IBM brings Apple's Swift to the cloud, simplifying enterprise app development

AppleInsider

Thanks to new tools from IBM, developers will be able to write and run applications entirely in the cloud using Apple's Swift programming language.

In a move aimed at simplifying the development of end-to-end development of apps, IBM is the first cloud computing provider to enable the development of applications in native Swift. The next phase of IBM's roadmap brings Swift to the cloud with a preview of Swift runtime, as well as a Swift Package Catalog to help developers create enterprise apps.

There are three ways developers can now use Swift on the IBM Cloud:

• Experiment in the Swift sandbox: Quickly experiment with open sourced Swift, ramp up your skill set and learn what Swift can do for your enterprise by checking out new enhancements to the Swift Sandbox.
  
• Develop and deploy: Start building end-to-end applications on Bluemix and quickly deploy them with Kitura, a new open source web server released by IBM, on both OS X and Linux.
  
• Share Swift resources: Leverage code across projects by creating packages and submitting them to the Swift Package Catalog on Bluemix to encourage sharing of new Swift resources with the global developer community.

IBM touted its role as one of the largest users of Swift for mobile app development. Big Blue said its understanding of Swift and knowledge in assisting enterprises will maximize "the true potential" for server-side Swift.

"Modern digital apps require a modern programming language. Swift is easy-to-learn, reliable, fast and interactive, the key traits that CIOs look for when building the next generation of enterprise mobile apps," said Michael Gilfix, vice president of IBM MobileFirst Offering Management. "Swift on the Cloud is an opportunity for enterprises to radically simplify the development of end-to-end applications and therefore reach new levels of productivity."

According to IBM, bringing Swift to the server will break down barriers between front-end and back-end development. This will enable enterprises to use a single language to build rich experiences and back-end business logic.

Swift on the server will also allow developers to have a simpler and more secure toolchain for end-to-end creation of applications.

Apple and IBM laid the groundwork for Swift development in the cloud last December, when Apple made Swift open source, and IBM introduced a simple browser-based way for developers to get started writing code.

wizard69

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.  AppleInsider said: I don't care what Apple insider said, I don't need to quote the whole thing.  This software is maddening.  

thewhitefalcon

I wonder if this might make it possible, in the future, to handle iOS development on iPads. 

cyberzombie

This sounds exactly like what I've been hoping for - to setup Swift up as a viable middle tier product. It's got a couple of years to go before I'd suggest new development not be done in Java, but this is a big step.

volcan

I just signed up. What a dreadfully slow service. I guess the server couldn't handle the load after the announcement.

The validation email took more than ten minutes to make it to my inbox.
The log in screen still won't load.  

502 Bad Gateway: Registered endpoint failed to handle the request.

edited February 2016

mdriftmeyer

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.  AppleInsider said: I don't care what Apple insider said, I don't need to quote the whole thing.  This software is maddening.  

No it won't. That's like saying C is obsolete with C++ around. Look around. Sorry, but Swift is a long road and quite frankly since its development is probably a solid 7 years in development, isn't that far along.

cwoloszynski

mdriftmeyer said:

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.    

No it won't. That's like saying C is obsolete with C++ around. Look around. Sorry, but Swift is a long road and quite frankly since its development is probably a solid 7 years in development, isn't that far along.

I think your pace is off; I expect that real web services in Swift will come out quite quickly.  It won't replace old developments but it will win new developments in the next 12 months.  There are so many patterns and frameworks to leverage, the team doing this (and Perfect.org) have a robust roadmap to follow.  I'd guess 24 months and you'll see real services launched using Swift.  It is a great language.  We'll see who is right in 2018. 

dick applebaum

cwoloszynski said:

mdriftmeyer said:

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.    

No it won't. That's like saying C is obsolete with C++ around. Look around. Sorry, but Swift is a long road and quite frankly since its development is probably a solid 7 years in development, isn't that far along.

I think your pace is off; I expect that real web services in Swift will come out quite quickly.  It won't replace old developments but it will win new developments in the next 12 months.  There are so many patterns and frameworks to leverage, the team doing this (and Perfect.org) have a robust roadmap to follow.  I'd guess 24 months and you'll see real services launched using Swift.  It is a great language.  We'll see who is right in 2018. 

What if Apple opened iCloud Services to web serving, including access to FoundationDB and other's cloud services (e.g. IBM) ?

This would allow end-to-end development opportunities in Swift for:

1. Mac App Developers and iOS Developers
2. IT Departments

Apple could even offer a modern socially-oriented service targeted at consumers -- kinda' an iWeb on the wings of a Swift ... Cue Slim Whitman ...

edited February 2016

slprescott

dick applebaum said:

What if Apple opened iCloud Services to web serving, including access to FoundationDB and other's cloud services (e.g. IBM) ?

Dick, as always, I respect your opinion on this.

I wonder, however, if the original Apple/IBM partnership may have defined boundaries around how each of the two parties can monetize the collaboration, and I expect that any hosting services fell to IBM and its enterprise-focused cloud services.

Having said that, I agree that Apple offering this as a new plank of iCloud would be an interesting development, and IBM still would have plenty of opportunity to sell customization services around their portfolio of 100 mobile apps... and now the associated backend applications.

john.b

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.

Python's native power is in parsing data.  It's nearly the perfect informatics language.  Everything Perl could have been if it weren't for Perl's ugly unreadable syntax and style. I'm liking Swift so far, but I see it as a separate language with its own purpose.  It will be interesting to see if Swift can move out of the shadow of being Apple's Obj-C replacement.

edited February 2016

foggyhill

dick applebaum said:

cwoloszynski said:

I think your pace is off; I expect that real web services in Swift will come out quite quickly.  It won't replace old developments but it will win new developments in the next 12 months.  There are so many patterns and frameworks to leverage, the team doing this (and Perfect.org) have a robust roadmap to follow.  I'd guess 24 months and you'll see real services launched using Swift.  It is a great language.  We'll see who is right in 2018. 

What if Apple opened iCloud Services to web serving, including access to FoundationDB and other's cloud services (e.g. IBM) ?

This would allow end-to-end development opportunities in Swift for:

1. Mac App Developers and iOS Developers
2. IT Departments

Apple could even offer a modern socially-oriented service targeted at consumers -- kinda' an iWeb on the wings of a Swift ... Cue Slim Whitman ...

It is probably in the cards one day, They've been with IBM for relatively short time and have moved pretty fast on this.

foggyhill

cyberzombie said:

This sounds exactly like what I've been hoping for - to setup Swift up as a viable middle tier product. It's got a couple of years to go before I'd suggest new development not be done in Java, but this is a big step.

Java is the most horrible thing known to man and I'll be happy to see it die (yes, a bit of hyperbole but I despise a lot of it, especially because it's been shoehorned where it shouldn't way to often).

slprescott

Can someone explain the Java-hate that has emerged over the past few years?

I'm not defending Java, I just honestly don't know what the issues are. 10+ years ago Java was considered a great language -- and enterprises still use it heavily -- so I just don't know what concerns have turned popular sentiment against it.

dick applebaum

slprescott said:

dick applebaum said:

What if Apple opened iCloud Services to web serving, including access to FoundationDB and other's cloud services (e.g. IBM) ?

Dick, as always, I respect your opinion on this.

I wonder, however, if the original Apple/IBM partnership may have defined boundaries around how each of the two parties can monetize the collaboration, and I expect that any hosting services fell to IBM and its enterprise-focused cloud services.

Having said that, I agree that Apple offering this as a new plank of iCloud would be an interesting development, and IBM still would have plenty of opportunity to sell customization services around their portfolio of 100 mobile apps... and now the associated backend applications.

If it were any company but Apple, likely, I would agree that there are boundaries around their collaboration ...  

However, since the Jobs2 era, I can't think think of any partnership where Apple has limited their options.

I think that SJ was a great negotiator ( though my daughter thinks he was an "arrogant asshole").

Don't know Tim, but based on his performance, I suspect that he is an even more effective negotiator than SJ.


Consider:

• Chris Lattner goes to work for Apple
• Swift is less than 6 years old
• Swift is less than 2 years available to Apple Developers
• Swift is the hot commodity
• Apple open-sources Swift
• IBM et al enter into partnerships with Apple

Why ... What's in t for Apple?

edited February 2016

dick applebaum

foggyhill said:

dick applebaum said:

cwoloszynski said:

I think your pace is off; I expect that real web services in Swift will come out quite quickly.  It won't replace old developments but it will win new developments in the next 12 months.  There are so many patterns and frameworks to leverage, the team doing this (and Perfect.org) have a robust roadmap to follow.  I'd guess 24 months and you'll see real services launched using Swift.  It is a great language.  We'll see who is right in 2018. 

What if Apple opened iCloud Services to web serving, including access to FoundationDB and other's cloud services (e.g. IBM) ?

This would allow end-to-end development opportunities in Swift for:

1. Mac App Developers and iOS Developers
2. IT Departments

Apple could even offer a modern socially-oriented service targeted at consumers -- kinda' an iWeb on the wings of a Swift ... Cue Slim Whitman ...

It is probably in the cards one day, They've been with IBM for relatively short time and have moved pretty fast on this.

I think you broke the code!  In today's world, timing is everything!  Apple has been doing Swift stuff internally for over 2 years -- so they're up to speed --  today they have at least a 2-year lead over the competition!

nmdias

mdriftmeyer said:

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.  AppleInsider said: I don't care what Apple insider said, I don't need to quote the whole thing.  This software is maddening.  

No it won't. That's like saying C is obsolete with C++ around. Look around. Sorry, but Swift is a long road and quite frankly since its development is probably a solid 7 years in development, isn't that far along.

Swift has been in development for the past 5 and a half years. Accordingly to Apple they needed as much as 2 more years to stabilise it (this was said in the announcement keynote) and that date matches the release date of version 3 of Swift. Which also, accordingly to the Swift's evolution repo will be the version that will accomplish this. The version from which after will introduce the least breaking changes.

I also don't believe it will replace C++ or Python. Well, Python...maybe in a far fetched sense... But C++?

It will certainly appeal to Swift developers as they will be more comfortable developing backend services with a known language. 

Anyways, Swift is pretty great been using it since it's release and is all I code now, 8 hours a day. We even rewrote everything from the ground up in Swift. So I guess that's a step right there. I'm pretty excited about services like this. IBM's, perfect.org... I'm crossing my fingers for Google to add Swift to their App Engine :P 

edited February 2016

cyberzombie

mdriftmeyer said:

wizard69 said:

With this sort of backing and rapid development Swift might just replace C++, Python and other popular web development languages.  It will be interesting to see where Swift is in three years.  AppleInsider said: I don't care what Apple insider said, I don't need to quote the whole thing.  This software is maddening.  

No it won't. That's like saying C is obsolete with C++ around. Look around. Sorry, but Swift is a long road and quite frankly since its development is probably a solid 7 years in development, isn't that far along.

I guess I'm just more optimistic. The reasoning that Swift's ARC is more memory efficient than Java's various garbage collectors (by 2x in most cases) doesn't hold much water these days. Just throw hardware at it. But Swift has optionals (which Java8 has in the Optional<T> class - but that's a kludge), tuples, and the splendid switch statement. In a couple of years the heavy lifting will be in place (IBM's WAS is a no-brainer, but I'd expect an Apache group will form around a Tomcat port...the same for Spring). No need for existing projects, but I could see enterprises start to put out pilot projects in Swift in 2-3 years.

The big issue I still see is that Swift doesn't have reflection, which I think it needs (as opposed to using tools to generate converters ala JAXB). Who cares if reflection adds a significant performance penalty in its use. Converting a JSON web request will still be orders of magnitude faster then the server SQL query to handle the request.

dewme

This sounds more like IBM is trying to position Swift runtime and its package model as an alternative to JavaScript plus Node.js plus NPM. It sounds like a good deal for developers who prefer Swift over JavaScript. 

edited February 2016

dick applebaum

dewme said:

This sounds more like IBM is trying to position Swift runtime and its package model as an alternative to JavaScript plus Node.js plus NPM. It sounds like a good deal for developers who prefer Swift over JavaScript. 

Without getting inside the head of Chris Lattner -- it's difficult to know if that was one of his original goals for Swift.

But as Apple began implementing Swift, it's likely that they (and IBM) saw the potential as you describe.

AIR, when Apple formally announced Swift at WWDC 2014, they defined it as:  ranging from a high-level System Programming Language to a low-level Scripting Language.

Consider the timeline:

• July 2010:  Lattner begins development of Swift
• June 2014:  Apple Announces Swift *
• July 2014:  Apple/IBM Announce Collaboration on MobileFirst *
• Dec 2015:  Apple open-sources Swift
• Dec 2015:  IBM Introduces Swift Sandbox
• Dec 2015:  IBM releases 100th MobileFirst app in collaboration with Apple
• Feb 2016:  IBM brings Swift to Cloud

* IBM doesn't do anything in 1 month -- likely, the collaboration on Swift & MobleFirst began some time in 2013

In retrospect, this appears to be a coordinated and well-executed plan!

edited February 2016

dick applebaum

The small (micro?) elephant in the [server] room.

With the availability of 64-bit ARM processors we may be at an inflection point for stand-alone servers, backroom servers, server farms and distributed servers ...

For example,  it may [soon] be possible to: 

• Increase the capacity of a server farm by an order of magnitude while reducing power, space and air-conditioning
• Provide a reliable backroom server with something as inexpensive and simple as an AppleTV
• Increase the performance and availability of servers by distributing the workload *
• Provide a viable home server

* FoundationDB anyone?

Apple has advantages in that it has been running Swift on ARM for more than  2 years -- and Apple can tailor its ARM hardware to the tasks at hand.

edited February 2016

rcfa

Anyone who did WebObjects in Objective-C starts having a deja-vue; then the Java craze came and destroyed it all. Swift is WebScript, Objective-C on steroids, what Java should have been.

There are a few things I miss in Swift from an experimental language called TOM that was developed in Holland; and of course I hope eventually we get to see a classic more ObjC like syntax; Swift is as ugly to write as C and Java; it lacks ObjC's elegant message syntax.