Along with the ARC transition, the move from PowerPC to Intel was incredibly smooth (relatively). But much of the work for that was done long before the change actually happened.

Both of those changes required very little from developers. Obviously a new language is going to take much more. The biggest challenge with Swift is keeping the programming interface simple. While it gets rid of the @ signs that people stumble over in Objective-C, Swift didn't really reduce the amount of learning required to create a program. Rather, it seems to be everyone's new favorite place to talk about exotic programming concepts, which results in terrifying blog posts like this one: https://developer.ibm.com/swift/2016/01/27/seven-swift-snare...

I wonder how long it will be before someone like Douglas Crockford shows up with Swift: The Good Parts.

I think the major advantage for new developers isn't the level of conceptual complexity but much clearer and cleaner syntax, which also happens to be more inline with most other modern popular languages. That's a major advantage.

No, it doesn't. C++ sucks.

The Gmail native iOS app is really nice, so the flow of innovation isn't consistently pointing towards the web. But part of the answer to getting "desktop deployment right" is sandboxing. In essence, the native app platforms are making operating systems more like web browsers, while projects like Mozilla and Chrome are trying to turn web browsers into operating systems.

Vishal Gurbuxani gave us the marketing perspective on this last month in Napa: http://www.slideshare.net/vgurbuxani/app-retreat-2012

One major problem: HTML5 is fragmented across platforms and is missing important ingredients needed by app businesses. In the end, it's just another platform, and not a very good one if a good user experience is something that you hope to deliver.

Read Vishal's slides for more, or come ask him about them at his session at Renaissance: http://renaissance.io/speakers/vishal_gurbuxani

Bill Dudney and I are organizing a conference for iOS app makers in San Francisco next month. It's unique because it's not being run by a platform company, a media company with a network of sponsors to satisfy, or a professional conference organizer. We're just two app developers with some ideas about how to make an event that's more productive for app developers like ourselves. Check out our speakers, and I'm sure you'd find that the people who are signed up so far are just as impressive. Join us?

Oops. It's back now. Sorry about that, I intend to upgrade that site but have been taking on some new things lately and haven't had much time. Also, I don't really follow Hacker News, but a friend pointed this post to me. Feel free to follow up with me any time on Twitter, email me at tim at radtastical dot com, or come see me at one of the meetups that I organize. http://meetup.com/sviphone

I've never heard that web apps require approval. Probably a typo? Also, it would be nearly impossible to "crash the whole iphone OS." An app runs in a sandbox, the worst thing it could crash is itself.

No typo, timburks.

Before the app store debut around July 2007, web apps was the only way to develop for the iphone.

With Webapps, Apple simply place your application on their http://www.apple.com/webapps page and gives you a chance to promote yourself.

Apple does not seem to promote those web apps as much as they should.

No, I think you're right. There seem to be separate queues for different categories.

Does the computer you are using have a hierarchical file system?

Of course, but would you really want to use a more advanced menu just to find a single-serving app on a mobile device?

I'd just let the market decide. With so many identical-looking screens of apps, most iPhones already are getting cluttered. I expect that we'll soon see a better (hierarchical) Springboard.

How about those book publishers who attempt to milk their printing presses by publishing multiple titles? Or bloggers who attempt to milk their blogging software by publishing multiple articles?