Marco.org

I’m : a programmer, writer, podcaster, geek, and coffee enthusiast.

On rumors that the next Mac Pro will be “really different”

Drew Baird, via Mac Rumors:

For what it’s worth - a couple of months ago I received a call from Douglas Brooks, Apples project manager for the new Mac Pro to address my concerns about the new machine. Obviously he didn’t tell me anything about the new MP, but asked me what I wanted to see. I told him expandability for extra graphics cards support, and memory expansion were at the top of my list amongst other things. His reply was simple:

“You are going to be really glad that you waited [to buy a new tower]. We are doing something really different here and I think you’re going to be very excited when you see what we’ve been up to. I can’t wait to show this off”.

First, some reasonable doubt here is warranted: it’s somebody on a forum recalling a phone conversation, months after it happened, in which an Apple employee allegedly gave vague hints about a future product. Even if Baird is credible, it’s probably safe to say that we can’t rely on these specific words. But the gist matches a common rumor that whatever replaces the current Mac Pro will be significantly different.

Lots of rumors have suggested significantly reduced internal storage and slots, relying on Thunderbolt for expansion, but the message from Apple has also been pretty clear that current Mac Pro fans won’t be disappointed by the update. I believe these are mutually exclusive.

Scaling down the Mac Pro without ruining the biggest reasons to buy it is no easy feat, and people who buy it aren’t really asking for it to be scaled down. We already have two Mac desktops that rely heavily on Thunderbolt for expansion, and most Mac Pro buyers don’t want them. Even Thunderbolt 2.0 still won’t be fast enough for extra RAM or high-performance GPUs, and the mass availability of Thunderbolt media peripherals still hasn’t happened — partly because Mac Pro owners still can’t use them. And while we can lose the optical bays without angering many people, it’s going to be tough convincing Mac Pro fans to give up internal hard-drive bays.

If the Mac Pro’s replacement doesn’t have at least 4 internal RAM slots, 2 PCI-Express slots, and 2–4 drive bays, Apple’s going to get a lot of angry professionals, and a lot of them are going to rush to buy refurbished 2010 Mac Pros.

One big question is whether they’ll still offer dual-socket configurations — their omission would anger many buyers, but not as many as those other changes, and the benefits could be substantial: they could stop relying on Intel’s less-frequently-updated 2P Xeons and make a much smaller, cheaper, cooler, more frequently updated lineup using the Xeon E3 series. But the E3, being only slightly different from Intel’s desktop chips, is limited to 32 GB of RAM, which wouldn’t be well-received in a system that has supported 128 GB since 2009.

Then there’s the Retina question. It feels like desktop Retina displays are still very far off, but by my calculations, Asus just announced a 31.5” one, as long as you usually sit at least 25 inches from it. (I sit almost exactly 25 inches from my 30” monitor, so this works.) The big limitation is GPU power, but…

Intel’s new Haswell processors promise to rectify this ailment, having made 4K support a headline feature of their integrated GPU…

Fortunately, Apple’s probably about to update their laptops to Haswell.

The other limiting factor for an external monitor is transmitting all of that video data over a cable: it would require Thunderbolt 2.0, which is coming this fall, at about the same time as the rumored Mac Pro replacement. Retina/4K-display capability in the Mac Pro’s GPU and interconnect, and a new display released with it, is a feature that many pros will pay quite a bit for.

The Mac Pro is usually only updated about every 18 months (cough), so if it can’t go Retina this fall, it might be a while. The technical needs and timing are tight for desktop Retina displays to make it in time for the Mac Pro replacement, but it does look possible, and that would indeed impress people and make us happy to have waited.

It would be unfortunate if the Haswell-generation laptops lack Thunderbolt 2.0 and can’t support an external Retina display, but the monitor is likely to cost over $2,000 at first, limiting its market to mostly Mac Pro buyers anyway. By the time the monitor is more affordable to the mass market, the laptops will be able to drive it.

Of course, it’s also possible, and probably more likely, that Apple will simply wait until the entire lineup has fast enough interconnects and GPUs to drive external $999 Retina displays before releasing them at all, but that’s probably at least 2 years away.

If Retina displays aren’t a feature of this year’s Mac Pro replacement, what will we get so excited about? What was worth skipping a major CPU generation and going 3 years without an update? And if we’re actually going to have less expansion than before in a system bought primarily for its expansion, what’s worth that?

I’m curious.

Vesper

Three friends of mine — John Gruber, Dave Wiskus, and Brent Simmons — secretly got together to build an app, and today they released the first version to the public.

You’ve probably heard about Vesper by now. I’ve been testing it for a while, and I think of it as a note shoebox with optional tagging. Many other apps, most of them for less money, allow you to write and store notes and photos, so it’s perfectly reasonable to ask why you should use Vesper over your existing text-notes app of choice.

Even for a 1.0, it’s pretty light on marketable features by 2013’s standards. It’ll lose a feature checklist comparison to almost every other popular text-notes app. Notably, Vesper can attach a photo to each note, or have photos that are notes without additional effort, which most Dropbox-syncing text-note apps can’t do. But Vesper can’t even sync to Dropbox.

From the outside, then, it’s easy to be dismissive or even resentful: How can these guys launch a relatively expensive text-note app that’s missing so many features of competing text-note apps?

Balls.

It takes balls to release an iOS app in 2013 for $4.99.

It takes balls to enter this extremely crowded category.

It takes balls to release a note-shoebox app in 2013 that has no sync, import, or export.

It takes balls to name your note-shoebox app after a cocktail nobody has heard of, then to age-rate the app “12+ for mild alcohol references” just so the cocktail’s recipe can be included in the Credits screen.

It takes balls to give your note-shoebox app a nondescript, abstract icon to match its cocktail name so nobody who sees just the name and icon will have any idea what it does or likely be enticed to find out.

And it takes balls for these three high-profile people, whose collaboration naturally earns extremely high expectations and polarizing reactions, to build and release anything together.

The best thing I can tell you about Vesper is that the app reflects its creators. I imagine they’ll add sync in time because it’s critical and very useful — otherwise, I don’t expect Vesper to get many more features.

But every feature in the app is extremely deliberate and thought-out: every mechanic, every restriction, every interaction, every animation. Every detail.

It’s a nuanced, polished app that’s pleasant to use and exudes craftsmanship. Simple flavors, executed extremely well. A vesper.

Fertile Ground

One of my favorite patterns in our industry is when the old and established are wiped out by disruption, irrelevance, or changing fashions. Like a forest fire, clearing out the old is very destructive and shouldn’t be taken lightly. But what’s left behind is a clean slate and immense opportunity.

I don’t think we’ve ever had such an opportunity en masse on iOS. After what we saw of iOS 7 yesterday, I believe this fall, we’ll get our chance.

The App Store is crowded: almost every common app type is well-served by at least one or two dominant players. They’ve been able to keep their leads by evolving alongside iOS: when the OS would add a new API or icon size, developers could just add them incrementally and be done with it. Established players only became more established.

iOS 7 is different. It isn’t just a new skin: it introduces entirely new navigational and structural standards far beyond the extent of any previous UI changes. Existing apps can support iOS 7 fairly easily without looking broken, but they’ll look and feel ancient. Their developers are in a tough position:

I don’t think most developers of mature, non-trivial apps are going to have an easy time migrating them well to iOS 7. Even if they overcome the technical barriers, the resulting apps just won’t look and feel right. They won’t fool anyone.

This is great news.

Apple has set fire to iOS. Everything’s in flux. Those with the least to lose have the most to gain, because this fall, hundreds of millions of people will start demanding apps for a platform with thousands of old, stale players and not many new, nimble alternatives. If you want to enter a category that’s crowded on iOS 6, and you’re one of the few that exclusively targets iOS 7, your app can look better, work better, and be faster and cheaper to develop than most competing apps.

This big of an opportunity doesn’t come often — we’re lucky to see one every 3–5 years. Anyone can march right into an established category with a huge advantage if they have the audacity to be exclusively modern.

I’ll be invading one as soon as I can. Here’s hoping I’m right.

Get Rid of the App Store’s “Top” Lists

At the bottom of my recent post on App Store pricing, I ended a three-paragraph footnote1 with this:

Abolishing the “top” lists from all App Store interfaces and exclusively showing editorially selected apps in browsing screens would do a hell of a lot more than trials to promote healthy app economics and the creation of high-quality software.

It was really a mental note to expand on that in a full article sometime, which I promptly forgot about. Fortunately, John August didn’t:

The App Store’s best-sellers lists hurt shoppers, developers and Apple. The charts create a vicious circle that encourages shitty business models and system-gaming. They’re a relic of a time when data was scarce. They should go away.

And Nick Dalton presented more alternatives.

The “top” list is a cop-out by Apple: it’s easy to implement and cheap to operate, and its results appear impartial and incontestible. I’m sure Apple thought this was a sweet solution.

But it has three major problems:

Since the “top” lists are featured so heavily in the App Store interfaces, many buyers appear to be buying from them as their primary store-browsing channel.3 These lists, and their mechanics, therefore deeply affect the entire app market in unsurprising ways.

The race to the bottom. Deceptive low-now, high-later pricing. Scam and clone apps. Shallow apps with little craftsmanship that succeed, but many high-quality apps unable to command a sustainable price. The “top” list encourages all of these — we’d still have them without the list, but to a substantially lesser degree.

Developers will optimize for whatever factor is being rewarded. The “top” list simply rewards developers for getting as many people as possible to buy or download the app once. There’s no reason to optimize for longer-term satisfaction or higher engagement after purchase.

It’s a lot like the Android market. Nobody — not Google, not the manufacturers, and certainly not the carriers — gives a shit if you hate your Android phone or put that cheap tablet in a drawer after a month. They’re optimizing for “top” lists, so they compete on price, flashiness, and huge retail incentives, usually at the expense of quality and long-term satisfaction.

Apple refuses to play that game in hardware. Why are they content to let it dominate their software ecosystem?

You can do better than “top” lists, Apple. Get rid of them and start rewarding great software.


  1. Because that’s how I roll. ↩︎

  2. Apple’s sub-par App Store iOS apps exacerbate this by often making it clunky or buggy to browse very deeply into the ranks. ↩︎

  3. This is all just speculation based on fuzzy data, since developers still don’t have another important metric: where buyers came from. It would be extremely helpful to even have a simple breakdown between three huge channels: browsing the App Store, searching the App Store, or following a direct link. (Of course, more specific data from each of those would be even better. But I’d happily take this elementary breakdown in the meantime.)

    Without this information, we have very little insight into why people buy our apps, which makes it harder to know where to invest our marketing efforts, how to price our app, or how to improve it. ↩︎

Far Too Much Analysis Of The Alleged New Mac Pro Geekbench Score

Mac Rumors dug this up last night. For reference:

Pre-production Macs routinely show up in Geekbench. This looks legitimate1 and reveals a number of new details: we already knew that this fall’s Xeon E5 update would include up to 12 cores per socket, but this appears to specifically confirm an “E5-2697 v2” model with 12 cores at 2.7 GHz. (The current best comparable model is the Xeon E5-2690 with 8 cores at 2.9 GHz.) An uncredited Wikipedia source confirms this as the highest-clocked 12-core chip in the lineup.

As far as I can tell, this is also the first benchmark we’ve seen of any Xeon E5 v2. It compares well: it’s pulling a score of about 24,000 out of a single socket, compared to about 16,000 from the previous Xeon generation — exactly in proportion to the core increase, even at a base clock of 200 MHz less.

The Ivy Bridge-based Xeon E5 v2 appears to perform slightly better than the Sandy Bridge-based E5 series (which never came to the Mac Pro), much like the Sandy Bridge to Ivy Bridge desktop transition, and the v2 can have up to 12 cores instead of 8. No huge surprises.

John Poole, founder of Geekbench’s parent company, isn’t blown away. This is going to be a common response: the new Mac Pro can’t blow us away in Geekbench relative to the old ones because there aren’t any dual-socket models. It’s one of the biggest compromises in the new design: easily-parallelized tasks won’t be much better, and may be worse, than on the old $5,000+ dual-socket Mac Pros.

But we don’t know if many Mac Pro buyers were getting the dual-socket models.

Many pro apps, notably including Photoshop, don’t effectively use tons of cores, so the fastest Mac Pro for many people in practice has been the single-socket, 6-core, 3.33 GHz model. (This is why my photographer wife and I chose this model for ourselves.) It scores “only” about 13,500 on Geekbench, but in single-threaded tests, it appears to only be about 10–20% slower than this new E5-2697 v2.

If there’s any disappointment to be had, it’s that Intel has made so little progress in single-threaded CPU performance since 2010.

If the new Mac Pro is offered with fewer cores but a higher clock, that might be notably faster in most workloads. That Wikipedia table claims that there will also be CPUs with 6 cores at 3.5 GHz, 8 cores at 3.4 GHz, or 10 cores at 3.0 GHz. I hope Apple offers the 8-core 3.4 GHz model, because that’s probably a better choice for most buyers than this 12-core chip.2

The new Mac Pro is also extremely power-lopsided: it will initially max out at 12 cores (almost certainly this exact CPU), which is upper-midrange by Xeon standards, but it comes with a ridiculous amount of GPU power. This is overkill to just be about future desktop Retina Displays — clearly, Apple’s pushing for pro and scientific apps to shift more of the heavy lifting to OpenCL.

If they succeed, the new Mac Pro will probably crush everything else in its price range (and the rest of the Mac lineup). In the meantime, or for people who won’t use OpenCL-accelerated apps, it will probably be an incremental Mac Pro update: similar CPU increases as every other Mac Pro update, minus most of the internal expansion. So if you have a 2010 Mac Pro, there may not be much reason to upgrade.

How compelling this will be depends on two big, unanswered questions:

  1. How much will it cost? (If every model comes with dual FireGL GPUs with lots of VRAM, this might be uncomfortable.)
  2. Will there be desktop Retina Displays? If so, will any other Macs be able to drive them?3

We’ll find out soon enough.


  1. Except for one tiny detail: the CPU is reporting itself as “Intel Xeon E5-2697 v2”, with a lowercase “v”.

    In Intel’s E3-series “V2” and just-released “V3” Xeons, the “V” is always capitalized. But in Intel’s leaked roadmap slide, it’s lowercase for the E5 “v2” series. So who knows? ↩︎

  2. If the 8-core, 3.4 GHz E5-2687W v2 is too hot to be included (it has a 150W TDP while the others top out at 130W), I’d also be happy with the 6-core, 3.5 GHz E5-2643 v2. It’d probably score “only” in the 16,000 Geekbench range but be much faster than this big 12-core chip at most tasks. ↩︎

  3. Notably, we didn’t hear about Haswell updates to the Retina MacBook Pro yet. I’m hoping that they’re holding them back until the new Mac Pro launch, they’ll also include Thunderbolt 2, and they’ll join the new Mac Pro as the only Macs that can drive a new external 4K Retina Display.

    I haven’t been following the consumer desktop chips, but it’s also possible that the iMac will receive a similar update for the same reason. It would also shake things up considerably if a Retina iMac comes out this fall, which would lure many potential Mac Pro buyers.

    But this is all wishful thinking, relying on the assumption that desktop Retina Displays in some form will be out soon, which isn’t supported by much. ↩︎

iOS 7 As Defense

Most of iOS 7’s redesign was probably decided on “valid” factors: the iOS UI was looking pretty dated and had accumulated a lot of cruft, and Apple clearly loves and believes in the polarizing new design. But it also comes with a number of convenient defensive advantages.

Since iOS’ launch in 2007, people have devoted a lot of time and money to copying the UI. Samsung, of course, is the biggest offender, but the copying has gone far beyond them: almost all modern smartphones and tablets have parts that resemble the old iOS UI. (Some have more parts than others.) iOS mimicry has even widely infected web design, especially those horrible “mobile” blog themes that try to look like an iOS app.

The entire tech industry has been homogenizing and commoditizing iOS’ appearance so much that after six years, iOS no longer felt exclusive, unique, or premium to most people: it felt like the norm. Apple needed to shake things up to keep their premium edge, and they went all-out.

Since WWDC, every iOS-imitating UI looks ancient. Soon, they’ll start to feel obsolete. Most imitating efforts will need to be redone or abandoned to look current. And what will happen if people try to imitate iOS 7?

Presumably, Apple has a few new patents for iOS 7’s interface and behavior. As we’ve seen, this won’t prevent copying, but it can at least increase the cost. Any efforts to copy the new UI are going to have a dark cloud of potential litigation hanging over them.

Copying iOS 7 is going to be a big problem for cheap hardware. iOS 7’s appearance and dynamics require a powerful GPU and advanced, finely tuned, fully hardware-accelerated graphics and animation APIs. This will hurt web imitators most, but it’s also going to be problematic for Android: while high-end Android phones have mostly caught up in GPU performance, and recent Android versions have improved UI acceleration, most Android devices sold are neither high-end nor up-to-date. The gap is much wider in tablets, and even “high-end” tablets usually have insufficient GPU power to drive their high-DPI screens.

Apple’s even attacking Android’s screen technologies by making heavy use of 1-pixel-wide lines on Retina screens. Most Android phones have high-DPI screens, too, but their feature-checklist battles have actually driven many of them to have too-dense resolutions for 1-pixel lines to be useful. And many Android phones, notably including almost all Samsung Galaxy models, cheat on resolution by using PenTile screens, which poorly render text and thin lines.

The theme is clear: iOS 7’s UI requires some of Apple’s biggest strengths, and efforts to copy it will be hindered by some of Android’s biggest weaknesses.

iOS 7 is also going to be a problem for cross-platform frameworks. Fewer assumptions can be made about the UI widgets and behaviors common to all major platforms. And any UI targeting the least common denominator will now look even more cheap and dated on iOS 7, since the new standard on the OS is so far from the old one.

But there’s also a huge win in developer attention.

In early January 2010, Google’s developer relations group sent me a Nexus One to experiment with, unprompted and with no strings attached. (This was very nice and generous of them.) Seeing a future where Android mattered, I started experimenting with different text settings for Instapaper’s web reading view to look good on its PenTile screen. On January 25, I published some of my progress.

Two days later, the iPad was announced. I never had time to think about Android development again.

Apple’s competitive message to developers during the keynote was clear: iOS is doing great, it’s the best platform to build apps for, and you don’t need to think about other platforms. iOS 7 reinforces this even further: whatever app developers were planning to do this fall is probably on hold now, because everyone’s going to be extremely busy updating and redesigning their apps for iOS 7. Anyone thinking about expanding into another platform now has a more pressing need to maintain marketshare on iOS.1

It’s a power move — one that I think Apple still has the power to pull, but just barely. We’ll see how it works out in practice.


  1. Very little of this article applies to games. Games usually build entirely custom UIs shared between platforms, they’re more financially viable than non-game apps on Android, and cross-platform game frameworks are so good that it’s relatively easy to maintain iOS and Android versions simultaneously.

    It’s obviously short-sighted to ignore games completely, because they’re very important to any platform’s app ecosystem. For a phone or tablet these days, lacking good games would be as fatal as lacking music or movie playback. But since games go cross-platform so much more easily than other apps and usually don’t take advantage of each platform’s unique strengths, they’re more of a universal tide that lifts all boats in practice.

    But games can’t carry a modern computing platform alone, just as music or movies can’t, so non-game app environments will always be important differentiators. ↩︎

All Or Nothing

A number of readers have asked me to clarify my tweet yesterday about NetNewsWire being “dead” to me, despite the recent 4.0 Mac beta. I have some complaints about the app itself, but they’re mostly irrelevant personal preferences. My statement was based on two other major concerns.

First, it’s simply too late. In three days, Google Reader will shut down and there will be no NetNewsWire client that syncs. The new Mac app is in an early beta, the new sync service is still in progress, and Black Pixel says they’re “revisiting the designs” for the iPhone and iPad clients, so we’re probably in for a pretty long wait for all three apps. This is a massive undertaking that appears to still be in its early stages. In the meantime, many NetNewsWire users — myself included — will switch to other apps, change our habits, and start building an affinity for something else.

That’s why I said it was “dead” to me: it will stop working for my purposes on Monday, it will be a long time before I can use it again, and when (if) that time comes, it may not be what I want to use.

My second concern is strategic, and it applies beyond just NetNewsWire.

In the wake of Google Reader’s demise, we now have many feed-sync services with public APIs, all of which are accumulating users and customers, including Feedly, Feedbin, FeedHQ, Feed Wrangler, Fever, NewsBlur, The Old Reader, BazQux, and probably one or two more that launched while I was writing this sentence and somehow already got added to Mr. Reader.

If your primary focus is traditional feed-reading (i.e. excluding browsing-centric, social-hybrid apps such as Flipboard), I believe launching a proprietary feed-sync solution in today’s environment is a huge strategic error.

A major strength of this multi-sync-service environment, as long as at least a couple of the services have widespread client support (which has already happened), is that users can pick and choose their favorite apps on different platforms. Google Reader gave us the same benefit: I was happily using NetNewsWire on Mac and Reeder on iOS for years, because I preferred NetNewsWire over Reeder on Mac and vice versa on iOS.

With multiple clients supporting the new sync services, we’ll retain most of the same luxury.

But if your app only syncs to your own service and nobody else’s, you’ve put up a massive barrier: for someone who likes feed-reading on multiple platforms, to switch to you, they’ll need to like your respective clients better than their existing choices on every platform they use.

It’s all or nothing. And the chance that someone will decide that you’re worth the “all” is far lower than the chance that they’d switch to just one of your apps that they like.

From Black Pixel’s vague public statements on sync, it appears likely that NetNewsWire will only sync with its own service.1 If so, they’re setting extremely high barriers: not only will customers face the all-or-nothing decision, but Black Pixel will need to ship three complete apps before any of them are very useful to people with all three devices. It’s extremely ambitious — probably too ambitious. If that’s their plan, I hope they’ll reconsider.

They’re not the only ones at risk of an overly ambitious strategy: many new services trying to replace Google Reader are taking the same gamble.

In the absence of major exclusive features or strong social lock-in, if you make a feed reader today, it better support multiple sync services. Conversely, if you’re making a service, it better have an API for third-party clients.

These strategies don’t apply in every market, of course. But the feed-reader market, with so many geeks and power users, is much more likely to own multiple devices, want native clients on each of them, and require syncing. Feed readers all do pretty much the same things and differentiate primarily on interface choices, so potential lock-in effects are minimal and the market appreciates client choice.

The post-Google-Reader market is already established as a multi-service environment, so the chances of any one company locking in another majority are (thankfully) low. If you try, you’ll be fighting an uphill battle.


  1. Black Pixel declined to comment on their sync plans for this post. ↩︎