A Good Conversation About Apple and APIs
When comments helps you grow
A couple of days ago, I hastily wrote a quite spicy blog post, about the removal of a good Spotify feature on iPhone. I got some pushback from Jason, and it led to a conversation I liked.
Me and Jason often disagree on stuff – but I always find his comments fair, interesting, and leading to me adjusting my opinions. One important takeaway in this for me, was whether or not this was just a case of someone whining, and putting a spin on, an API getting deprecated. Which is something that sometimes has to happen, but will still often lead to complaints which I often disagree with.
His first comment
I mean, it also could be there was an old way of accessing volume controls, Apple built a new way to work with HomePods, AppleTV, and likely Matter devices. They then deprecated the old API in favor of the new one to maintain one pathway, that yes, means access to its products comes along, but also means one, more modern, more standards compliant API to control devices.
I don’t trust Spotify comms for shit, and don’t know that’s what Apple did, but that story is equally like Apple– get rid of an old API for a new one that does more stuff, misses some functions but adds a lot more, etc. It would also line up with Matter and the addition of these devices that didn’t exist when the original implementation was done.
And for what it’s worth, the bugs that Spotify mention are present with the volume rocker over AirPlay. So, I’m not sure it’s not just Apple’s shit not being as good as it should be, period, versus disadvantaging someone.
My first reply:
The thing is, you can point to a million small examples where Apple makes (decently rational) choices that just so happens benefit themselves at the expense of others. Many of them are of the type “We made this choice, because privacy. That it hampers our competitors, is just a coincidence. So is the fact that we didn’t do this other thing which also would’ve been good for user-privacy, but would hurt us.” If you only look at each thing individually, Apple’s behaviour can often be defended – but you have to look at the larger patterns. Hehe, I agree that it’s a good idea to be skeptical of comms – but I don’t trust “Spotify + Sonos combined” less than Apple, I think. And Apple could’ve decided to comment, but has chosen not to.
And for what it’s worth, the bugs that Spotify mention are present with the volume rocker over AirPlay. So, I’m not sure it’s not just Apple’s shit not being as good as it should be, period, versus disadvantaging someone.
This is a good point! 👆🏻 But I still stand by my stance that these things would be better if the behemoths didn’t insist on competing in more and more markets. (I wrote more about this here.) We would have much cleaner incentive structures, if Apple’s “only” focus was on making their platforms (and hardware) as good as they could make them. (And “competition” would be an important ingredient here.)
His second comment:
I am deeply skeptical and in deep disagreement that a company cannot change APIs. And a load of comments about what Apple should do with it’s platform from a practical and technical standpoint amounts to “never break backwards compatibility; never test a new way of doing things before giving it to third-parties” under the idea that it’s anti-competitive. And I just think that’s total nonsense and nearly entirely orthogonal to meaningful anti-trust or anti-competitive behavior.
I don’t think Apple’s pattern is “harm competitors and advantage us”. I think Apple’s pattern is mostly “do the best thing for us, then figure out what we can make work for others on the platform.”
There are absolutely areas where I don’t agree with their actions, but 95% of the time someone complains about API changes or API deprecation, they seem totally wrong. The exception being when the changes happen without sufficient documentation– but that’s a rollout, relationship, and release management issue, not an indictment of the change itself.
My second reply:
I was a bit unclear: I’m not saying that Apple necessarily sets out to harm others. But, to me, it’s clear that where they fall on a bunch of decisions, is affected by the complicated incentive structures they’ve constructed. And that includes API decisions.
I don’t think it’s fair to say that because I can disagree on parts of an API change, I «don’t think companies can change APIs»… But yeah, I also often think complaints about API changes are overblown! And perhaps I’m guilty of the same in this case. 🙂
However, we can take another (pretty) recent API change as an example: When Apple updated the cloud storage APIs for the Mac, don’t you think it was easier to decide not to include support for external storage, when they know it remove a reason to choose Dropbox over iCloud Drive, and to not buy more internal storage on Macs? There are absolutely good parts of the new API! But if, say, Apple’s only incentive was to make macOS be better than Windows in handling of cloud storage, I feel like they would’ve tried harder…
Re: «never test a new way of doing things before giving it to third-parties" under the idea that it’s anti-competitive.»: I know that Apple likes to not communicate — but it would be so much easier to trust them, if they said things like «We’re working this private API accessible to third-parties in a privacy minded way», etc.
His third comment:
but it would be so much easier to trust them, if they said things like «We’re working this private API accessible to third-parties in a privacy minded way», etc.I don't really agree. Because they have a decades long track record of doing just this-- starting with a small first party feature, then expanding that feature alongside public APIs within a few years. In fact, I'm 98% sure that Craig Federighi and/or Greg Jozwiak have said as much in interviews-- that once they make an API public they have to support it for a long time (though not forever), so they often want to use it first internally before it's locked down enough (not privacy locked down, but functionally) before exposing it. I think their track record on that is great. As for Dropbox and external drives... well without getting into a massive thing, let's just say I do not agree with that assessment at all and I think that may be an even worse example to look at given how truly deep into the system Dropbox was getting. This is a case where the API wasn't really there at all, and I don't think Apple would have built _anything_ if Dropbox didn't already exist. Dropbox was hooking deep into the bowels of the system in more and more complex ways. I don't think Apple doesn't have this working on external drives to make Dropbox less appealing-- Apple would prefer to be able to support that feature themselves and compete on it. With all of these things, I think there's a lot of attribution to malice or competitive analysis that really comes down to time, difficulty, cost, and value to users. And in my opinion, if the result is fully rational before ascribing anything to some sinister anti-competitive incentive, it's silly to think the anti-competitive incentive is a driving force.
My last reply:
Hmm, interesting – I don’t think Apple’s track record (and things that has been revealed in court cases) gives them the benefit of the doubt, heh. 😅 But I wouldn’t mind examples that could convince me otherwise! (However, it’s important to keep in mind when discussing this: Even though I’m skeptical towards Apple on a bunch of things, I vastly prefer them to most of the other large tech companies, heh.)
I think their behaviour surrounding the App Store, locking down of the NFC chip, sudden “militarisation” of App notarisation, only Safari being allowed extensions, and other decisions, big and small, is pretty telling. Now, this year’s changes in iOS 18 actually addresses some of my complaints, specifically with the framework for Control Center + Lock Screen buttons. But while I have plenty of things I don’t like about the DMA, some of the good changes would never have happened without it.
(How good these are, are of course speculative. And I also think it will take quite a long time to shake out. I’m annoyed when people say “Huh, alternative app stores haven’t taken off”. Like, give it a couple of years, and we’ll see – _if_ it isn’t allowed to be nerfed into oblivion by Apple.)
One example which I have hopes will be a good example of you being right, is the journal entry API. I really liked that they made it accessible to third-party journaling apps from the get-go. But I was equally disappointed that third-party apps wasn’t allowed to provide fodder for it – so it would pick up something from the Podcasts app, but not Overcast. However, there’s absolutely a good chance this will be added later, and be a good example of your point. But it would be better if they said as much, as it could suddenly take 10 years as well. 😛 (BTW, That they’ve started to monetise the Podcast app as well, is also something that worries me. It creates yet an incentive for them to get users to choose their app over others. And I just don’t see how you can be so sure that they’re impervious to these incentives!)
As for Dropbox and external drives…
I think Dropbox pushing the boundaries to create a good product for customers is a good thing. And if it pushes Apple to create APIs they would’ve have created otherwise, that’s also good. A bit like Rogue Amoeba! I’m not saying Apple shouldn’t’ve made a new API. And I’m not saying Apple didn’t add support for external drives specifically to make Dropbox worse. I’m saying Apple’s incentives affects where they land, and prioritise, on these edge cases – in a way that makes things worse than they could be, for customers and smaller companies (which I value highly).
With all of these things, I think there’s a lot of attribution to malice or competitive analysis that really comes down to time, difficulty, cost, and value to users. I’m saying that Apple has put themselves in a position where “competitive analysis” is allowed to seep in and affect the choices regarding “time, difficulty, cost”, and in a way, “value to users”. But those things are, of course, also important factors.
I’d love more of this – even though the conversations don’t have to become this long, hehe.