It's All Software
daringfireball.net
daringfireball.net
I'm reminded of how Jean-Louis Gassée, head of Apple's Macintosh development in the late 1980s, responded when asked how they plan to connect Macs to the network. He held up a piece of phone wire and said "this is our network". This was a reference to Apple's proprietary LocalTalk protocol which could be used over regular RJ11 connectors.
The gist of this story is that Apple completely ignored the need for interoperability with the Mac. The iCloud has a whiff of the same hubris.
I think back there's a common belief in nerddom that it's gotta cover all these technical topics or else it's bad. Most people that use iCloud will not care about anything else other than the fact that their data 'magically' appears where it's needed. This is where I completely agree with Apple and diverge from the tech crowd - the best experience should just work.
The problem is that the dream of iPhone everywhere is dead for most people. Android isn't looking back. iPhone will never surpass Android in marketshare again. For most people, if the iCloud means an iPhone device, that means iCloud won't be used.
You shouldn’t forget that connection but I also wouldn’t exaggerate it. Also, iOS still has a sizable market share, maybe more than the Mac ever had? That helps, too.
There are circles within circles though... and I think that's what Apple is marketing to.
Market share is the runner-up prize for those who value a large community of vocal spendthrifts.
People aspire to an iPhone, they settle for Android.
I think the more abstract distinction comes from what each company wants you to do with their products, though. Apple wants you to use their hardware. What you use it for doesn't matter so much as you using it, and every decision they make, including leveraging the cloud, is informed by this.
Google, on the other hand, wants you to use the web, because the web is where they get paid. And the web is where we connect to each other.
Google's model makes that connection its fundamental primary motivation, while for Apple, the connection is incidental, just another feature.
This is unfortunately less true than it used to be. It matters to Apple if you want to run unapproved software on an iPad, or heaven forbid install Linux on it. I expect these restrictions will be coming to Macs over the next few years.
Google, on the other hand, wants you to use the web
Google wants you to use Google services and see ads that make Google money. They don't especially care whether that means web apps with AdSense ads or native apps with AdMob.
Doesn't it make more sense that Apple is trying to build a large and consumer friendly platform (approved apps) and that it would be an unjustifiable expense to support installing alternative OSes (Linux)?
Why is there always the assumption of nefarious motivations when it's more simply explained? They are only trying to make money, after all.
Considering the influx of open-source hackers and ex-windows users that migrated to Mac after the switch to Intel, this is just... nuts.
Or maybe just FUD.
I doubt it - iOS has only got less restrictive over time (From no 3rd party apps, to 3rd party apps in Objective-C, C++ and C, to 3rd party apps in any language)... Why would they would go in the opposite direction on the Mac and alienate all their current Mac users?
Google, on the other hand, just wants to serve you ads. That could be on HP, Dell, Apple, or Samsung hardware (although, I'm sure they'd prefer you bought Android or Chrome devices).
Personally, I prefer to buy Apple devices and use Google services. Apple makes the best hardware, but I don't want to be forced into buying them forever. With Google I can be assured access to my data from just about any future hardware vendor.
*Google is blurring the line a bit with Android, but still makes their real money from their services.
The Kindle example veers off slightly from Google because people pay for their ebooks.
Apple is a weird company because for all the eyebrows they raise with talk of closed systems and hubris and whatever else, they primarily make their way in the world by selling goods to consumers for money. The infrastructure enabling that is a crazy thing, but the heart of it is pretty old fashioned.
I do think my data is safe and private at Google (not so much trust in Yahoo and Facebook), but I am certainly looking forward to iCloud. Example: Gmail is a really good and free email-provider, but it also really sucks if you go through times that suck like a break-up and the adsense advertising is mirroring the unpleasant mails stored in your inbox.
If you aren't paying for the product you are the product.
Google also makes it fairly easy to export all your data if you decide you want to pick up and move.
The app store certainly makes it easy but its no where close to the easy installation and update of webapps - typing a name and hitting enter or clicking a link.
Wouldn't it be more appropriate to say that Apple's vision of the cloud is "data in the cloud. apps are local" whereas Google's vision of the cloud is "data and apps are in the cloud"?
All of Google's client-side software also comes down from the cloud the same way your data does.
And Gruber's point is that web apps still require internet access to work, albeit with improvements like cached storage in HTML 5. Whereas apps on the mobile device already work and are already taking advantage of sensors onboard the device, but the internet (cloud) makes them even better. For example, Garageband for iOS would be really hard to do as a webapp. Not impossible, but hard.
come on, google search is smarter than the app store search
I'm saying the App Store has categories, "Staff Picks", "iPad App of the Week" etc and it helps in discovery, just like Yahoo did back in the day.
Google would have to tread carefully here or they would step on advertisers' toes or invite anti-competitive scrutiny due to their dominance in search.
I'd beg to differ. It may not be identical, but App installation on my iPhone/iPad/Macintosh from the App Store is pretty much "Click, authorize, done"
I'm not saying the iOS process isn't dead simple - it is - but its still a step behind web apps.
App store installation does have one other advantage: a sense of ownership. The user feels they own the app rather than merely visiting it -- psychologically that is a powerful difference.
1) With local storage registration isn't necessary in a lot of cases. It is only needed so the user can have the same data on multiple computers.
2) We have Facebook Connect, Google Login, Twitter Anywhere, etc today. It's a 2 step process, "Login", and "Authorize".
The UX thing is why I'm slowly warming up to the Chromebook idea. I own an iPhone, and I love Macs, but I think ultimately the constraints of native applications will work against the users. We should cherish the million frameworks showing up instead of raising the "too much choice" flag just yet, as they will (hopefully) yield a development experience as enjoyable as the current tools for iOS, Android and Windows Phone seem to.
I think choosing between the two depends on what type of app you plan on making, the type of users you target, business goals and a ton of other options.
I love how he's correctly reframed the native vs web discussion but am disappointed he decided to make it a war again. It's not a holy issue: choose what's best for your business.
As for the user experience argument, I'm still looking for a valid excuse why the Facebook mobile website is so astronomically poorer in all aspects to the Facebook for iPhone app. In theory, they can and should function exactly the same when viewed on such a device, but they are leaps and bounds apart.
Ironically this is almost the polar opposite of Gruber's view. Gruber argues in this piece that web apps are local client apps running on your mobile device (and he urges you to view source). Both forms of apps (native and web apps) can mask latency in the same ways.
We have a professionally used iOS app that relies on server data (the magical cloud) and the number one user complaint is speed. Creating your client on iOS rather than in HTML/JS doesn't give you any magical "Latency and responsiveness" user experience unicorns.
http://en.wikipedia.org/wiki/Fallacies_of_Distributed_Comput...
Cloud is the service is the soul and is the only truth, but it also works best invisibly.
Gruber: "The first clause isn’t true, I’d argue. Shared document editing is not an inherent advantage of Google Docs being web-based; a native client could do it just the same."
Well, sure it could (subethaedit demonstrated that well) but Topolsky clearly stated that no native apps exist that allow it right now. Great, a native app could. Whoopee. No one's doing it yet. :/
It's true the distinction between networked native apps and networked browser apps is logically arbitrary; but the practical distinction is that the web is an open standard, cross platform, widely adopted, whereas Apple's native apps are tightly integrated with the device, giving them faster response and a better managed user experience. If iOS is everywhere, then you get the best of both worlds.
Can Apple topple the web? If they are better at providing what users want, yes.
well, then assuming a web page === a web "app"
the number of websites available to users is probably one million times the number of apps...
the number of new web sites that are started each day is probably one thousand times the number of new apps....
apple has a ways to go
and yes, i'll go there...you can't "kill" the real web if you do not allow users to access copious amounts of hardcore pornography
The current state of the web is such a crappy experience for regular consumers that Apple's model could make some headway against it and Google wouldn't want to be moving at the pace of standards against them.
The only way I see of furthering NaCl's agenda is if Adobe, Autodesk or some other company with a little cash to burn decides to write one of their apps (say, a Photoshop Express, or maybe a Sketchbook Pro) with it and prove in the market that there's a future in that approach. Otherwise, it will forever be just a cool hack.
Yes there is, and I can't believe Gruber missed this. I use the iOS mail client to access my GMail account. That's the point of Gruber's argument, that GMail is a service and can have multiple client apps of which the browser app is just one.
Something like Google Maps, or Topolsky's example of Google Docs, supports the point well enough.
Installing an app on a device is still the old way of doing things. All aspects of the application should come from the network - including the UI.
Why? Because then we can focus more on the nature of HCI itself instead of focusing on developing for different devices like we do now. Developers should be developing for screen sizes - not devices.
IMHO it's inevitable that we will all use one global standard for the client UI in the future. The web is the beginning of that and it's a powerful concept.
And let's not forget that every single one of Apple's iDevices will have Safari installed.
The user does not and should not care about where his or her application logic is being run so long as this device is affordable.
The vision you present adds questionable cost savings and substantial latency: (NY->SF->NY)/c will always take 32ms, best case. Taking this trip asynchronously instead will make sense for a very long time.
Apple understands that shuffling data between devices is a real burden to users, so they've attacked that problem instead of chasing an esoteric one.
He strikes me as a serious devotee first, analyst second.
Jesus (to quote you), people like you are worse than fanboys.
Deleted comment
This has never been the case, and I don't think it will for some time to come, if ever.
Innovation on the web moves at standards-speed (de facto and open). If 80% of your users are stuck on IE6 or lower (circa 2007), you will be hard pressed to make anything web-based as usable as the iPhone and it's native apps that showed up around that time.
Apple's cloud isn't for serving everyone; just users of their products. Consequently they can move faster than standards with their native binaries, and be out in front of web standards (with graceful degradation for non-Safari browsers) for web clients.
i mean....does anyone actually install pointless web-in-a-box apps like the rotten tomatoes app (for example because it bugs the hell out of me to decline it each time)?
my dsire to consume content is not going away, and having an app-per-url is a nonstarter