Official tailored Swift for Windows support promised in 5.3

The team working on the Swift programming language has said its goals for the upcoming version 5.3 will include "adding support for Windows and additional Linux distributions". Swift is a programming language developed by Apple and first released in September 2014. It won immediate popularity as an alternative to Objective-C …

[Anonymous Coward]

Why bother?

I doubt 99% of Windows developers will even notice or care, and those that do will have a similar experience to developers on Apple platforms: I.e., language features changing so often it's impossible to keep up, and you can't stick with a single version for too long either because the surrounding tools obsolete older versions just as quickly.

So far Swift has had 29 releases in five years. Compare that to the nine releases of C# in 20 years. Even the rapid release schedule of C++ is out of control with features being deprecated and replaced more often than I care to count. Instead of jumping on the bullshit release-often bandwagon, can language developers actually starting thinking about things for more than 30 seconds before implementing them so we can have languages and libraries that are actually designed properly and are more useful than annoying?

</rant> :)

[anonanonanon2]

Re: Why bother?

As a late adopter of swift, mostly because it changed so rapidly at the start, I can say it's not making so many massive changes anymore, and we've had to do very little in terms of updating code to newer versions over the past couple of years.

I now start new projects in Swift, but Obj-c still has some advantages, especially if you want to include cpp libraries, and I find myself going back to that sometimes.

[karlkarl]

Re: Why bother?

"I now start new projects in Swift, but Obj-c still has some advantages, especially if you want to include cpp libraries, and I find myself going back to that sometimes."

This is so true. It also is one of the reasons why I predict Objective-C will outlive Swift. When Apple replaces Swift with Boop-lang or whatever, for native interop you still need a language *based* on C.

Perhaps if Apple did put a tiny C or C++ compiler into Swift and allow it to consume C/C++ headers / libs, they might be onto something.

I hate writing bindings basically ;)

[Glen 1]

Re: Why bother?

" language features changing so often it's impossible to keep up, and you can't stick with a single version for too long either because the surrounding tools obsolete older versions just as quickly."

something something javascript frameworks

[Charlie Clark]

Swift for ML? That ship sailed a while ago, I think

Windows Swift for Tensorflow, used for working with ML models, is a good use case

Hardly, it will be competing against the established Python toolchain, which with things like Jupyter Notebook and Pandas, just brings more to the party. As all the heavy lifting is handled by Tensorflow, there's little to be gained from minimal improvements in the language marshalling data to run on some GPUs somewhere.

[Version 1.0]

Re: Swift for ML? That ship sailed a while ago, I think

Another language ... how many languages are people programming in these days? Or should that be "this week" - given the rate at which "new languages" appear, I expect that at some point in the future every programmer will have their own personal language ... a couple of billion computer languages by the end of the century, is that a big deal?

These days most new languages look like C rewritten in a bar with free beer over the weekend.

[a_yank_lurker]

Re: Swift for ML? That ship sailed a while ago, I think

"how many languages are people programming in these days?" - This has always been a problem with there being numerous languages and versions in circulation since the mid 60s. It seems like it is a requirement for academics and businesses to develop languages that are really vanity projects with some unfortunately becoming widely used. Compound this with bungled designs (aka JS) and other stupidities and you a language landscape that is mess.

[Charlie Clark]

Re: Swift for ML? That ship sailed a while ago, I think

I don't mind new languages per se. And there are always ones worth taking a look at: Rust for example seems to have found a niche and plenty of friends, and there will always be a need for a new functional language…

Swift might have had more success if Apple had made it open source from the start and didn't restrict the GUI to Apple devices. As it is, it's nice for research and I'm sure some people will be looking to borrow good ideas from it.

[Fruit and Nutcase]

FAO Taylor Swift

Can you bring some Swiftmas cheer to the team working on the Swift programming language by sponsoring their work?

[Anonymous Coward]

“but unfortunately the idea of taking your Swift application for macOS and/or iOS and recompiling it for Windows will not fly.”

Whatever framework/language you use it never flies, the concept just doesn’t work. All platforms end up with a substandard lowest common denominator experience.

You want the code that actually does work shared, fine, but stop being lazy and develop UI separately for each platform. (NO, web is not the solution either).

[Anonymous Coward]

What Are Languages For??

....basically all procedural languages address the same problem. The choice of a procedural language might be about performance (say C), or it might be about productivity (say Python), or it might be forced because of history (say COBOL), or it might be about individual taste.

*

Some languages address very specific problems, which narrows the choice. SQL comes to mind -- it focuses on relational databases, and it is declarative (not procedural).

*

But surely the bottom line is delivering reliable functionality over time to some community of users. If the community is very small (say a research group) then I suppose the language choice doesn't matter. Otherwise I guess C and SQL are going to out there for a long while to come. Swift....maybe not so much!