Apple Aperture: Senior QA (2004-2005)
techreflect.net
techreflect.net
There was a lot of functionality/crap in there that I did not use, a lot of functionality was hidden away behind context menus and a lot of promised features around integration with plugins and 3rd party never really materialized, but in terms of core function it was a total workhorse.
Apple ultimately went to great lengths to purge it from existence, even for those of us with paid licenses who'd been told (or at least safely presumed) that we'd be able to install it on new machines via the app store, even if it wasn't available for purchase. In reality, they removed the binary from the cache. At one point, I went through a ridiculously elaborate process of obtaining a certain kind of physical license from eBay and spent hours emailing random Apple employees - likely some of the folks mentioned in this story - getting them to make exceptions and pull favours to make it possible for me to install it one last time before Catalina pulled the rug out for good.
Fuckin' Apple.
Which is all a shame because I loved photography for a couple decades, and now I have small kids and I’m sure I’ll regret it, but it’s just such a pain in the ass to do these days if I don’t want my entire life tethered to services and data in the fuckin cloud.
So much so, that I too eventually stopped with DSLR after losing it.
I actually like it so much, that I even use it since I switched to Linux recently. I have a Windows VM with one app installed that I boot up every month or so to process photos.
One of my favourite things about Aperture+PS was being able to spawn an edit process that, when you save the file in PS, created a duplicate version with the changes back in Aperture. The more I think about it, this was the secret sauce. Is this replicated in DxO?
As an aside, the author's "Technical Deep Dive" is worth a read: https://medium.com/@cormiertyshawn895/deep-dive-how-does-ret...
Stories like this is why I rarely bother with proprietary software anymore.
I remember using it when I got my first digital cameras :-)
Remember, everything Apple does, it does to advance the interests of Apple. The company. Not the products, not the departments, not the employees, and definitely not the customers. The same is true of any large org.
The knockout success of iPhone grew the business by magnitudes almost overnight. iPhone Apple, frankly, is too big to be wasting its time now on niche products with limited potential for growth, even if those products are very successful within their own niche. If a product doesn’t fit with the current business strategy and portfolio, it’s out.
It’s one of the basics to running a successful business: don’t be distracted. And definitely don’t be sentimental. Poor Aperture just got left in the iPhone’s dust.
I think the situation currently is an overcorrection to a vaguely remembered pre-Jobs past.
What does that mean in the context of a huge corporation? From my perspective, $1K in profit is still $1K, even if you're making $100K elsewhere.
Tim Cook certainly has limited time and needs to prioritize, but he also doesn’t need to touch every part of the company. Independently-profitable projects can be spun out into their own teams as needed.
Having said that I think Apple hasn't prioritise their Software Engineering resources well.
When Steve Jobs came to Apple, he made the Mac line “focused” by cutting it down to four models. That had clear benefits: less hardware for the OS to support, better economies of scale in manufacturing, and less wasted retail shelf space.
How does killing Aperture benefit Apple’s other products in that way?
(We don't have to like it, or work at or buy from corporations.)
$100k in profit is not $100k in profit to a company that makes $1B a day: it is a massive operational risk. Not only is the revenue a rounding error, but every product on the docket represents a meaningful sacrifice of agility and exposure to unknowable risk.
One of my smartest mentors once told me (and then reminded me a dozen times) that a company can only have three priorities at any given time. If an initiative does not make the top three cut, it is tabled for the future. Your executive team - and indeed, the entire org - should have those priorities front and center as they make decisions and spend their time. "Does what I'm about to do advance one of the top three priorities?" is a clarifying question.
Every person in an org can have three personal priorities, no more. Every person should be able to confidently tell you what those priorities are. Those priorities should all meaningfully tie back to the company priorities.
My mentor was, some 20 years ago, the in-house corporate counsel for a large telco. Amongst his responsibilities was M&A and foreign expansion. You'd think these would be priorities, but you'd think wrong. His initiatives were consistently tabled, politely but assertively. It was the best thing for the company.
On the flipside, companies and leaders than cannot embrace laser focus on top priorities tend to be slowly dying, even if they don't realize it yet.
In the end, $1B a day is $1B a day to lose, if they don't focus on their top priorities. Frankly, I think it's some combination of branding and charity that they haven't just stopped making computers.
Most geeks know the value of everything and the cost of nowt. Especially those costs measured in risk, be it risk of a code exploit in some mature stable part of the OS that almost nobody uses, or the risk of ceding a trillion-dollar primary market to competitors while preoccupied chasing secondary tails.
Having blown my first startup by sloppy unfocused mismanagement, I can attest to this lesson the hard way.
(And yeah, the only reason Macs still exist is because Apple has to protect the whole productivity ecosystem that puts Apps in its iOS Store. ’Cos the moment it cedes that foundation to a competitor, the whole company is lost.)
That isn’t a real deadline, it’s a management ambition. And a badly set ambition, based on some demos and the belief they could scream good code into existence.
Apple is getting better at this - see how it now pulls announced features from releases if they aren’t ready - but it’s still a terrible approach.
To make parents work weekends without their children for an entire 6-month period just to try and hit a management fantasy is a symptom of a toxic culture.
(And is also counter-productive. The slow and buggy v1 that resulted wasn’t good enough to capture the market — but it was enough to finally wake up Adobe, who dramatically accelerated efforts on Lightroom, which launched first on Mac and went on to dominate the market)
When Apple announced Aperture, it instantly galvanized the executive support in favor of Lightroom and Shantanu made the call for us to announce Lightroom and ship a public beta in 6 weeks.
Thankfully we ran the engineering team with an iterative and low technical debt threshold. We had been shipping iterations to a private beta group.
The Lightroom public beta beat Aperture to the market, ran on PowerPC and Intel (Aperture was only on Intel Macs), and was much much faster.
Do you have any idea what happened? Aperture 1 was slow but it got better and since it was expensive to switch I never tried Lightroom until Apple cancelled Aperture. I was surprised to find that it was so much slower and eventually ended up getting a refund because even the Adobe support person eventually admitted that it wasn’t fair to suggest buying a new Mac just to be able to use the application.
I guess from a mentality perspective, sure. But the actual quality of their software has gone down so much over the past few years...
Plus now there's a fork of Lightroom "classic" and a weird cloud-only version, and also Adobe's subscription pricing. There are a few other apps that have added digital asset management, but overall, there's really nothing that has taken its place.
Edit: most recent I can find is that a year ago they weren’t working on it https://forum.affinity.serif.com/index.php?/topic/78025-when...
Is this the corporate version of PTSD? I'm sure the Maps debacle really had an affect on the overall decision process to the "release software prematurely, or delay it until it is working" question.
It's still mostly "we ship yearly, and everything has to be ready by September". That's how we got four or five iOS 13 updates within a month of release fixing glaring bugs and underdeveloped functionality.
People left rudderless but with strong and inflexible demands on their teams will often revert to a control freaky micromanagement style, out of pure risk abatement. Bad managers are products of bad cultures; there are (thankfully) very few actual psychopaths in the world.
I assume these managers got promoted after I left. It's hard to measure manager effectiveness, and with projects lasting multiple years, a couple of wins can make it look like you're an ace manager even though the team basically just ignored you.
Any bad employee will damage the business and those around them. But where a typical bottom-rung grunt has a very small scope of influence and is relatively easy to fire, even a mid-level manager has more than enough power to ruin entire departments; and once they’re c-suite the game is all but done.
A typical engineer will gladly stab another in the back for something so trivial as what nearly-identical-anyway programming language or framework or IDE to use. Meanwhile managers, like lawyers, accountants, doctors, teachers and every other profession, prioritise solidarity within the profession. We can debate whether that’s good or bad but there’s no doubt why they prosper while engineers are the first to be outsourced or exploited.
Have you worked for any length of time in those fields? Don’t know if you ever worked as a lawyer but “solidarity” isn’t a core value.
Solidarity to the profession. They may squabble amongst themselves but they’ll never compromise the general prestige of lawyering. Nor tell the client that what they do is easy and anyone can learn it, even if that’s true.
That didn't happen. And while the author doesn't say that Aperture 1.0 was bad, he doesn't appear to have thought too highly of it either. Separately, it's hard to imagine such a rushed and mismanaged process producing a stellar end-product (although, I'm sure stranger things have happened).
What was Aperture like at release? Were later versions a substantial improvement?
"So, we shipped the product and introduced it at the show as planned. Despite a long list of warts and terrible performance, I thought it turned out well given the schedule and the circumstances. I continued to use it as my primary photo app, even though I’m not a hard-core professional. Others on the team were too bitter, such that they refused to use the final product."
- cricket
Definitely a tool where you wanted the x.x.2 before you’d update a major or minor version, because, well, it probably held 150,000 photos at a time where almost nothing else you did on your personal computer had that amount of data or irreplaceable value per file.
--- Aperture science ---
Many prolific shooters considered this the best photojournalism / travel / wedding / EDL bulk photo library and photo triage tool outside of the camera bag, bar none. Still is if you have kept a Mac tuned for it.
The key benefits for me, not found most other places at the time, were:
- auto photo sorting (“stacks” of burst photos you could expand and choose the best from then collapse again, like burst works on iphone now), along with press compliant metadata and tagging, GPS, TZ adjust, etc.
- ultra fast photo triage (rapid star ratings and ability to see photos above n stars for fast pruning of a 3000 photo shoot to the best 30 shots), and
- a (just okay) RAW engine with non-destructive photo editing, as well as masters and ‘versions’ per photo including automatic RAW+JPEG pairing for digital cameras
All this on a love-it-or-hate-it “managed” library, where you could go look at the originals in folders, but never needed to if you let go of folders and thought in metadata sets. (It could also oversee folders external to the library.)
I switched to it from Lightroom, occasionally tried to switch back, always felt like molasses compared to the ‘just right’ benefits of Aperture if you shot a lot of photos to get a small set of publishable shots, but also wanted to keep every shot forever.
--- Photos 2.0 ---
There’s been a dark few years here where the Photos library that was capable of Aperture sized content didn’t offer any of those key benefits in the Photos app, but at least remained accessible across all your devices. While Photos continued to do the RAW+JPEG pairing, it lost most other things.
Latest Photos finally puts back in some of the level of photo editing that was in the original Aperture.
That said, seems though when implemented through Photos on iCloud, there’s a soft performance limit in the 130K to 160K range that things start to act wonky, one photo out of 100K won’t sync, etc.
Be interesting to see what would happen if Apple threw some QA at the high volume uses cases again.
I was lucky in that during both of my tours, my management was humane, but that's sort of the luck of the draw.
I don't understand why people tolerate that kind of treatment. Never in a million years would I work nights or weekends or allow anyone to scream in my face. That's just not life.
My manager and some on my team at Apple generally discouraged me from working on weekends - I’ve worked some on weekends mainly to de-risk accomplishing certain tasks due to me being in an optimal state to tackle some issues & having some spare time. Occasionally I have to fix operations issues on weekends too, but those get resolved generally pretty quickly.
All in all it’s a YMMV situation. Most people in my org keep to normal hours, have same development timelines, and are pretty happy (no yelling/screaming, most operate somewhere in the 10-4 time range give or take, 5 days a week).
On the other hand, I've never built a product from scratch so quickly. Google and Facebook didn't exist when I started my current job. Now they're gazillion-dollar companies. I've had projects on my wish list longer than Aperture was a product.
That’s an interesting nugget I’ve not heard before about history of SwiftUI!
The same thing happened to me last year, and basically this is the reason I have been shadow-banned from HN. I stayed for 2.5 months only, but it was still too much, because I thought I got a decent salary. It cost me twice as much since then because I wasn't able to work since 4 months and I have no clear idea how I will crawl out this huge hole despite taking antidepressant, tranquillizer and going to psychotherapy every week.
DON'T PUT A PRICE TAG ON YOUR MENTAL HEALTH, and get out of a terrible job until it's too late! It's not worth it, believe me!
I'll point a friend of mine to your comment. He's constantly complaining of work. Once a bright engineer. Now his eyes look soulless.
> “Since this was a professional product, they knew the quality had to be high.”
I sighed heavily because instead we got Photos.app. Which is just bad and sad.
But then I continued to read about all the management screwups and development hell. And I now I know exactly why Photos.app is such a disaster.
I always run into this issue and I'm not alone, some people put their libraries in disk images just so they can properly eject their external disks: https://annoying.technology/posts/ebd70c6c5715bcd8/
- https://grumpy.website/post/0SZmcHq7V
- https://grumpy.website/post/0Rz1BBnjz
Not to mention that in the time it takes it to sync photos Google Photos can upload a few Instagrams worth of media.
But you're right otherwise.