Ersatz Free Trials

On Monday Apple announced that they are officially supporting so-called “free trials” for non-subscription apps. The reaction has been a breathless celebration that Apple has finally relented and given developers something we’ve been asking, no begging, for since the dawn of the App Store.

But what really changed? Not much. Apple announced no functional changes to the way the apps are categorized, how pricing is conveyed to customers, or how the physical transaction of downloading, trialing, and potentially purchasing an app takes place. What they did announce is a change to the App Review Guidelines, adding a bullet item to section 3.1.1 describing a kind of ersatz substitute for actual free trials, built on the in-app purchase system:

Non-subscription apps may offer a free time-based trial period before presenting a full unlock option by setting up a Non-Consumable IAP item at Price Tier 0 that follows the naming convention: “14-day Trial.”

This change to the review guidelines is fantastic, because it will give app developers greater confidence that such a workaround will continue to be approved by Apple. But the practice of offering free trials in this manner is not new, and is not particularly great by any stretch of the imagination.

The Origin of Ersatz Free Trials

To my knowledge, the first developers to come up with the idea of using in-app purchases to approximate free trials were The Omni Group. In September 2016 they wrote about a novel solution for the long-standing absence of free trials (and upgrade pricing) on the App Store, which is based on providing a baseline free download and unlocking premium functionality through in-app purchases:

With the original download free, we can implement any pricing options we want to offer customers through In-App Purchases. We can offer our standard unlocks of Standard and Pro, of course. But we can also offer a free 2-week trial which unlocks all of the features of Pro and Standard, letting you freely choose between them.

Many of us who had also been waiting for App Store support for free trials and upgrades waited in anticipation to see whether it would really work. Would Apple actually approve such a use? Would customers understand it? Would the App Store Infrastructure reliably handle the approach? The answer, as it turns out, was yes. Sort of.

Omni’s approach worked well enough that developers of other apps soon followed suit. MindNode 5, Acorn 6, and Sparkle all launched as “free” App Store titles that can only be substantially improved with in-app purchases ranging from free, timed trials to paid, permanent feature unlocks.

The approach seemed to be gaining momentum so in mid-2017 as I was looking forward to the release of MarsEdit 4, I decided that I would embrace the same idea. Since December, 2017, I have sold MarsEdit 4 as a free app with in-app purchases for free trials, free upgrades for recent purchases, discounted upgrades, and full-price upgrades. In many ways the change has been a revelation. It’s a great relief to be able to offer my customers nearly as many trial and pricing accommodations as I can offer directly through my own store.

The Problem with Ersatz Free Trials

While I’ve enjoyed many of the upsides of the Omni approach, I’ve also had the opportunity to appreciate the many downsides. You might say it’s “a pretty sweet solution” for offering free trials.

I think it’s particularly important, in the face of all the celebration this week about Apple’s perceived changes to the App Store, to understand the many ways in which this solution falls short of what many developers still hope for: bona fide support for real free trials in the App Store.

In summary: none of the mechanics of supporting ersatz free trials are substantially supported by the App Store. Every aspect of the solution is bolted on to a system which was not designed for, yet is somewhat admirably being used to simulate real support for free trials. Let me elaborate by listing several shortcomings and how they affect both users and developers in significant ways. Just off the top of my head …

  • Paid apps are listed as free, even though payment is required to unlock core functionality. This is confusing to many users and leads some to a feeling of bait-and-switch, and that they’ve been betrayed by the developer. This is particularly problematic with apps whose price points make them most suitable to free trials. MarsEdit is $50, so some users who download the “free app” are understandably annoyed when the first thing they learn is that it will cost a significant amount to unlock it.
  • Bulk purchase programs are unavailable. Apple’s Volume Purchase Programs for business and education are based on a system of allocating a certain number of “primary” App Store products to an institution. In the case of a free app with paid in-app purchases, there is no mechanism by which a school or company can for example purchase 500 copies of MarsEdit from the App Store. They can “purchase” 500 free copies and then proceed to unlock each copy individually through the in-app purchase dialog in each app. This is a particularly unfortunate limitation for apps that are uniquely suited either to education or to business uses.
  • Family sharing is unavailable. For the same reasons that bulk purchases are off the table, a developer who wants to allow families to purchase an app once and share it among their family’s devices and accounts is unable to do so unless they sell their app with a fixed, up-front cost in the App Store.
  • Not applicable to all app types. Although Apple doesn’t explicitly state it in their revised App Review Guidelines, I strongly suspect that a continuing requirement for ersatz free trials is that the app must continue to function in some way as a perpetually free, unlocked app. For document-based apps such as Omni’s, they went with an approach whereby the app becomes a read-only document viewer when it is not paid for. In MarsEdit, I took a similar tack by allowing all features to function except for publishing changes to blogs. In many cases it is possible to contrive a free/paid functionality divide, but for some apps it would be very awkward, or maybe impossible to do so.
  • Apps are ranked and featured in the wrong charts. A problem rooted in these paid apps being listed as free is that there is no natural place for them to be honestly ranked among the App Store’s two-tier division of apps into “Paid” and “Free” charts. An app that is $50 and sells very well will never make its way to the top of the “Paid” charts, and if it is lucky enough to beat out actually free apps in the free charts, it will only confound users who are surprised to learn that one of the top free apps actually costs money. The presence of a “Top Grossing” category provided a sort of compromise category for such apps, but Apple removed the ranking from iOS 11, and appears to be set to remove it from the Mac App Store in macOS Mojave.
  • Transaction mechanics are pushed onto developers. One of the primary advantages of the App Store to developers is being able to get out of the business of managing direct sale transactions. With the paid-up-front approach, users browse the store, conduct a transaction with Apple, and download the app. In exchange for taking on this work, Apple is rewarded with a 30% cut. With ersatz free trials, almost every aspect of this complexity is pushed into the app, where developers have to laboriously devise a mechanism for conveying app limitations to users, blocking pertinent functionality, transacting an in-app purchase, facilitating the unlock of app functionality, and so on.
  • Free trials cannot be easily reset. It is typical outside of App Store marketplaces for developers who offer free trials to periodically reset free trials so that users who, for example, enjoyed a free trial on version 1.0 of an app, can give it a fresh look on 1.1. The use of in-app purchases for accommodating free trials would, strictly speaking, require that developers perpetually add new SKUs to the App Store representing a different “free trial” product for each of the timeframes in which a developer wants to reset things.
  • Apps cannot be made to “just work” out of the box. One of the main rationales for offering free trials is to get prospective customers to download and start using the great features of an app as quickly and with as little effort as possible. With ersatz free trials a customer must first authorize Apple to allow the download of the free app, and then they must commence a confusing in-app purchase process during which they will be asked again whether they want to start a free trial.
  • Real Free Trials

    Now that I’ve listed a number of significant problems with ersatz free trials, let’s talk a little bit about what real support in the App Store might look like, and how it would alleviate the problems I’ve described.

    For starters, real free trials would allow developers who currently list their apps as “free” in the App Store to list them by their actual price. The App Store could convey that information both more honestly and more informatively to users. Instead of “Free with in-app purchases,” MarsEdit could be identified succinctly as “$49.95 with 14-day free trial.” These apps would no longer be erroneously featured among free apps, but would rank alongside other paid apps, where they belong.

    Having a bona fide price associated with the main App Store SKU would re-open access to the bulk purchase programs and family sharing. You know you want 500 copies of MarsEdit for your company? Go ahead and purchase 500 copies. The fact that the App Store happens to support free trials would be irrelevant to your conducting this transaction with Apple.

    Real free trials would open the functionality up to any developer who chooses to participate, regardless of their app’s functionality. Instead of forcing developers to come up with arbitrary lock-downs on functionality in the app, they would simply flip a switch in App Store Connect, ideally specifying a trial duration. When free trials are downloaded from the store, the receipt would have the trial information baked right in.

    Putting the logic in the store itself would also empower developers to start or stop offering free trials whenever they like, and to reset free trials across the board with major updates, in the same way they can choose to reset star ratings today. And all the tedious mechanics of offering, transacting, and enforcing free trial limitations would obviously be back in Apple’s court, where they can efficiently support such functionality in one place instead of requiring every developer to re-implement the same kind of support in every app.

    Finally, and probably equally importantly to users and developers alike, real free trials would enable users to effortlessly download and use all the features of an app without having to labor through any of the administrative tedium that is currently required by ersatz free trials. Happy customers trying excellent apps and ultimately paying for them is something that we can all get excited about.

    I hope this article has been helpful in illustrating why Apple’s review policy announcements, while very welcome indeed, do not constitute a major shift in their support for free trials in the App Store, and do not substantially change the status quo. Many of us are stretching the limits of the App Store to provide something that comes close to real free trials, but we would all be far better off if Apple announced a substantial change in supporting them. That didn’t happen this week.

Late to the Mac

A few weeks ago I was asked by Stephen Hackett to contribute a story, to his members-only newsletter, about my early experiences with computers and/or with the Mac. I was happy to help Stephen out because he does great work at 512 Pixels, because he is a friend, and because I like to talk about myself.

I agreed with Stephen that my contribution would be exclusive to his newsletter for a few weeks, and that afterwards I could publish it on my own blog. Here it comes!

Stephen’s version in the newsletter was spruced up with cool pictures of the mentioned computers, and possibly edited in ways that made me sound better than I do here. This is the raw, uncut version. Enjoy!

I’ve been a rabid Mac user and developer for 25 years, so it’s fair to say I’ve been a part of this community “for a long time.” But in my heart, I still consider myself something of a newbie. After all, I didn’t get my first Mac until I was 18 years old. My entire childhood was spent using, enjoying, even obsessing over computers made by other companies.

In 1982, I was 7 years old and lived with my mom in the rural town of Dunsmuir, California. While Silicon Valley’s famous fixation with technology was well underway, it hadn’t substantially reached the folks in my small, relatively poor home town. Meanwhile, my dad was living separately from us in Santa Cruz. He had gone back to school as an older student to obtain a computer science degree, and started his professional career working for IBM in San Jose. His enthusiasm for computers would turn out to be crucial to my interest in computers, and to my eventually programming them.

On one of his visits, he brought my first computer: a Timex Sinclair 1000. It was barely the size of a small book, used a television set for a display, saved files to audio cassette tapes, and featured a whopping 2KB of RAM. It’s fascinating that even the content of this article I’m writing would not fit in the memory banks of that computer. My dad gave me some copies of Compute! magazine, which featured BASIC program listings, some of which I guess were applicable to the Timex Sinclair. One of my earliest programming memories is of him walking me through the rough notion of source code as a sequence of instructions, and the computer as the obedient executor of those commands.

That Timex Sinclair wasn’t much of a machine, but it may still have made me the very first person in Dunsmuir to own a “home computer.” Apart from typing in the occasional program from those magazines, I never really learned to program it. But having those computing concepts introduced when I was so young, and in such a fundamentally hands-on way, set me up for the deeper understanding of computers and programming that I would later gain.

In 1983, my mom and I moved to Santa Cruz to live with my dad. In the Bay Area, we were far from the only family with a home computer, but my dad’s enthusiasm certainly meant our home saw a greater variety of them than most! My dad bought me a Commodore 64, which I adored, but the real thrill was his Kaypro IV. This beast was a semi-portable, all-in-one computer that ran the CP/M operating system, and featured a monochrome green display. I’d sneak onto it while he was out of the house, scouring his piles of floppy disks for anything of interest. I spent hours playing Ladder, a Donkey Kong-like game whose graphics consisted entirely of ASCII characters animated on the screen.

When the Amiga 1000 was released in 1985, my dad bought one, and I threw myself completely into it. I experimented with programming, studying the “ROM Kernel Manual”, but never learned how to put together a whole program. Still, for many years, I was “an Amiga guy.” I met other Amiga users, traded software, and reveled in the various technically superior capabilities of that platform.

Through the Santa Cruz BBS community, I got hooked up with an extensive network of UC Santa Cruz computer nerds, and found my way onto the internet by around 1989. That’s how I fell in love with UNIX. Throughout my teenage years, I played with various UNIX systems through the Adm–3A terminals in the university’s computer labs, and eventually ended up owning both a Sun 3/50, and an SGI Indigo. Obviously, I loved computers, but in spite of my diverse experience with various systems, I’d never really learned to program, and I’d never really used a Mac.

By 1993, I had a few friends through the Santa Cruz computer scene who were then working at Apple, about a 45-minute drive away. David Van Brink was a software engineer on the QuickTime team, and he opened my eyes to the virtues of the Mac. As much as I had enjoyed other computers, from that Timex Sinclair, through the Amigas, UNIX systems, and occasional other platforms, I had never witnessed a computer seeming particularly empathetic to its users. The Amigas came closest, but even at their most charming, they were obviously meant for nerds. The Mac, I finally came to realize, truly was “the computer for the rest of us,” and I was ready to become one of us.

David shared his employee discount with me and, after convincing my dad it would be useful for my college work, I came up with the cash for my very first Mac: a PowerBook Duo 210. This curious machine came from an era when Apple supported the notion of a “dockable” computer that could be used as a portable, but also plugged in to a desktop case connected to a full monitor, keyboard, etc. I could only afford the computer itself, so I made do with its limitations, but it was marvelous. Finally, more than 10 years after first being introduced to programming and computers, I owned a Mac, and was eager to write software for it.

I used a program called THINK Reference to study the Macintosh Toolbox API, and got to work writing my first app, Super Robots, which was a blatant clone of a text-graphics based app for UNIX called “robots.” (Fun fact: because the PowerBook Duo had a grayscale screen, I developed and shipped the app without ever having seen it for myself in full color.)

After shipping an app, and becoming vaguely familiar with the Mac operating system, I took the step that would ultimately be most impactful on my career and life-long love of computers: I went to work for Apple! At the encouragement of Qarin Van Brink (David’s sister), I applied at a Silicon Valley contract agency that was known to supply many of Apple’s QA testers. A few improbably successful interviews later, I was working in Cupertino on the System 7 integration team.

I eventually made my way out of testing and into engineering, where at long last, I started to sort of maybe learn to program. I also started learning how to really use and appreciate the Mac, and how to empathize with users. Compared to many Apple fans, I was late to the Mac party, but I’ve been making up for lost time. For 25 years, I devoted myself to using and programming Macs. Here’s hoping I can eke out at least another 25!

Change Infinite Loop’s Name to Apple Playground

When Apple’s flagship “Infinite Loop” headquarters was built in 1993, the name implied many promises of eternal iteration. Regrettably, none of these promises are likely to be fulfilled.

While it’s impossible to predict whether _some_ characteristic of the landmark might one day meet the standard, none of the most obvious candidates have passed the test:

  1. The buildings form a connected loop, so you can technically walk from one building, to the next, to the next, etc. This gives a feeling of an infinite sequence, but it’s apparent to anybody that there are in fact only 6 buildings, labeled IL1 through IL6. Add in IL7 if the Peppermill (now BJ’s) is accounted for, and you get seven. Seven is nowhere close to infinite.
  2. Given the size of the building compared to Apple’s relatively modest Silicon Valley workforce, it seemed at one time that it might house infinite employees. As time wore on and Apple’s successes grew, the number of employees who could be packed into the buildings’ narrow confines was shown to be … decidedly finite.
  3. Lifers at Apple might have once expected to work a virtually infinite number of days (and nights) on the campus, but as time wore on it became evident that these people either retire or move on to other companies. Finally, the remaining hope for infinite workdays was dashed by the construction and opening of Apple Park, where many current Infinite Loop employees will now work.

In fact Apple Park exceeds, by every reasonable measurement, the “infinite” aspirations of Infinite Loop. The extent to which one can walk around it infinitely is grander. The number of employees it can shelter, while shying considerably from infinity, is nonetheless greater. And the career longevity of folks who call Apple Park home today is, I’ll concede, about the same as it was at Infinite Loop.

The pragmatism in naming Apple Park is evident. In corporate headquarters, nothing is infinite. Not even for Apple. In naming this major headquarters upgrade, it makes no such allusions.

So what’s the perfect name for a smaller hoop of a campus, residing a stone’s throw away from the mighty Apple Park? Apple Playground, of course. You’re welcome.

(Radar #38078647)

The Nosiest Assistant

I bought a HomePod, and have been using it lightly for about a week now. On the face of it, I’m the ideal customer: an Apple fan who’s invested in other Apple devices, subscribes to Apple Music, is more-or-less at peace with Siri, and is primarily in the market for a better sounding home audio speaker. Hello, HomePod! I couldn’t resist.

I unboxed it and installed it in our dining room. I tested it with some of my favorite music, by Neutral Milk Hotel. It sounded fantastic to my ears, and seemed to live up to all the hype. My decision to put it in a common room was exciting for my children, who are six and nine. The six-year-old danced around, piping semi-naughty utterances such as “Hey Siri, play inappropriate music,” before giggling maniacally and looking to me for a reaction.

Next, we enjoyed about eight repetitions from the theme song for Milo Murphy’s Law, a children’s show starring Weird Al Yankovic. The kids fought over which one got to “ask Siri” to play it again next. For a short while, it looked as though HomePod would be a welcome addition in our home, but it soon wore out its welcome.

I expected shortcomings, as anybody who’s been following news about the device should, but none of the expected problems were deal-breakers. Here’s how many of those issues affect myself and my family:

  • Multiuser Support. It doesn’t support multiple users. It does support “personal requests” for exactly one user, so anybody who lives with family or roommates, and doesn’t want everybody to have access to their reminders, notes, etc., is compelled to turn off support for these requests.. Fortunately, Apple provides a setting for this and even asks during setup what the default should be. This is not a big deal to me, because I am almost always with my iPhone and Apple Watch, to which I make most of my “personal requests.” More on that later.
  • Multi-HomePod Support. It doesn’t support multiple HomePods. Yet. Apple promised technologically stunning audio performance with the HomePod, and they seem to have delivered. It’s supposed to be even better if you can afford to purchase two or more HomePods and let them work in concert (har, har) with one another. Me, I’m just looking for one HomePod to sound great in one room, so I’m not too bothered.
  • Direct Audio Input. HomePod can’t stream Bluetooth, you can’t plug an iPhone or any other USB device into it, and you certainly can’t pipe direct audio from RCA jacks or a headphone-style 3.5mm stereo jack. It can play music from your iTunes Music Library and Apple Music, or it can stream from an AirPlay-compatible device. This takes care of most my needs, but doesn’t cover all my wife’s wishes. We concluded we would need to keep our existing Bluetooth/USB playback device around as a backup plan.
  • Siri is Siri. Let’s face it: much of the criticism is warranted. I rely heavily upon Siri, and on the whole I’m pretty happy with it. Which is not to say I am not frequently frustrated by its mistaken interpretations and unfathomable inabilities. Perhaps the most embarrassing incompetency, particularly for a device that should feel at home in a kitchen, is its continuing inability to manage more than one timer concurrently. Chalk it up to Stockholm Syndrome, but I’ve got all manner of mental workarounds for Siri’s quirks. I get by, and I didn’t expect HomePod to make any of these challenges disappear.

What I didn’t expect was how incredibly sensitive Siri on the HomePod would be, and how obnoxiously persistent it would be about responding to my requests, no matter how far I am from the device, nor how unsuited it is for the task at hand.

My first experience with Siri’s obtrusiveness came while I was listening to loud music in my living room, being played by the HomePod in the dining room next door. I sat on the couch with my laptop, and paused to give myself a reminder about something. Let’s say it was “Hey Siri, remind me in 30 minutes to start making dinner.” I raised my Watch, issued the command, and was startled to hear the music suddenly drop to near silence. What’s going on? A moment later, Siri boomed from the next room:

You can turn personal requests on or off in the Home app associated with this HomePod.

Holy what! That was unexpected. An incredible demonstration of the perceptiveness of the HomePod, from a distance and over loud music, and a demonstration of the fundamental problem that is slowly driving me to disdain this thing: it will not stay out of my business! Any attempt to “Hey Siri” another device is met by a loud interruption by Siri either of the music, or of the silence of the room. It’s bad enough that it assumes all requests are being made to it, but it’s even worse that it insists on chiming in even when it isn’t capable of serving the request. Just to remind everybody that it’s not configured for personal requests. After several more unexpected activations, I conceded that things were not going well in the common room. I moved HomePod up to my office.

In my office, the HomePod is less useful as a music playback device, because I’m usually deep in thought on some programming or writing task, and I don’t do well with music playing in that context. However, at least in my office I could turn on “personal requests” and accept that HomePod might be my occasional fine-listening device, as well as a ubiquitous Siri assistant. At least it was worth a shot!

At this point I should mention that while my home office is primarily where I work on my independent software business, that I am also a moderately stressed out person who is trying to get a handle on reining in my busy, anxious-feeling brain. To that end, I commit some time in a typical workday to a very casual attempt at mindfulness meditation. I sit down on the floor, take a few deep breaths, close my eyes, and ask my Watch to “set a timer for six minutes” so I can unlock some personal decompression before getting on with my day.

You can guess where this is going: the first time I attempted to meditate in my office after moving the HomePod in, I had just settled down to lap up the water from the trough of relaxation when HomePod blares out at a loud volume: “Counting down from six minutes.” I value the timer on the Watch in part because it rouses me from my activity at first with a slight tapping and vibration on my wrist, rather than with a loud alarm sound. But as I had already settled down into my relaxation pose, I decided to soldier on, bracing myself for the room-shattering awakening that was soon to come.

At this point I’m on the verge of disabling “Hey Siri” altogether. Which is unfortunate because, to my mind, it’s still one of the most impressive features of the device. It’s just incredible to be able to utter music commands like “set volume to 30%”, “pause”, “continue”, “skip track”, etc., even at a whisper from across the room. If I disable this feature, I’m not sure I’ll find the merits of the HomePod as an audio speaker compelling enough to weave into my regular music listening habits.

I have filed a bug with Apple (Radar #37572955) requesting that it be less obtrusively interrupting, particularly when it can’t do what I’ve asked. Although I have high hopes for improvements to come with future software updates, you never can be sure when Apple will get around to fixing the issues that most vex you. I’m conflicted to be at once so impressed with the audio, form factor, and general behavior of Siri on the HomePod, while on the other hand having trouble finding a place in my house where it can show off all those features without also driving me and my family to frustration.

Update: After publishing the above, I received a note on Twitter from Connor asking what version of watchOS I am running:

I realized I didn’t know, off the top of my head. I figured I was updating my watchOS regularly, but couldn’t recall when I’d last done so. I reported that I was running watchOS 4.1, and he suggested that I might see an improvement after updating to 4.2.2. Yikes! I’d really fallen behind.

After updating to 4.2.2 I went back to my office and ran a few artificial tests. I immediately noticed the difference, as my Watch responded to “Hey Siri” requests, and the HomePod remained silent. I was so surprised that I wanted to confirm the HomePod was still listening. I lowered my Watch and issued a request, which was answered by HomePod. Moments later, I raised the Watch again and tried to address it, and this time the HomePod butted in again.

I think they’ve made dramatic improvements specifically to watchOS between 4.1 and 4.2.2, but there is still room for improvement. I had run into the problems with HomePod with my iPhone as well, which I confirmed is running the latest version of iOS. Still, the dramatic improvement with the Watch gives me hope that things will continue to improve, and the situation is not quite so dire as I had first experienced it.

Unified Swift Playgrounds

Usually I write about “programmer stuff” on Indie Stack, and “other tech stuff” here.

Today I had it in mind to write about my belief that Apple should get rid of Xcode Playgrounds and instead focus on adapting the iOS Swift Playgrounds app to the Mac.

I thought at first I would post it here because it’s not strictly about programming, but I ended up posting it there instead:

Unified Swift Playgrounds

May be of some interest even to folks not strictly interested in programming!

Gus Mueller on Extra Intuition

Manton and I just published the second episode of our members-only Extra Intuition podcast: I Know it was 15 Years Ago. We’re joined by Gus Mueller of Flying Meat to chat about … whatever comes up! Many thanks to Gus for taking the time to do the show.

It’s fun to have a chance to deviate from the usual format of Core Intuition. In some ways it’s more relaxed, like the show was in the early days. We are looking forward to mixing it up with interviews and other conversations that we don’t think are as suitable to the main show.

Our first episode, We Did Meet in Person features our recollections of meeting each other and starting to podcast together.

Core Intuition Membership

Today Manton and I released the 300th episode of Core Intuition. We published the first episode on May 30, 2008, and every episode since has been completely free for our listeners. Starting with Episode 301, that’s … going to stay completely the same. Except…

Core Intuition is now offering a membership program. We have been lucky over the past several years to have the financial support of many great sponsors, but we also want our enthusiastic listeners to have the option of supporting us directly. In the long term, we don’t know if we can count on our sponsorship luck to continue indefinitely, and would like to be able to continue doing the show regardless of how those fortunes shift.

We think that many of our listeners would support us without an incentive, but what’s the fun in that? That’s why we decided to start a second podcast, exclusively for members. Extra Intuition will feature extra discussions, interviews, and frankly, we’re not sure what. We’re just excited to have an outlet for some of the stuff we want to talk about, but doesn’t exactly fit the format of the main show.

Our first episode of Extra Intuition is already live, and it features a discussion about the early days of our friendship, and how we decided to start Core Intuition. If that sounds intriguing, please consider becoming a member so you can check out the show!

Sandbox Transparency

Apple’s sandboxing technology provides a mechanism for developers to specify “entitlements” that an app needs in order to provide functionality that users want. For example, on the Mac, an app can specify the entitlements to “print” and to “make network requests.” This system of granular privilege designation is a great baseline both for developers, to avoid accidentally overstepping intended bounds, and for users, to protect against apps intentionally or accidentally causing harm.

One of the biggest problems with Apple’s approach to sandboxing is that the accountability component has been left entirely to Apple itself. Developers are held accountable for the specific entitlements they request only when they distribute software through the iOS or Mac App Stores. In the review process, Apple may determine that a specific entitlement requested from an app is inappropriate for that app’s domain, and demand that the developer remove the entitlement before being approved. Or, in rare cases, they may approve an entitlement that other developers are not typically granted.

Yesterday, Gizmodo reported that Uber had been granted an entitlement for their iOS app that allowed them to capture an image of an iPhone’s screen at any time, even when the Uber app was not the active app on the phone. This is a big deal, because users don’t typically expect than an iPhone app that is not active might have the ability to eavesdrop on anything they are doing.

I have long felt that the sandboxing infrastructure on both iOS and Mac should be used to more accurately convey to users specifically what the apps they install are capable of doing. Currently the sandboxing system is used primarily to identify to Apple what a specific app’s privileges are. The requested entitlements are used to inform Apple’s decision to approve or reject an app, but the specific list of entitlements is not easily available to users, whose security is actually on the line.

I think the next step for sandboxing, on both iOS and the Mac, is to expose the list of entitlements that apps possess, in a way that is reasonably understandable to all users, and even more open to scrutiny by power users. Any user who is wary of an app should be able to examine its entitlements so that any unusual privileges can be evaluated. With this level of transparency, you can bet that Uber’s ability to arbitrarily record the screen would have been revealed much earlier.

Being more transparent with entitlements would also pave the way for overcoming an unfortunate side-effect of sandboxing: the elimination of whole classes of power-user level apps. If users were empowered to know what the privileges of an app are, through a combination of user prompting and an interface for inspecting entitlements, then it would be reasonable to grant more indulgent entitlements to developers.

Mac apps such as TextExpander essentially became unqualified for the Mac App Store with the advent of sandboxing, because they require access to system services such as monitoring the user’s keyboard input, in order to provide valuable macro text substitution. If entitlements were transparent across the board, and users were consistently informed about the extent of an application’s capabilities, it would empower users to make more reasonable decisions about the software they run. It would empower them to allow apps like TextExpander that are currently disallowed by the App Store’s sandboxing policies, and to reject apps like Uber that may be unexpectedly allowed to capture footage of users’ activity even while running other apps.

The Watch is a Phone

Since Apple announced that the Apple Watch Series 3 would be offered with an LTE option, a significant amount of collective energy has been spent scrutinizing the various pricing plans offered by telecom companies around the world.

Because I live in the United States I will focus on the de facto standard $10/month pricing that AT&T and Verizon have both announced. Depending on who you ask, this charge is either completely reasonable, or a complete rip-off. It may be a little of each.

Arguments for the rip-off point out that most US plans come with a fixed amount of data, and the extra $10/month doesn’t buy you any extra data. The charge is merely for the privilege of connecting directly to the provider’s wireless network from a Watch.

Arguments for reasonableness concede that there is an infrastructural cost to supporting another whole device on the network. Even if the data and phone number are shared, a new standalone device exists in the world, and it demands to be catered to by the network’s services.

I am not excited about LTE on my Watch, but I decided to buy an LTE edition anyway. Because Verizon is offering three months of free service as part of a sale promotion, I also decided I will give ubiquitous Watch connectivity a shot. Who knows? Maybe I’ll love it.

One thing that caught my eye during the checkout process was the fact that my Apple Watch has already been assigned a phone number of its own. Even though it will be configured in Verizon’s system to share my phone’s phone number, their process for activating and supporting devices on their network apparently requires assigning them a phone number. I don’t know if this is true of all carriers worldwide, or if this is a peculiarity of American systems, or of Verizon in particular.

In any case, I thought it was interesting that my Watch will have its own phone number, even if it goes unused. Longer term, I could see merit in configuring Watches to be the only device for a phone number. For example, a work phone at a company that communicates primarily by voice might be satisfied to equip their employees with a Watch and a pair of headphones.

I also think the fact my Watch will have its own phone number increases the justification for charging a nominal monthly fee. For as long as any Apple Watch is on Verizon’s network, a full-fledged US phone number will be “off the market,” so to speak. In most respects that matter to Verizon, the Apple Watch is not an add-on accessory to the iPhone that happens to want independent access to the network. The Watch is a phone.

Apple Watch Series 3

When Apple introduced the Apple Watch three years ago, orders were almost immediately backordered. Apple made a special offer to developers, granting us the opportunity to order a 42mm Sport edition with a blue strap, with minimal delay. I jumped at the chance and have been wearing the same “Series 0” ever since.

I’ve come to appreciate Apple Watch for all the ways it extends my iPhone: less intrusive notifications, casual Siri access, and effortless Apple Pay. I also appreciate the fitness monitoring, and have switched to using it as my exclusive running watch, in spite of its lack of GPS. It’s a pretty darned good debut product, and it has served me well.

That said, I dropped and cracked the screen over a year ago. With repair costs approaching the cost of replacement, I have long thought that I would hold out for the next big update before buying a new one. As I watched Apple introduce the Apple Watch Series 3 during their event last week, I knew that the time had come to upgrade.

Selecting a replacement has been more difficult than I expected. I always slightly regretted not getting the smaller, 38mm version of the watch, but I think I will err on the side of fat-fingered tappability and stick with 42mm. A harder decision has been choosing between the LTE Cellular version of the Watch, and the slightly cheaper WiFi/GPS-only model.

I thought it would be easy, because I don’t care about LTE. Or at least, I don’t think I do. Since I got in the habit of running with a belt, I’m accustomed to having my iPhone with me all the time. This ever-so-slight encumbrance has brought with it two unexpected benefits: I feel safer, knowing that I can always (in most locales) make an emergency phone call, and I feel empowered to take photos mid-run when something particularly interesting catches my eye. This has come in handy when jogging in foreign countries, through nature, or when I want to capture a selfie while running in a New England blizzard.

The LTE Watch certainly would provide the same emergency calling comfort, but am I willing to give up access to a camera for the freedom from wearing a running belt? I suspect not.

Nonetheless, I’ve decided to go for the LTE edition. Why? For one thing, apart from the option to add a cellular plan, the LTE models also feature twice the storage. I tend to err on the side of extra capacity, so even if LTE were not an option, I would probably choose to pay $70 for this alone.

Sweetening the deal, Apple offers configurations with the LTE models that are not available with the GPS version. After reading Serenity Caldwell’s iMore article about the merits of the Sport Loop, I’m convinced I’ll want to give this a try. Choosing the LTE model means I can preconfigure the Watch with it, instead of paying extra for the Sport Loop, and getting another rubber Sport Band that I’ll never use. (I’ve been loyal to the Nylon Band since they were introduced.)

In short, even if I never use the LTE functionality, the LTE model is the right choice for me. Who knows, maybe I’ll end up trying the cellular access, and it will be a big hit. Maybe I’ll cast my running belt aside, and not worry about lost photo-taking opportunities. Maybe LTE on my wrist will be a revelation, and I’ll laugh at myself for ever having doubted its usefulness. Maybe I’ll come around.

The only downside is that gaudy red crown.

Update: If you’re in the US and a Verizon customer, they’re offering to waive the activation fee and give 3 months of free service, but only if you buy the Apple Watch from them. Their site is a mess but you can get at most of the models here. Good news is their ship date for the model I chose is a few weeks ahead of Apple’s!

Multilingual Completion

At times I’ve been frustrated that when I’m trying to type a Spanish word on my iPhone, I’m only offered completions in English. For example, if I wanted to type the Spanish word for ostrich, avestruz:


It finally occurred to me that I could add a Spanish keyboard to my Keyboard settings:

IMG 2574

Now, when I’m in an app or on a web site where I’m likely to type in Spanish, I can easily change the keyboard and get Spanish-language completions.

It gets better: if a second-language keyboard such as Spanish is merely enabled, it affects the keyboard completion behavior such that completions from either your first or second language is included. Here I am typing “aves” in the English keyboard, but completion is smart enough to infer that I’m looking for “avestruz”:

IMG 2575

I had hoped that iOS’s willingness to cross language barriers might also apply to Siri dictation, but alas no. “All these truths” is the closest it can muster when I speak “avestruz”:

IMG 2570

However, having the Spanish keyboard at hand means I can quickly switch keyboards and dictate as expected:

IMG 2572

If you find yourself frequently typing a foreign-language word and hoping for some help from your phone, be sure to install a keyboard for the language so that iOS knows you’re interested in its suggestions!