July 20, 2022
On Wed, Jul 20, 2022 at 05:19:33PM +0000, ryuukk_ via Digitalmars-d wrote:
> On Wednesday, 20 July 2022 at 16:40:14 UTC, H. S. Teoh wrote:
> > > On Wednesday, 20 July 2022 at 14:31:49 UTC, ryuukk_ wrote:
> > > > https://github.com/carbon-language/carbon-lang/tree/trunk/docs/design#choice-types
> > > > > They got tagged union built in! as well as .Enum and pattern
> > > > > matching!
> > > > > What are we waiting for? We failing behind
> > [...]
> > 
> > We are waiting for highly-motivated and skilled people like you to submit DIPs and PRs to implement these features in D.
[...]
> Let's not suggest anything then, let's accept your faith and become irrelevant
> 
> I'll just use a different language, i'm fine with that personally, what ever empowers me, i'm only just a mere language user after all, who cares about the users when you have compiler developers telling language users to submit PRs to implement their suggestion ;)

I think you have a wrong perception of what's going on here.

I'm not a compiler developer, I'm also a mere language user, I just happened to have occasionally contributed some PRs, mainly to Phobos, a few to druntime and dmd (simple ones, mind you).  I'm not personally associated with the language maintainers, and I do not get paid by the D Foundation for my contributions; I'm not an employee of the Foundation, I'm just a volunteer.  So you're yelling at the wrong person here. :-D If anybody, you should be complaining rather to Walter.

This is an open source project primarily run by volunteers (perhaps with the exception of the language developers and some of the people paid by the D Foundation).  This isn't some commercial company that pays me to do work for them; so telling me what I should do with my free time is rather presumptuous, and quite insulting.  I have no obligations to serve you or anyone else on this forum, because you're not my customer -- I don't have any customers, I don't even run this show. You'll gain nothing by yelling at a fellow D user.

Besides, being a volunteer-run project, what matters isn't how loudly somebody demands for things to happen, what matters is who is willing to get their hands dirty and actually write the code to make it happen. If someone is not willing to do that, then the volume of their voice matters little in the grand scheme of things.  Time and again people have shown up on the forums demanding this or that, start a flamewar or two, but in the end, nothing came of it.  The advances that actually happened came from people who were willing stoop down and do the actual work necessary to make it happen. Quite often they were the ones who hardly said a thing in the forums because they were busy actually writing code instead of demanding others to do what they want.


T

-- 
Being able to learn is a great learning; being able to unlearn is a greater learning.
July 20, 2022

On Wednesday, 20 July 2022 at 17:46:15 UTC, Ola Fosheim Grøstad wrote:

>

Reminds me of how Google created Dart as dynamic language, then as an afterthought made a U-turn and redesigned it to become more like Java.

Can't blame them for that, it's the right thing to do, even javascript does it. Dynamic languages scale poorly beyond 100 LOC.

July 20, 2022
On Wednesday, 20 July 2022 at 18:01:09 UTC, H. S. Teoh wrote:
> On Wed, Jul 20, 2022 at 05:19:33PM +0000, ryuukk_ via Digitalmars-d wrote:
>> On Wednesday, 20 July 2022 at 16:40:14 UTC, H. S. Teoh wrote:
>> > > On Wednesday, 20 July 2022 at 14:31:49 UTC, ryuukk_ wrote:
>> > > > https://github.com/carbon-language/carbon-lang/tree/trunk/docs/design#choice-types
>> > > > > They got tagged union built in! as well as .Enum and pattern
>> > > > > matching!
>> > > > > What are we waiting for? We failing behind
>> > [...]
>> > 
>> > We are waiting for highly-motivated and skilled people like you to submit DIPs and PRs to implement these features in D.
> [...]
>> Let's not suggest anything then, let's accept your faith and become irrelevant
>> 
>> I'll just use a different language, i'm fine with that personally, what ever empowers me, i'm only just a mere language user after all, who cares about the users when you have compiler developers telling language users to submit PRs to implement their suggestion ;)
>
> I think you have a wrong perception of what's going on here.
>
> I'm not a compiler developer, I'm also a mere language user, I just happened to have occasionally contributed some PRs, mainly to Phobos, a few to druntime and dmd (simple ones, mind you).  I'm not personally associated with the language maintainers, and I do not get paid by the D Foundation for my contributions; I'm not an employee of the Foundation, I'm just a volunteer.  So you're yelling at the wrong person here. :-D If anybody, you should be complaining rather to Walter.
>
> This is an open source project primarily run by volunteers (perhaps with the exception of the language developers and some of the people paid by the D Foundation).  This isn't some commercial company that pays me to do work for them; so telling me what I should do with my free time is rather presumptuous, and quite insulting.  I have no obligations to serve you or anyone else on this forum, because you're not my customer -- I don't have any customers, I don't even run this show. You'll gain nothing by yelling at a fellow D user.
>
> Besides, being a volunteer-run project, what matters isn't how loudly somebody demands for things to happen, what matters is who is willing to get their hands dirty and actually write the code to make it happen. If someone is not willing to do that, then the volume of their voice matters little in the grand scheme of things.  Time and again people have shown up on the forums demanding this or that, start a flamewar or two, but in the end, nothing came of it.  The advances that actually happened came from people who were willing stoop down and do the actual work necessary to make it happen. Quite often they were the ones who hardly said a thing in the forums because they were busy actually writing code instead of demanding others to do what they want.
>
>
> T

I think you are the one having the wrong perception of what's going on

If nobody suggest anything, if nobody tells what they need, then how do you expect someone with the skills to step in and say "Hey i know how to do  that, i can maybe offer a hand"

I also volunteer, and i also sent various PRs to the language server, i don't brag about it, that is not point of my posts

Without demand, there can't be offers, maybe the way i phrase them is poor, i'll make an effort about it

But if everyone turns down every suggestions/discussions how can you create momentum?

Not everything needs to start with a PR or DIPs, sometimes it is someone fed up with how he has to write code in a certain way and wishes feature X existed so he let people know

Telling people to stfu because we don't have $$$ or they should do it themselves doesn't sound positive, and kills discussions/hype/momentum
July 20, 2022

On Wednesday, 20 July 2022 at 18:13:44 UTC, Kagamin wrote:

>

Can't blame them for that, it's the right thing to do, even javascript does it. Dynamic languages scale poorly beyond 100 LOC.

Can't blame them for not being focused and making a U-turn? It wasn't a small scale hobby project. Do you know how many people worked on it? :-)

July 20, 2022
On Wed, Jul 20, 2022 at 06:15:22PM +0000, ryuukk_ via Digitalmars-d wrote: [...]
> I think you are the one having the wrong perception of what's going on
> 
> If nobody suggest anything, if nobody tells what they need, then how do you expect someone with the skills to step in and say "Hey i know how to do that, i can maybe offer a hand"

That's fine.  But if nobody steps up to do it, then what?


> I also volunteer, and i also sent various PRs to the language server, i don't brag about it, that is not point of my posts
> 
> Without demand, there can't be offers, maybe the way i phrase them is poor, i'll make an effort about it
> 
> But if everyone turns down every suggestions/discussions how can you create momentum?

I didn't turn down your suggestion.  That would imply that I have some kind of decision-making power over the language, which I don't.  The decision-making is done by Walter and Atila, I have no say over that.


> Not everything needs to start with a PR or DIPs, sometimes it is someone fed up with how he has to write code in a certain way and wishes feature X existed so he let people know

That's fine too.  But it requires someone to pick up the tab and actually implement it, otherwise nothing will change.


> Telling people to stfu because we don't have $$$ or they should do it themselves doesn't sound positive, and kills discussions/hype/momentum

I didn't tell you to STFU.  Maybe something got lost in the ether, sometimes forum posts convey words but not the expression. All happened was, you asked "what are we waiting for" to implement feature XYZ, and I replied to your question of what we're waiting for: i.e., what must happen before we will have feature XYZ. That is, somebody needs to sit down and implement feature XYZ. Until that happens, feature XYZ won't exist in D.  It's as simple as that.  You can feel free to continue to make suggestions, complain about the lack of XYZ, etc., which is fine, I'm not going to stop you. (I *can't* stop you, even if I wanted to, anyway.)  But feature XYZ isn't going to suddenly drop down from the sky all implemented and ready to go until somebody actually sits down and does the work.


T

-- 
Question authority. Don't ask why, just do it.
July 21, 2022

On Wednesday, 20 July 2022 at 18:20:17 UTC, Ola Fosheim Grøstad wrote:

>

Can't blame them for not being focused and making a U-turn?

They had a mistake and fixed it, happens all the time.

July 21, 2022

On Wednesday, 20 July 2022 at 18:15:22 UTC, ryuukk_ wrote:

>

But if everyone turns down every suggestions/discussions how can you create momentum?

What is a needed is a plan with dependencies mapped out and priorities assigned, and focused process, but the D evolution history is more impulse driven and the planning aspect has been vague… this is the most limiting factor at this point.

In contrast Carbon has one, and primarily one, huge advantage: they have a big internal business critical use case in Google that will drive new features and tooling. Management will get resources allocated to improve on the weak language spots as they start to use Carbon internally (features driven by internal demand). As such the evolution cannot follow impulses, but has to be driven by plans (which will ensure steady progress).

It remains to be seen whether Carbon will be useful for small projects or end up like Ada: too tedious to be used for small and medium sized projects?

It is very difficult to tell at this point, but there is some arrogance in claiming to be a "C++ successor" at V0.1 and then push a somewhat flawed ML/Rust/TypeScript mashup syntax and less flexible semantics than C++.

One big future problem for D is that Carbon has this huge internal use case and most likely will get solid open source tooling funded by Google and a bunch of tutorials written by self-promoting bloggers.

Today we have the situation that people look at C++, finds it overwhelming, looks at Rust, finds it difficult to get into, then looks at niche alternatives (Zig, D etc). With 3 major system level languages I think many devs will stop looking further when they have looked at the three major contenders and just pick the one they find easier to deal with.

This might be a good time to consider a D3 move.

July 21, 2022

On Thursday, 21 July 2022 at 08:15:56 UTC, Kagamin wrote:

>

On Wednesday, 20 July 2022 at 18:20:17 UTC, Ola Fosheim Grøstad wrote:

>

Can't blame them for not being focused and making a U-turn?

They had a mistake and fixed it, happens all the time.

No, they didn't make a mistake, it was designed as a JavaScript replacement. They had an experienced language designer, but put in so many mainstream requirements that he felt that the resulting language was watered down and "bland". Then internal demands for something Java-like (I think they used Dart for their Google Ads software) forced it in a more Java-like direction. There is really no other reason for Dart to exist today than Google's internal use and Flutter. It is a "me too" language with no real identity.

July 21, 2022

On Thursday, 21 July 2022 at 11:06:37 UTC, Ola Fosheim Grøstad wrote:

>

No, they didn't make a mistake, it was designed as a JavaScript replacement.

My understanding is that it was designed as a language for webapps. Simply replacing JavaScript isn't much of a goal.

July 21, 2022

On Thursday, 21 July 2022 at 12:16:31 UTC, Kagamin wrote:

>

My understanding is that it was designed as a language for webapps. Simply replacing JavaScript isn't much of a goal.

The Dart VM JIT was supposed to go into the Chrome executable as a more performant VM JIT than the one for JavaScript, but it was dropped. The people who designed Dart also worked on StrongTalk, a faster SmallTalk variant, and that VM design was later turned into HotSpot for Java.
So the Dart designers had knowledge of how language design affects VM/JIT performance.

(Dart also compiles to JavaScript as a cross-platform feature, but that was not the original primary use case.)