So D will have LTS or not?

[Martyn]

On Monday, 3 July 2023 at 07:07:13 UTC, linger wrote:
> Looks like The D Language Foundation treat D as a toy 
> language(or only a language not a production), and the members 
> don't care user experience LSP/documnet/vscode/github which 
> users really care. thanks @webfreak did this.
>
> DLF more like said you can do it just do it, if you can't we 
> don't. but most of us are endpoint user not a language 
> developer.
>
> https://forum.dlang.org/post/u6jjkn$2e7b$1@digitalmars.com
>
> https://forum.dlang.org/thread/avvmlvjmvdniwwxemcqu@forum.dlang.org

I am not sure if we need another thread on this as the previous 
one is still rather active.

As a regular reader of the forums I have posted little. Recently 
I shared my views supporting for an LTS cycle.

I am not the best person to present examples. I am simply sharing 
my support for an LTS. I think the original poster of "D has 
become unbearable.." made valid points. The recent one from 
Z3Solver is a good read.

For me, as someone who has small "to-the-point" server programs 
in D, I can only share my concerns when wanting to use D for big 
projects that requires a combination of libraries with DUB like 
Gtk, Zeromq/Nanomsg, JSON serializer, Database (postgres), Web 
(Vibe.d) etc. I see nothing more than compilation hell as each 
library release targeting different compiler releases. When each 
release can be breaking changes, it is not unreasonable for me to 
have this concern.

*It is also worth noting posts about people wanting to use D for 
projects work are told to avoid doing so. One of these is one of 
my posts in 2020*

Some references: -
https://forum.dlang.org/thread/cffdqyvthqvurbcpxvbo@forum.dlang.org
https://forum.dlang.org/post/kfknqdjfkjbljhqrjarp@forum.dlang.org


Thing is, I do not see an LTS being created any time soon. It is 
more than to simply `git branch..` and here ya go! It will 
require a few people to maintain it. On top of this, it will be a 
cultural change in the D ecosystem especially for library 
developers to target target LTS, and to provide a version number 
standard to indicate which LTS it refers to.

    At least that is my opinion on the matter.


I want D to be successful. To me Dlang is 'the language' that 
ticks 90% of the boxes and easy to transition to being a C and C# 
developer. I am at an age, now, where I am not sure I want to 
learn Rust - as I hear is a steeper curve coming from those C\C++ 
background. I know I can do it, obviously - but I was hoping for 
D to be the preferred choice for me.


Rust is getting more and more momentum. Recently it has been 
included into the Linux kernel which, to me, has given it some 
extra brownie points in terms of free advertising. With Z3Solver 
recent comment about facebook ditching D and replaced with Rust 
does say a lot, along with his comments on crates not breaking 
(or worrying as much about breaking) if he upgrades, etc, comes 
back to my views on using D for larger projects.

There is a lot of talent in the D community and I plan to invest 
some time learning the internals of D, hoping I could be one of 
the LTS maintainers in the near future. Of course, it is very 
unlikely I can do this on my own... and won't get round to "being 
ready" for such a task anytime soon.