> The gaming phone is neat idea and I’ve been actually thinking this is the way Microsoft gets back into phones. Leverage their XBox platform to build a phone meets Nintendo DS like device. Thats what I thought the Surface Duo was going to be. If MS could extend their embrace of PWA (since they don’t have a mobile app ecosystem of their own) this starts to look a lot like the Epic phone idea. However, in either case doesn’t this just create another closed platform to compete with existing ones?
It’s not obvious why it would have to be closed. That may be the case if Android is the starting point, but Epic and the streaming game services would seem to benefit from an open platform.
You could argue that they might be tempted to try to close the platform so as not to leave money on the table, but I think that’s just a mistake. By making the platform open, they can deflate the value of closed platforms, and capture the gains of doing what they already do, but without having to partner with Apple.
I see no reason there shouldn’t be an open platform with multiple stores on it which sell proprietary software. This would be much better than a proprietary platform owned by Apple, grudgingly forced to support alternative stores.
> Everyone would start with Android forks it seems and want to control an app store.
I think that’s a tempting, but bad strategy. I assume this is why we don’t see it done.
> From your examples I see 3 categories of things:
1. control over operating system feature (cloud vs local backup) 2. API access to platform compilers, linkers, loaders, at runtime (to build programming environments) 3. decentralized distribution (no single fault, p2p, etc)
> As a user not a developer, the only one I feel is missing is runtime programmability which would enable all sorts of first class development tools on my ipad or iphone. I wouldnt be limited to xcode, or swift playgrounds. I do feel this is coming soon, it’s just Apple being careful with safety and security. Swift playgrounds is dogfooding these APIs that will eventually open up.
That has been my strong desire as a user too and I’ve argued this exact point. It seems like closing the loop so that code goes iOS -> App Store - iOS, rather than Mac -> App Store -> iOS, with Swift / SwiftUI only apps would be a huge streamlining. I have imported fairly complex complete swift only Apps into playgrounds on even a 5 year old iPad, and it’s surprising how well they work, and how responsive the compiler is.
That said, there are still many other kinds of programming that are not supported in that model.
Much as I like Swift, programming models and languages are in a Renaissance right now, and I’m not going to stop wanting to participate in other language ecosystems even if Apple does this.
Therefore, this is a point of differentiation for an open platform.
> local backup vs cloud is interesting. I thought Apple allows this through a USB cable?
They do. But that’s an early 2000s solution! Who wants to plug in their device to back it up? Also what if you don’t have a desktop machine?
> At some point I stopped bothering with local backups because it takes up precious local space and I don’t trust my own ability to maintain computer hardware.
Very reasonable, and probably a good approach for many people, and frankly what I have done too.
> “The cloud” does this better.
Does it? iCloud backups are not e2e encrypted, and are available to law enforcement and anyone who can hack Apple. I trust Apple more than almost any other company in the space, but they are not infallible, definitely not always going to win in court, and nobody is unhackable.
Obviously a local NAS or independent cloud service is not immune either, but it could certainly be e2e encrypted, or have open source code that could be independently verified etc.
Also, access to the file system and system in general means that all kinds of other interoperability between local devices becomes possible.
I think Apple’s approach is a great turnkey system and I don’t want to see it demolished, but I think many other possibilities do exist.
> To me, cloud backup is just 1 of N operating system features that depend on cloud services.
Agreed - it was just meant as one example.
> I suppose an open platform could be service-less by default.
Yes, or have layers of infrastructure that you aren’t forced to use.
EDIT: not trying to say that Apple is the holy grail or anything, just trying to figure out what an open ecosystem would add that changes the calculus for people using iphones.
> What does decentralized app distribution get us?
It allows us to explore different trust and discovery models, and it frees us from a single point of technological or policy failure.
I delegate a lot of trust to Apple. They are pretty good. I don’t want to have that diluted, but I do want us to be able to develop alternatives.
In many ways you can see this as a classic innovators dilemma situation. Apple is over-serving with an integrated solution, but can’t cover all of the use cases, and so the advantages of a modular solution start to increase.
They managed to stave this off for a long time by optimizing execution, and because the plateau was further away than competition realized, but that doesn’t necessarily work forever.
It also doesn’t mean they will now be destroyed, because execution and lowest common denominator appeal count for a lot. They may continue as the market leader of mass market devices, but in a market where there are more options and they are less dominant.
It is interesting to make a connection between integrated solution and over-serving. To me, integration seems like a defense against feature sprawl and over-serving. Not saying they can't be disrupted, I just don't think it comes from their choice to integrate.
Disruption wouldn't come from Apple not covering all the use cases, but from making a really important use case harder to access than it should be. Amazon's Echo is a classic attack in this regard. If Amazon bets that Siri-like interfaces are important it should build a device that mostly does that, and does it better.
I don't see the smartphone as a whole being disrupted, especially by an open platform, since I still don't see a use case that Apple can't copy. If there was an innovators dilemma based attack it would need to radically simplify the interface. The problem is a multi-touch screen and display make the core value the diveristy of things the phone can do. A product cannot really collapse over some subset.
It’s not obvious why it would have to be closed. That may be the case if Android is the starting point, but Epic and the streaming game services would seem to benefit from an open platform.
You could argue that they might be tempted to try to close the platform so as not to leave money on the table, but I think that’s just a mistake. By making the platform open, they can deflate the value of closed platforms, and capture the gains of doing what they already do, but without having to partner with Apple.
I see no reason there shouldn’t be an open platform with multiple stores on it which sell proprietary software. This would be much better than a proprietary platform owned by Apple, grudgingly forced to support alternative stores.
> Everyone would start with Android forks it seems and want to control an app store.
I think that’s a tempting, but bad strategy. I assume this is why we don’t see it done.
> From your examples I see 3 categories of things: 1. control over operating system feature (cloud vs local backup) 2. API access to platform compilers, linkers, loaders, at runtime (to build programming environments) 3. decentralized distribution (no single fault, p2p, etc)
> As a user not a developer, the only one I feel is missing is runtime programmability which would enable all sorts of first class development tools on my ipad or iphone. I wouldnt be limited to xcode, or swift playgrounds. I do feel this is coming soon, it’s just Apple being careful with safety and security. Swift playgrounds is dogfooding these APIs that will eventually open up.
That has been my strong desire as a user too and I’ve argued this exact point. It seems like closing the loop so that code goes iOS -> App Store - iOS, rather than Mac -> App Store -> iOS, with Swift / SwiftUI only apps would be a huge streamlining. I have imported fairly complex complete swift only Apps into playgrounds on even a 5 year old iPad, and it’s surprising how well they work, and how responsive the compiler is.
That said, there are still many other kinds of programming that are not supported in that model.
Much as I like Swift, programming models and languages are in a Renaissance right now, and I’m not going to stop wanting to participate in other language ecosystems even if Apple does this.
Therefore, this is a point of differentiation for an open platform.
> local backup vs cloud is interesting. I thought Apple allows this through a USB cable?
They do. But that’s an early 2000s solution! Who wants to plug in their device to back it up? Also what if you don’t have a desktop machine?
> At some point I stopped bothering with local backups because it takes up precious local space and I don’t trust my own ability to maintain computer hardware.
Very reasonable, and probably a good approach for many people, and frankly what I have done too.
> “The cloud” does this better.
Does it? iCloud backups are not e2e encrypted, and are available to law enforcement and anyone who can hack Apple. I trust Apple more than almost any other company in the space, but they are not infallible, definitely not always going to win in court, and nobody is unhackable.
Obviously a local NAS or independent cloud service is not immune either, but it could certainly be e2e encrypted, or have open source code that could be independently verified etc.
Also, access to the file system and system in general means that all kinds of other interoperability between local devices becomes possible.
I think Apple’s approach is a great turnkey system and I don’t want to see it demolished, but I think many other possibilities do exist.
> To me, cloud backup is just 1 of N operating system features that depend on cloud services.
Agreed - it was just meant as one example.
> I suppose an open platform could be service-less by default.
Yes, or have layers of infrastructure that you aren’t forced to use.
EDIT: not trying to say that Apple is the holy grail or anything, just trying to figure out what an open ecosystem would add that changes the calculus for people using iphones.
> What does decentralized app distribution get us?
It allows us to explore different trust and discovery models, and it frees us from a single point of technological or policy failure.
I delegate a lot of trust to Apple. They are pretty good. I don’t want to have that diluted, but I do want us to be able to develop alternatives.
In many ways you can see this as a classic innovators dilemma situation. Apple is over-serving with an integrated solution, but can’t cover all of the use cases, and so the advantages of a modular solution start to increase.
They managed to stave this off for a long time by optimizing execution, and because the plateau was further away than competition realized, but that doesn’t necessarily work forever.
It also doesn’t mean they will now be destroyed, because execution and lowest common denominator appeal count for a lot. They may continue as the market leader of mass market devices, but in a market where there are more options and they are less dominant.