The Year That Kicked My Ass

A Christmas ornament from my sister-in-law that sums up my year.

January started as normal as can be expected when malicious grifters start making basic decency a radical idea. It turns out the anxiety associated with these political events would be the least of my problems throughout the year.

It felt great to finish up a 12 month project and release the first version of Tapestry. I celebrated with a trip to Louisiana visiting my wife’s birthplace, exploring islands and bayous, and eating more seafood than I thought possible.

An impressive spine.

In April, I turned 65 and signed up for Medicare. I was about to learn how important this was.

Towards the end of that month, I started feeling some tingling in my left index finger and some pain in my neck, especially after working at the computer all day. Initially, I chalked it up to the normal aches and pains of growing older, but the pain just wouldn’t go away.

The next month was marked by tragedy. On May 17th, while taking our dogs for a walk before dinner, our girl Jolie was attacked by dogs that had escaped from their yard. It took every ounce of my strength to get two 50 pound dogs without collars off of our 15 pound pup, but I rescued her, did some quick triage for her open wounds, and rushed her to the vet for four hours of surgery. We were both wrecks, but made it to see another day.

We’re both getting too old for this shit.

Jolie started to recover from her injuries, but she was a 15 year old with a weak heart. On June 4th, I found her unconscious outside the door of my office. She died peacefully and the loss was added to the year’s pain tally.

I also had adverse effects from the dog fight: the pain in my neck had gotten much worse. The adrenaline rush made me move my neck and arm in ways that turned an irritating pain into a persistent one.

In July, we travelled to San Diego to see an outdoor concert. I was living with neck pain all day, every day, and when I couldn’t lift my head to watch the show, I knew I needed help. On the 16th, I had my first appointment with a local chiropractor. X-rays showed degenerative spine disease, which is consistent for someone my age: pain being caused by old cervical vertebrae and pinched discs.

I was staying active in spite of the pain in my arm and neck. My swimming stroke sucked thanks to my limited arm movement and neck pain limited the length of my bike rides.

On August 3rd, while riding my e-bike to Trader Joe’s to do some grocery shopping, I was hit by a car. Someone blocking the road at a 90 degree angle decided to backup while only looking at the camera on their dashboard. They didn’t see me riding in the rightmost lane of traffic.

I ended hitting the D pillar of a SUV with my left shoulder and tearing my AC joint. Then I was thrown from my bike and landed hard on asphalt. The impact broke five ribs and I immediately had a new source of pain on my left side.

The paramedics arrived and got me to the closest emergency room. That’s when we all discovered I had another problem: a punctured lung that was causing my chest cavity to fill with air. This presented itself while lying down waiting for a CT: it’s impossible to express the panic of not being able to breathe or talk. Luckily, my wife was in the room and screamed for help that resulted in a temporary chest vent while I was rushed to a trauma center. Another ride with the paramedics, this time with lights, sirens, and lot more speed.

There was a team waiting for me, and I got a dose of ketamine, followed by a chest tube that was inserted while I was (barely) conscious. As the surgery was ending, the head nurse asked me how I was feeling, and my response was “I’M TRIPPING BALLS”, which got a laugh from everyone in the operating room. It also helped me understand a billionaire that needs the substance to feel joy in his life.

I spent a total of three days in the hospital as the doctors monitored my chest fluids. My main source of pain at that point was the broken ribs: sneezing, coughing, or laughing hurt like hell. What didn’t hurt was my neck and arm: as one nurse joked when I was telling them about my situation: “Hey, you got a free adjustment!”

The view from my hospital room and it’s missing a K.

I felt good enough to spend some time working on Tot 2: all of the App Store purchasing code was done while in a hospital bed. It was a nice distraction and helped us ship the update at the end of August.

Soon after the release I read a blog post that rang true: Irrational Dedication. Both of the Iconfactory’s major releases during the year were willed into existence. Tapestry after a year of work for a new product category (“timeline apps”) that was difficult to explain. Tot while working through various stages of pain.

It took about six weeks for my ribs to heal completely. While that was happening, September presented another health issue to deal with: this time for our boy dog, Pico. What started as a small bump on his butt quickly grew into a large Mastocytoma (Mast Cell Tumor). At the end of August he had surgery to remove the mass and he got a new nickname: “Zipper Butt”.

“Daddy, you should try this look.”
Click or tap to view.

We were about to put a twist on the old adage about dogs looking like their owners: this owner was about to look like his dog.

This was also the time where my original neck pain returned. It turns out the brain can’t handle more than one pain input at a time – the broken ribs put the nerve pain on the back burner. Chiropractic treatment was providing only temporary relief, so I tried acupuncture in October.

Then, in November, all hell broke loose. At the beginning of the month we took a car trip to Tucson for a family event. I spent most of the trip through the desert with shooting pains through my arms: agony for hours on end.

A week or so later, I started noticing problems with my ability to walk and a numbness throughout my torso. The nerve pain felt like the onset of paralysis. Shit was getting serious.

My primary care physician prescribed muscle relaxers which had no effect. My chiropractor scheduled an MRI on the 14th and we got the results on the 17th.

The MRI showed that I had a mass in my spine that was pressing on the fluid that protects and nourishes the spinal cord. My neck was screwed up more than anyone expected and needed immediate attention. A referral to oncology at Hoag Hospital got us into the ER on the 19th.

MRI showing the white spinal fluid around the dark spinal cord being invaded by a mass of gray tissue.
Click or tap to view.

There was just one problem: my goddaughter was getting married on the 19th. On a sandy beach, at the end of a rocky path. And I could barely walk.

I’ve been a part of her life since birth and not being able to share this important moment broke me completely. I spent most of the 18th sobbing and feeling shitty about the cards that life had dealt me.

The tests included a two hour full–body scan in a noisy and cramped MRI. Plenty of time to contemplate life and realize that the last time I had been at this hospital was when my goddaughter was born 36 years earlier: a day spent translating for two women who were about to be grandmothers for the first time and didn’t speak each other’s language. (Little known fact: I’m an Italian godfather.)

All the tests confirmed the spinal mass and provided a plan for treatment. I was given steroids to reduce inflammation and felt immediate relief: it was the first time I had been without neck pain in about eight months. Next, a cervical laminectomy would remove part of my spine and permanently relieve the pressure on the spinal cord that was the source of my pain. It would also allow the doctors to obtain a sample for pathology: to determine if the mass inside my spine was benign or malignant.

Twenty-nine staples later and my neck felt a lot better.
Click or tap to view.

The operation was a success and I was home in time for Thanksgiving. I was so thankful for my wife, family, friends, and medical professionals that were helping me through this rough time. And for the end of a week with opioid constipation.

After the holidays, it was not a shock to learn that the mass was malignant. Everything we had seen suggested that the source was lymphatic. Additional tests, including a PET scan and a lumbar puncture (a.k.a. spinal tap), made it clear that I have a follicular lymphoma in both my blood stream and spinal fluid.

The good news is that this is not a particularly aggressive variant and has therapies that have been effective for decades. It’s going to be something that takes months to treat and will require some hospitalization. But the doctors and I are both optimistic about the outcome.

The surgery to relieve neck pain continues to heal: I still have a bit of muscle soreness but the persistent pain is completely gone. Another reason to be hopeful for recovery.

I still have the nerve damage that caused my initial paralysis. The hope is that as the spinal mass shrinks, my walking and numbness will improve. And the only way to make that happen is with both physical therapy and chemotherapy, both of which I started on Christmas week. Happy holidays!

Luckily, I didn’t have any major issues during the first infusion, but a week later I’m still feeling the effects: overall fatigue, a queasy stomach, and a weird taste in my mouth. Dietary restrictions like giving up red meat, fried foods, and processed sugars seemed important a week ago. Now, the medicinal marijuana my nephew got me for Christmas feels much more significant.

It’s clear there is a long road ahead of me, and while I may have less spine, I am not spineless. The irrational dedication I mentioned earlier is now focused on getting myself back to health.

My personal goal is to swim to a buoy in the Pacific Ocean. It’s going to take a lot of effort to make that happen and I know that stating your objectives is the best way to meet them. (One of the reasons for this blog post, in fact.)

My goddaughters heard about my aspirations and handmade an inspirational gift for Christmas: candles of the buoy itself and the kelp and Garibaldi underneath. I’m going to burn it all down.

Burn, baby, burn.

I had originally wanted to end this essay on that positive note, but the year had other plans. The week after Christmas, Pico started showing signs of abdominal pain and inappetence. He had developed a mass on his liver and spleen, and given his age, the prognosis for recovery wasn’t good. I always knew that saying goodbye to my constant companion of the past 15 years was not going to be easy, but never imagined doing it with all this other shit going on in my life. Consider my ass well and truly kicked.

Even if I’m getting out of the year on emotional fumes, I lived to see another one. My little boy won’t be there to dance around excitedly as I get out of the water this summer, but he will always be a reminder that I never give up.

Tahoe Electron Detector

No, we’re not doing science at California’s most beautiful lake.

We’re looking for bugs.

A popular cross-platform app development framework called Electron is using private and undocumented API that’s causing system-wide slowdowns in macOS Tahoe.

We’re hearing from customers that some of our apps are running slowly on Tahoe and I suspect that this bug has something to do with it. Unfortunately, it’s hard for customers to check which version of Electron is being used and see if that might be a cause. So I decided to do something about that…

Luckily there’s a script written by Tomas Kafka that lets you check all your apps quickly and easily. I took that script, updated some parts that required Xcode to be installed, and wrapped it up in an AppleScript applet that’s easy to download and run:

Download TahoeElectronDetector.zip

When you run the app, you’ll see a short introduction:

Introduction of TahoeElectronDetector check explaining what will happen.

The first time you run the app, you’ll see a warning that the app was prevented from modifying other apps on your system. This is “normal” because the app needs to read other apps to do its job:

After all apps are checked, you’ll see the results:

Results of TahoeElectronDetector check with red X shown in front of apps that do not pass.

Eventually, you’ll see ✅ in that window and know that one or all of your Electron apps have been updated.

If you’re one of those people who’s wondering when it’s a good time to upgrade to Tahoe, you can run TahoeElectronDetector on older versions of macOS and give yourself an idea of when it’s safe to move to the new operating system.

Additionally, there’s a website that lists the status of the most popular apps. This will be helpful in locating newer versions since some of them will not update automatically.

If you’re a Mac developer who’s hearing from customers about weird slowness, feel free to point them at this web page or give them a copy of the app to check their own system. If you need the source code, it can be downloaded here.

And if you’re a developer, this is your periodic reminder not to use private and undocumented parts of an API. They will break, and in cases like this, it will be spectacular.

One Size Does Not Fit All

In a previous essay, I briefly expressed some thoughts about why Liquid Glass is inappropriate for the Mac:

I’m having a much harder time seeing how Liquid Glass will benefit other platforms like the Mac or Apple TV (where Apple doesn’t even make the screen). Forcing tactility where it’s not needed or wanted feels like a misstep.

I’ll now go into depth regarding these thoughts.

In 2010, John Gruber wrote The Future of the Mac in an iOS World for the Macworld back page. He explained why the Mac was still so important in the new world dominated by the iPhone:

It’s the heaviness of the Mac that allows iOS to remain light.

When I say that iOS has no baggage, that’s not because there is no baggage. It’s because the Mac is there to carry it. Long term — say, ten years out — well, all good things must come to an end. But in the short term, Mac OS X has an essential role in an iOS world: serving as the platform for complex, resource-intensive tasks.

It’s been fifteen years since he wrote that — have we reached the point where the Mac can come to an end?

I’d say not.

Apps like Final Cut Pro on iPad are an impressive achievement, but they lack the features, file management, and expansive screen real estate of the Mac. It’s a great tool for casual editing on the go, but I can’t imagine Apple saying that the Mac app is end-of-life without there being a huge uproar.

Another example is Xcode: even though the hardware on iPad and Mac shares the same processor, there is still no port of the user interface. And even if you can whittle the complexity of an IDE down to fit on a smaller screen, you’d still have problems with locking down the app. On the Mac, Xcode doesn’t run in an app sandbox and does not use the hardening that prevents certain security exploits. iOS mandates the use of both these things: it’s not even an option for app developers.

Xcode also sports 54 entitlements that let it do things other apps can’t. Things like authorization, device pairing, and inspecting the memory of other processes. A port of Xcode to iPad would immediately make it an attack surface for hackers.

So while iPadOS has obviously gotten more capable, I don’t see it displacing the Mac desktop as the place where the heavy lifting gets done.

Now let’s talk about that heavy lifting.

It’s done by professionals who have highly tuned desktop spaces and workflows. Windows and controls are in just the right place so focus can be wholly devoted to the task at hand. Everyone’s workspace is different and as unique as the person who created it.

If you’re someone who’s only using email, a web browser, and some messaging apps to get stuff done, changes to your desktop appearance aren’t going to be disruptive. It’s also likely that you’ll appreciate changes that make it look like your phone.

If you’re doing anything more complex than that, your response to change will be much different.

A photo of a female truck driver named Sharon Kimbrough at the wheel of a semi-truck.
Have you ever wondered what the inside of a semi-truck looks like? Watch this short tour!

Professionals on the Mac are like truck drivers. Drivers have a cockpit filled with specialized dials, knobs, switches, microwave ovens, refrigerators, and pillows that are absolutely necessary for hauling goods across country. Those of us who are making movies, producing hit songs, building apps, or doing scientific research have our own highly specialized cockpits.

And along comes Alan Dye with his standard cockpit, that is beautiful to look at and fun to use on curvy roads. But also completely wrong for the jobs we’re doing. There’s no air ride seat, microwave oven, or air brake release. His response will be to hide these things that we use all the time behind a hidden menu.

A photo of the cockpit of a 2025 Porsche 911.
A beautiful Porsche 911 cockpit that’s perfect for high performance driving. But be sure to compare the angle of the steering wheel with the previous photo.

It’s no wonder our reaction is somewhere along the lines of “fuck off”. Or maybe something a little more polite and eloquent. The bottom line is that one size does not fit all: we don’t want a Mac that looks or works like a phone, tablet, watch, or TV.

Worse, this situation is going to be like notifications on the Mac: a minimal design that mimics other platforms, and completely annoying in day-to-day use.

Liquid Glass is currently in a barely presentable state on iOS. It’s going to be like iOS 7 and take another year to sand down the rough edges. And then several more years to tone down the design, as with Aqua.

With the Mac typically lagging other platforms, I don’t expect to see any design improvements on my desktop for several years. It’s going to be an unpleasant and lengthy slog with various accessibility workarounds in place until the standard design looks decent. Or maybe, like with notifications, that will never happen because Alan Dye knows best.

My Mac has been a truck since the beginning of this century. When my desktop computer got Unix and Aqua it was the perfect platform to craft my cockpit. It’s going to be really hard to abandon that and create a new one, but the way things are heading, it feels likely.

FB19914338

I submitted the following feedback today. If you ever plan to change your business model from a paid up-front to freemium model, read this report and avoid a day of headache and stress.

Title: The sample code for a business model change was written by someone who’s never submitted an app to the App Store

Please describe the issue and what steps we can take to reproduce it:

The source code example using in Supporting business model changes by using the app transaction does not work if you’re using current Xcode and App Store conventions. Additionally, the sandbox environment uses the same outdated conventions.

And when you use that sample code, that you cannot test in the Xcode transaction simulator or in the TestFlight sandbox environment, it will fail spectacularly on launch day. You will be inundated with support requests from people who are expecting to see a payment for the previous version AND you’ll be in a state of panic because YOU HAVE NO IDEA WHAT THE HELL IS GOING ON. And did I mention that you can’t test this in production?

The sample code implies that the originalAppVersion is a string that’s separated by periods (“.”). The sandbox environment returns a value of “1.0” which reinforces this notion that it’s a value that separated by periods.

It is not.

If you’d read the Xcode documentation, you’d know that it automatically generates an app’s Info.plist. This has been the default setting for quite while – most developers have no idea this is a configurable option: they fill in the “Version” and “Build” number in the target’s General settings and are done with it.

For most, the build number will just be a single number that increments each time you submit to TestFlight (and eventually to the App Store).

When GENERATE_INFOPLIST_FILE is enabled, it sets the value of the CFBundleVersion key in the Info.plist file to the value of the build number (CURRENT_PROJECT_VERSION, or the “Build” in General settings). And that means your Info.plist is getting a CFBundleVersion without periods.

So what happens when you use this code?

let versionComponents = appTransaction.originalAppVersion.split(separator: ".")
let originalMajorVersion = versionComponents[0]

Well, if you’re an inexperienced Swift developer, your app is going to crash with an array index that’s out of bounds. Those of us who are more careful in our receipt processing code will skip over the originalMajorVersion because versionComponents is empty.

And that’s when the emails from customers start arriving.

Luckily, there is this nugget of information describing originalAppVersion:

The originalAppVersion remains constant and doesn’t change when the customer upgrades the app. The string value contains the original value of the CFBundleShortVersionString for apps running in macOS, and the original value of the CFBundleVersion for apps running on all other platforms.

So even though CFBundleVersion was originally intended as a major/minor/patch format, its current use is as a single integer that increments when you submit to TestFlight. So the code above is expecting “1.0” and is actually getting “83”.

(And why the hell is it different on macOS? You do realize that cross platform apps are a thing, right?)

Again, you have no way to test this theory other than going through App Review (with an expedited review if you’re lucky). And if you’re even luckier, you’ll have folks on Mastodon that will confirm that this sample code is a piece of shit. A few hours later you’ll breathe a sigh of relief when folks start telling you that things are working fine.

And then the next day, you’ll write this bug report and post it publicly because no one else should have endure the stress caused by this sloppy code.

UITabAccessory Backward Compatibility

The addition of UITabAccessory in iOS 26 is welcome. It does, however, create a problem as far as backward compatibility is concerned. How do you present the new accessory view on older versions of iOS?

This backward compatibility is especially important for Triode. A lot of folks turn an older device into a dedicated radio player. I have an old iPad in the kitchen, for example.

So what do you do on the other side of the availability check where you set UITabBarController.bottomAccessory?

You’ll need to create two subclasses: one for UITabBarController and another as base class for all the view controllers you add as tabs (mine is creatively named as TabViewController).

In the UITabBarController subclass, you’ll do the check for availability in viewDidLoad and for versions older than iOS 26, you just add the accessory view to the tab controller’s view hierarchy using view.addSubview(accessoryView). You’ll also style the accessoryView as needed (e.g. adding a backgroundColor and cornerRadius). The same gesture recognizer is the attached to the accessoryView regardless of how you add it to the tab bar.

Then, in viewDidLayoutSubviews, you use tabBar.frame to position the accessoryView relative to the tab controls.

The other piece of the puzzle is doing the automatic inset adjustments on the tab controller’s views. In your common subclass (e.g. TabViewController), you’ll implement viewDidLayoutSubviews. On iOS 18 and earlier, you can check if the view or its first subview is an instance of UIScrollView. If it is, set contentInsetAdjustmentBehavior to .always and make new UIEdgeInsets to match the metrics you used in your tab bar controller.

(Side note: if you are having problems with the bottomAccessory on iOS 26 not animating as you scroll, make sure that your UIScrollView is the first subview. If something like a search field is the first view, it won’t work correctly.)

On iOS 26, Triode’s tab bar looks like this:

And thanks to the work above, folks on older systems can use the same accessory view:

You don’t get the fancy animations and effects, but folks on older devices will appreciate having the same capabilities. And better text contrast ;-)