Swift 5.5 for FreeBSD.

If the people most interested in FreeBSD (those who read this forum)
Already wrong assumption.

Interest: https://wiki.freebsd.org/WantedPorts#Q-T

Just as an example: MakeMKV was on this list as well. MakeMKV has a forum, there was a FreeBSD thread with very little interest. I was personally interested in using it on FreeBSD, so I created multimedia/makemkv. Other people are using it. So,
why would anyone bother to do a port?
BECAUSE YOU WANT TO HAVE IT. Simple as that.
 
It should be noted that Swift on non-Apple platforms will not have access to Cocoa or any of those "batteries included" systems that Swift leverages. This likely contributes to a lack of interest.

This means that even if we did have Swift support, we would need to put in so much work for language bindings. We couldn't even use many existing ones either because they do drag in (and are encouraged to do so) Cocoa systems in one way or another.

IBM originally expressed interest in Swift but quickly backtracked for this reason. In general I would recommend against Swift; not because the language is bad but simply because its upstream developers don't offer any real support on 3rd party platforms. Maybe it is why Objective-C never made much headway outside of Apple either.

i.e imagine C-Sharp without .NET or Java without JRE/Swing.
 
It should be noted that Swift on non-Apple platforms will not have access to Cocoa or any of those "batteries included" systems that Swift leverages. This likely contributes to a lack of interest.
This is a special case of the general case why I'm not interested in vendor-specific "open" platforms like Dart, Kotlin, or Swift.

Big decisions in community-driven open projects are made in the open and driven by the members of that community. I may not always like the results, but I know they won't be splash damage from idiot Harvard MBAs reading the latest trendy business book.

Swift is and will be whatever Apple execs decide. The community may get some input, but its consideration will be secondary at best. I'm not going to waste my precious free time working for Apple for free.
 
It should be noted that Swift on non-Apple platforms will not have access to Cocoa or any of those "batteries included" systems that Swift leverages. This likely contributes to a lack of interest.
There seems to be a disconnect between what you are discussing and Swift's use on the Server.

There is not a lack of interest in Server Side Swift, just a lack of interest for anyone using it on FreeBSD. Clearly the server side community has been growing as last year Amazon introduced a Swift Lambda and released Smoke, their server side Swift framework. This year they introduced a native Swift AWS SDK (mostly server side focused). Not sure why you are discussing Cocoa (or SwiftUI for that matter), as neither is used for server-side applications (the primary use of FreeBSD and Linux.

This means that even if we did have Swift support, we would need to put in so much work for language bindings. We couldn't even use many existing ones either because they do drag in (and are encouraged to do so) Cocoa systems in one way or another.
When was the last time you developed a server side application in Swift or even looked at its use for server side development? To which particular language bindings are you referring?
IBM originally expressed interest in Swift but quickly backtracked for this reason.
IBM does a large amount of Swift development, they just stopped development of Kitura, their own server side framework. That happened mostly because they got there too early. Apple had not yet produced the Linux tools needed to have it be a first class development platform. Since then, Apple has added much better tooling, and third parties (including Microsoft) have added LSP support for Swift.
In general I would recommend against Swift; not because the language is bad but simply because its upstream developers don't offer any real support on 3rd party platforms. Maybe it is why Objective-C never made much headway outside of Apple either.
You say this based on what? Apple is an active participant in the Server Side Swift Working Group, and has steadily added open functionality for this market.
i.e imagine C-Sharp without .NET or Java without JRE/Swing.
Foundation, SwiftNIO, libdispatch are all part of the open release. No one tried to promote Objective-C as a server side language. People wrote runtimes that were supposed to be compatible with Apple's but they were not from Apple, and given that even many macOS/iOS developers only used Obj-C for UI code (with C/C++ for most other stuff), not much of a surprise that there was not substantial adoption.
 
This is a special case of the general case why I'm not interested in vendor-specific "open" platforms like Dart, Kotlin, or Swift.
Swift's use on the server has nothing to do with Apple's platforms, so I am not how it is "vendor-specific".
Big decisions in community-driven open projects are made in the open and driven by the members of that community. I may not always like the results, but I know they won't be splash damage from idiot Harvard MBAs reading the latest trendy business book.
The Server Side Swift Working Group has the following members:
With Tom Doron (@tomerd, Apple) representing the Swift core team.

Not a single Harvard MBA in the group. Two of the SSSWG are from Apple, two from Amazon, two from MongoDB, two from Vapor, and Adam Fowler is an independent who did the first Swift AWS SDK (Fabian Fett wrote the first Swift based Lambda before he went to Apple).
Swift is and will be whatever Apple execs decide. The community may get some input, but its consideration will be secondary at best. I'm not going to waste my precious free time working for Apple for free.
Swift is an open source project, and the use of Swift on the server is driven by a large community with Apple being only a small part. Were Apple to suddenly change the language in a fundamental way, nothing would stop the community from forking it.

However, I am curious how developing server side applications using Swift would be "working for Apple for free"?
 
You say this based on what?
Again, mainly from the information gathered by IBM pulling out of server-side Swift (and not just Kitura it seems).

This article mentions a few challenges cited by IBM.

https://www.infoq.com/news/2020/01/ibm-stop-work-swift-server/

No one tried to promote Objective-C as a server side language

Honestly, a capable language shouldn't need "promoting". I am not saying Swift (or Objective-C) are not capable; it just seems that interest in these languages outside of Apple is lethargic at best.

FreeBSD has loads of crazy niche languages supported that some guys care about. It is almost a little funny just how ignored Swift is. Even Brainfuck and ancient obsolete languages see more interest.
 
Unfortunately, what's listed (the link within the table) is a demonstration of disinterest, with no indication of how many people are interested.
I would agree with you.

In response to a simple question trying to gage community interest, I got nine people responding. Not one expressed any interest in Swift (on any platform, let alone on FreeBSD).

Instead, I got people lecturing me on how the ports process here works, complaining that I was selfish for asking others to do a port, valuable career advice (from people with no knowledge of me at all), someone who pronounces what Apple will or will not do (with no evidence or support of his statement), and two people who specifically are not interested in having it happen (one who explained that even if it did happen it would have no value because it would be so incomplete that one would not be able to use it anyway).

All super helpful.
 
Again, mainly from the information gathered by IBM pulling out of server-side Swift (and not just Kitura it seems).

This article mentions a few challenges cited by IBM.

https://www.infoq.com/news/2020/01/ibm-stop-work-swift-server/

The article talks about the state of the world about two years ago (It was released in January of 2020, based on decisions IBM made in October and November of 2019). Even in the article Chris says:

Apple is working hard to address these issues, to make Swift more open, and to help build the server ecosystem — and this has recently stared to pick up pace. Tom Doron has been a driving force from Apple in promoting the server ecosystem through the Swift Server Work Group and leading Apple's efforts in the server space. Additionally Ted Kremenek has recently posted On the road to Swift 6, which describes a strong statement of intent on steps to expand the ecosystem and make it more open — including driving more focus around the fledgling Language Server Protocol (LSP) project, which will enable other IDEs to better support Swift development.
All of which have borne fruit. Since those decisions were made, the open source Swift Package Manager was released with solid build support for Linux, and Apple released SourceKit-LSP enabling third party IDE support for Swift (of which there are now several). Adopting SPM in Xcode makes cross development for Linux much easier, as does the substantially improved Docker support. Amazon jointly developed an AWS Lambda Swift runtime and their Prime Video team built their core system on Smoke, a server side swift framework they have since released publicly as another open source, server side swift framework. In other words, in the two years since that article was written, the issues that made IBM drop its own server side Swift framework, have been resolved.
Honestly, a capable language shouldn't need "promoting". I am not saying Swift (or Objective-C) are not capable; it just seems that interest in these languages outside of Apple is lethargic at best.
Objective-C, even in the iOS/macOS world was primarily used for building user interfaces. Given that none of the tools nor any of the libraries that made it interesting on those platforms was open source, it should not surprise anyone that it did not receive adoption outside of those platforms.

You say adoption of Swift outside Apple's platforms is "lethargic at best", where as the evidence of increasing support on the server seems pretty clear. In the time since that article was released, the dominant cloud provider has released Lambda support, its own server side Swift framework (that they use for a mission critical set of services for a critical part of their business) and has introduced its own, native Swift AWS SDK. Support went from limited on a single Linux distribution to substantial on three (including AWS's native Linux) and reasonable on Windows. Not my definition of "lethargic".
FreeBSD has loads of crazy niche languages supported that some guys care about. It is almost a little funny just how ignored Swift is. Even Brainfuck and ancient obsolete languages see more interest.
Based on the response to this thread, I am not that surprised. The attitude towards Swift here seems to be characterized as somewhere between disinterest and genuine hostility.
 
Why are you blaming and shaming others, when you can't or don't want to do it yourself.

I've bought a book on Swift a long time ago. It seems like it's for developing for Apple and i-products more than anything. So, I dropped interest in learning about it. That, and that I read, that the next version was so different, there was no point in spending time on a version so new, yet already so obsolete.

Stop trying to shame us, when more of that responsibility falls on you. I would like to see a few ports, or things done a certain way, and while I may not be able to do it, I don't come in here and start blaming and shaming people for it.

Apple won't maintain a FreeBSD port, never. What you could expect from them is to upstream changes necessary for correct building and execution on FreeBSD.
And if hundreds would ask, Apple still wouldn't maintain a "FreeBSD port". That would require them to do/submit commits to FreeBSD's ports tree.
I was a bit curious why Apple wouldn't maintain a port. They have done a lot for FreeBSD's base. If I didn't see this response, I would say for the person who wants Swift to ask Apple to maintain a port, since he wants to see it there.

It seems to me that majortom needs to ask Apple to make Swift portable so it can be more easily maintained on FreeBSD's portstree.
 
What responsibility, exactly?
majortom 's responsibility to do something helpful to get Swift ported more than anyone else's. Instead of shaming and blaming, for something he wants, while others aren't very interested in it.

I have a hard time putting some thoughts into words. It's not really about responsibility even though I said that. It however, falls on someone who's interested more than on those who aren't. He's acting like, it's our responsibility, and not his, even though he's the one who wants it. I see Swift as too niche by being too Mac/i-product centric.

To tell you the truth, I wanted to see newer languages get ported, and it was over my head in comparison to understanding how simple programs get ported. Asking Apple to make Swift more portable, and for example being able to be built with bmake would do a lot. A maintainer's job would have to be made easier, and portability upstream could help a lot with that.

At one time, I thought Swift would be the thing to learn, and wanted to see it get ported. I thought Rust would be cool, and while it has strengths, it seems on another world than what FreeBSD is in. I still think Julia would be cool if it were brought back, even without Python support. Julia would be awesome, but I don't have the brainpower to learn such a thing. I would try again to learn Python, since it's there.

I at least looked at Julia's needs from a Makefile standpoint, and didn't get as far as only the fetch file and compile to breaking because it lacked dependencies. What the maintainer put together for the Makefile, it was more complex than what I was capable of doing.
 
I understand some of the enthusiasm, but I don't understand the approach of telling us it's our responsibility for something that you want. This is not helpful at all, and it's a bit rude.

I wanted to see Julia get ported, and I wasn't even able to jump on it to use it, when it was. I might have turned it on, and typed some example commands or math equations. I don't remember. It's cool, even if I can't make use of it, or hardly make use of it. It's like me going into a music store as a child, and touching a few piano keys.

Even if I were a master of using Julia, it's not my place to make demands.
 
Why are you blaming and shaming others, when you can't or don't want to do it yourself.
Please show me anywhere in my posts I asked anyone to do anything other than express interest if they had any. Also, please show me where I blame anyone for not doing something I want done?

I asked if there was any interest, and I got none, but did get lots of people complaining.
I've bought a book on Swift a long time ago. It seems like it's for developing for Apple and i-products more than anything. So, I dropped interest in learning about it. That, and that I read, that the next version was so different, there was no point in spending time on a version so new, yet already so obsolete.
Your information from “a long time ago” is clearly up to date. I am pretty sure that Prime Video’s core services written in Smoke, run on “Apple and i-products”, so you must be right.
Stop trying to shame us, when that responsibility falls on you. I would like to see a few ports, or things done a certain way, and while I may not be able to do it, I don't come in here and start blaming and shaming people for it.
Again, please quote from any of my posts, where I asked you, anyone on this forum, or for that matter anyone from any part of the part of the FreeBSD organization to do anything other than express interest.

Just a single line.
I was a bit curious why Apple wouldn't maintain a port. They have done a lot for FreeBSD's base. If I didn't see this response, I would say for the person who wants Swift to ask Apple to maintain a port, since he wants to see it there.
I would say the reason that Apple has not done a port to FreeBSD, is that almost no one here, or on the Swift Forum has expressed any interest in having such a port. I came here to see if there was interest, and what I found was that there seems to be a great deal of hostility to even asking if people might want it.
It seems to me that majortom needs to ask Apple to make Swift portable so it can be more easily maintained on FreeBSD's portstree.
Swift is support on macOS, Linux (three distributions), Windows and Web Assembly. That seems pretty portable. Given that no one here seems to have any interest in it (“It’s only for i-products”, “I would recommend against any one using it”), why would anyone spend any energy doing anything to bring it to this platform?
 
I understand some of the enthusiasm, but I don't understand the approach of telling us it's our responsibility for something that you want. This is not helpful at all, and it's a bit rude.
Please show me where in any of my posts I said it was your responsibility to do anything other than express interest (or even lack there of) in having it on the platform.
Even if I were a master of using Julia, it's not my place to make demands.
What demands have I made? Are you even reading my actual posts or just parroting the comments of others?
 
You keep insisting and suggesting that it be ported. It's later you say Apple should port it.
When Apple first open sourced Swift, someone ported it to FreeBSD. Then,…. Crickets.
Unfortunately, still nothing on FreeBSD.
Just wondering if there is something, philosophical or otherwise preventing it from happening?
I am not asking why the previous port was removed, or how a new port would be maintained, or even if there are people who could port Swift themselves.
Maybe that would answer the previous question of yours.
Thanks for your help. I guess Linux it is.
You do not even express interest in using the result, just in suggesting that someone else do work.
Ports don't magically appear, and people politely told you, that someone has to do the work to have a port made. You don't want to hear how ports are made, yet someone has to know how ports are made for ports to be made. I'm sick of your attitude. Go use Linux. Please, do us a favor.

You got rude really quick, just because someone told you, that someone has to maintain a port for it to happen.

"I want a port, but I don't want to hear how ports are made." Ask Apple to maintain a port, as there are still 2 ways, Apple can maintain it, as someone said may not happen, or someone else can port it. You want it ported? Well, your attitude will be sure there's more resistance that wasn't there before. SOMEONE HAS TO DO IT, but you don't want to hear about it, or why it's no longer, which would be relevant to SOMEONE DOING IT. Ok, people here can ask Apple, but now there's less people who would ask, or in fact be more resistant to it.

Wanting Swift is one thing, your attitude is another.

I don't want to see Swift on FreeBSD just because of your terrible attitude. I had expression in seeing it in the past, and I was neutral afterwards by seeing how its utility was mostly for i-products and Mac.


I at least expressed interest in questioning whether Apple would port it. If Apple won't port it, and you insist it be ported. It needs a maintainer still. And the last time it was maintained, it was maintained by a volunteer. So the way to get it ported, if Apple doesn't, would be to start where their Makefile left off.
 
Back
Top