On Sunday, 17 November 2024 at 09:47:23 UTC, Adam Wilson wrote:

It would be for personal projects where it's just me and nobody else. I can't tell others to invest a bunch of (very scarce) time in a language unless they'll be able to use it in the future. Moreover, we sometimes have to return to the code in five years, and the profession is moving to a requirement to share our code publicly, so the time horizon is very long.

I'm not going to start yet another discussion of all of the related concerns. It's just not an option to use a language professionally when it's reached the point that the lack of manpower allows something like this to happen.

That's what has me worried. The volunteer hours dry up, releases slip, and then a year or two later the project is on life support.

On Sunday, 17 November 2024 at 10:14:02 UTC, Sergey wrote:

Julia has the same problem it's always had. Otherwise I'd have been using it for a long time. It's too easy to write slow code and too complex/distracting to fix it. I've long viewed it as a fundamentally flawed language. However, it's what I'm moving to for some projects, because it has the community you mention and is so easy to get up and running.

Other languages are in principle good alternatives. Swift, Kotlin, and Go would require me to create my own ecosystem. I don't want to do that. F# is about the only language that's fun, doesn't suffer from performance problems, and has a good ecosystem in place, but it comes with a heck of a learning curve. Rust is so far from a sane choice that it's not worth talking about.

On Sunday, 17 November 2024 at 19:57:34 UTC, Lance Bachmeier wrote:

I think it's disingenuous to treat a lack of release as a lack of progress. As it so happens in this case, it's the fact we rely currently on one person to do the release, and it's a very non-automated process. That is going to be remedied.

It's very obviously not on life support, just look at the commit activity here: https://github.com/dlang

While you are free to draw your own conclusions however you want, I think either you are not looking at the right metrics, or you have some other reason to be leaving D.

-Steve

On Sunday, 17 November 2024 at 22:24:03 UTC, Steven Schveighoffer wrote:

That's not what I'm doing. The exact phrase was "it's reached the point that the lack of manpower allows something like this to happen". We recently had discussions about Visual D and GtkD being dead. These are major projects and nobody's picked them up. Contributors have left and nobody's taken their place. Adam left to work on his fork. The list of problems could go on but I don't have time for that discussion. And now a release is nearly four months late.

I can't imagine anyone comparing the current situation to where things were six or eight years ago and not be deeply concerned.

I'm not leaving D. I've moved projects to other languages if others are involved. I cannot look at the current state of affairs and tell someone else they should use D.

On Monday, 18 November 2024 at 18:13:16 UTC, Lance Bachmeier wrote:

The reports of VisualD's death have been greatly exaggerated.

https://github.com/dlang/visuald/commits/master/

GtkD, I'm not sure about, I don't use it. Looks like it has activity about 5 months ago.

I don't know that "contributors have left" in the case of VisualD, but I can say that at least for my projects, none of them are "dead", even though in some cases I haven't updated them in years.

I've picked up several projects that people have left. raylib-d I just yesterday was probably the second binding to update to the new 5.5 release. mysql-native I also manage, but it has been languishing for a while with not much time for me to work on it. But it's pretty solid, I use it for one of may work projects.

A project being abandoned doesn't mean it's dead forever. Open source gives anyone the opportunity to pick it back up. And sometimes, things don't get updated because they just work.

Adam is still working on his projects, which are still written in D. Maybe at some point it will purposely break with upstream? I don't know. I tend to doubt it, but I could be wrong.

The release delay is not based on any kind of abandonment. I think there has been too much read into this, it's not a usual situation.

I am not deeply concerned. I'm actually pretty positive on the outlook.

D is somewhat special in that a very few amount of people can manage huge projects. It's also a curse that the ecosystem can change rapidly based on the whims of a few.

The compiler is going strong. The ecosystem is continually improving. ImportC will open up new doors that make things even easier. PhobosV3 looks like a good opportunity to clean up the library. I hope editions can be finalized soon, but when that happens, it will make things drastically easier to change.

D is one of 9 supported languages in gcc. LDC gives world class optimization support for bleeding-edge D code, as it is released soon after the DMD release.

12 years ago I had to create an embedded server on a small ARM board. I chose C++ over D, because D just didn't have the capabilities.

Today, I would pick D without question. I know there are others who also are exploring this area.

I've been helping to develop a new GC for D, and in this, my usual mode of debugging using writeln statements just wouldn't cut it. I've been extremely impressed by the support of D in gdb.

Apologies for misreading here!

I am the opposite. I have taught a class of brand-new programmers to use D. I think there is potential for D with the right motivated people to be the go-to solution for an entire segment of programming. It just hasn't happened yet.

People join the Discord all the time, and say "how come I haven't heard of D before, it's amazing!"

I truly don't understand the doom-and-gloom sentiment from long-time D users, I see nothing but improvement.

Anyway, I get that these kinds of uncertainties can be demotivating, but it just bugs me when people post sad predictions of failure when everything usually has a reasonable explanation.

Don't get me wrong, there are some losses which are tough, I can count a few contributors that I am sad they have left, and I know they have left permanently. I have faith that the things they are no longer doing can be picked up by others.

I want to say something else, something I was just telling a group of students recently: when it comes to programming, learn whatever you want. If you want to learn D, learn D. If you want to learn python, learn python. If you want to learn brainfuck, you should think twice, but sure, go ahead! All of these skills will just add to your experience and your knowledge. I see a lot of fear of learning some language or framework or library because it won't further their career. I can't see how that's possible. Even if you don't end up using the language/framework/etc, the act of learning it opens new ideas and opportunities for growth!

-Steve