francisco ably replaces VAG Rounded in the clear legend (and for what it’s worth, Apple maintains their decision to keep separate keyboards for Swedish/Finnish, Danish and Norwegian lettering, bucking the trend of laptop manufacturers smearing all Nordic languages onto the same key caps into an unreadable mess leaving some keys looking like an explosion in the glyph factory). Overall, it’s just plain good, and returns to the Apple laptop keyboard tradition of yore of being unremarkable in its silent exceptionalism.
I am not the right person to objectively judge the performance of this model, due to my upgrade path. A lot of things happen in five years. But thanks to the ability to load it up with significant GPU memory (a presumed funnel point; individual Safari tabs would often glitch out massively post-Catalina upgrade), RAM and SSD, there’s once again a lot of headroom for many medium-sized tasks, and I haven’t ran into being bottleneck by really anything. It plows through everything with ease. This makes sense: the thermal envelope has been raised by reconfiguring the inside to free up more airflow and allow the fans to do a better job, leading the CPU to throttle less.
The SSD upgrade pricing is extortionate, but notably less so than before a price adjustment earlier this year, and the 8 TB upgrade option is somewhat reasonable in its ridiculousness due to the lack of such options across the industry. I am expecting this model to last at least as long as the previous model, and to deliver solid performance for at least the first four years.
The model also has much improved speakers and microphone; I will not be mentioning them due to not listening a lot to speakers directly. But the going consensus is that they are genuine technical achievements and worth reading up on. And on the flip side, the 720p camera is laughable, and a pro laptop launched at this date should include Wi-Fi 6 (802.11ax), already present in this year’s iPhones.
There are four Thunderbolt 3, USB-C ports and one 3.5 mm headphone jack, and that’s still it. USB-C is more commonplace now, so the market has moved the needle and improved Apple’s justification.
It still carries all the previous problems of a budding universal connector, though; it connects directly to almost nothing, even for the many people who have made the deliberate transition. Not only early 2015 MacBook Pro models but also most laptops in the marketplace today come with additional ports for USB-A (the traditional plug), HDMI and, not unusually, an SD card slot. MacBook Pro also hosted the magnetically attached MagSafe 2 power connector, which absense is sorely felt not only in ease of detachment but in a charging LED indicator. (Its cable shroud is not known for its reliability, but I predict neither will the identical form used on the USB-C power cables be.) There are third-party alternatives available, but all of them look significantly clunkier, and almost none provide the full 96W offered by the included AC adapter.
Phil Schiller offers a reasonable-ish defense when pushed – that in their place, the connectors that deliver the widest landscape of opportunities are the Thunderbolt 3 ports, an argument clothed in the “giving the pros what they wanted” angle Apple chose for the release of this model. To me, it sounds off – Apple already makes plenty of sacrifices in the name of a more compact, more clean product. If giving the pros what they wanted was a high priority, delivering at least one of MagSafe, a USB-A port, an HDMI port or an SD card slot should have made the cut. (As would making the Touch Bar optional.) And the workaround dongles everyone will have to if not carry regularly, then at least acquire, do nothing to bolster a compact setup with a clean profile, and everything to highlight its inadequacy.
The major determinant of the product, the 16-inch display, has mostly gone unmentioned. Like Touch ID, it performs admirably in a boring way; unlike it, there are interesting details to mention. The slimmer bezels do a lot to make every other MacBook model look dated. True Tone (adjusting the white point to ambient lighting) is present and accounted for, and works as well here as every other Apple device it’s been implemented in (including the Touch Bar). The leap in resolution is not tremendous, but the added space is welcome. Although Apple is late to bring its ProMotion variable refresh rate (or high refresh rates of any kind) to any MacBook, it does offer manually selectable additional refresh rates that are integrally divisible with common refresh rates and thus avoids pull down.
The Catalina/iOS 13 feature Sidecar — using a recent iPad as a secondary display — is not new to this device, but it is the first time all the requirements have lined-up for me to be able to test it. It’s a wireless connection powering a retina-level display; knowing AirPlay’s performance, it shouldn’t work, but it is undeniably smooth with no discernable lag and runs miles around AirPlay in my brief testing (I’m not in the target audience). Look for matching wireless technology in a future Apple TV to upgrade AirPlay’s performance. I did see the Touch Bar having issues showing up, and also the triggering of another glitch, bringing us swiftly to the well-known nugget:
I have seen two glitches, both related to waking the MacBook Pro up. Some of the time, the display color reproduction gets out of wack with a saturated posterization and toggling True Tone, the refresh rate or automatic brightness adjustment is usually enough to get it to snap out of it. (The Sidecar tie-in is that changing its settings could result in triggering this condition, as well as causing the MacBook Pro display to forget its resolution and revert to the default.)
Worse, a significant fraction of the time, waking it from sleep will show either a login window that hasn’t finished initialization (no placeholder label in the login password field, for example), or a partially intact scene from one of the full-screen spaces, atop which a cursor forever beachballs, blissfully ignoring clicks, keys, Touch ID or Apple Watch unlocks. There’s no indication that this hang ever resolves.
(Both of these issues present on macOS Catalina 10.15.1, Supplemental Update included.)
A machine is more than the sum of its parts. It can also be a way of communicating changed priorities. There is no contest that this is the most powerful MacBook Pro ever offered; it would be mad for any new MacBook Pro to be less powerful, but how this model manages to be that much more powerful than its immediate predecessor while still using the same CPUs is noteworthy.
It also reconsiders blind allies, some of which I was starting to give up on as lost causes. It makes decisions that I love, and it makes decisions, still, that make me roll my eyes. In ways that previous MacBooks Pro have been sub-par, it is thoroughly, stupendously good, but it is also shockingly expensive compared to what it should cost. And at least during testing, it had impactful, worrying glitches, my recommendation hinging on their timely resolution.
In a sentence, it is both perfect and stupid, rolled into one. It is not the computer I wished for, but it has things in it that I wished for, and it is enough to get me through the next few years and feel like charting that course wasn’t a fool’s errand — money given to a company that’s still willing to listen to its customers, even if it could work on its humility and attitude.
As the old Swedish saying goes: If you don’t get what you love, you get to love what you get.
(As tested: 2.4GHz 8‑core Intel Core i9; 64GB of DDR4 RAM; 4TB SSD; AMD Radeon Pro 5500M with 8GB of GDDR6 memory.)