.NET 5.0 is going to be multiplatform[1] and getting alot more sugax-syntax[2]. Can't this make people pick C# rather D? I'm a bit afraid D can gain less popularity due to that or even make people switch. What are your thoughts on that?


[1]: https://devblogs.microsoft.com/dotnet/announcing-net-5-0/
[2]: https://devblogs.microsoft.com/dotnet/c-9-0-on-the-record/

On Thursday, 12 November 2020 at 15:57:44 UTC, Jack wrote:
> .NET 5.0 is going to be multiplatform[1] and getting alot more sugax-syntax[2]. Can't this make people pick C# rather D? I'm a bit afraid D can gain less popularity due to that or even make people switch. What are your thoughts on that?
>
>
> [1]: https://devblogs.microsoft.com/dotnet/announcing-net-5-0/
> [2]: https://devblogs.microsoft.com/dotnet/c-9-0-on-the-record/

1. Users themselves target .NET to their own proprietary system?
2. What is the license?
3. When you compile to native code, how much extra do you need to add in your distribution and what will the total size be.
4. Can you target bare metal targets?

On Thursday, 12 November 2020 at 16:04:21 UTC, IGotD- wrote:
> 1. Users themselves target .NET to their own proprietary system?

I fail to parse that question. I'm afraid. Do you mean to ask whether .NET users are porting the runtime to other operating systems themselves?

> 2. What is the license?

Most of .NET is licensed under MIT. It is my understanding that the entirety of the cross-platform code is released under that license. I do not know whether the Windows Desktop version of the runtime includes extra components that are still proprietary.

> 3. When you compile to native code, how much extra do you need to add in your distribution and what will the total size be.
> 4. Can you target bare metal targets?

Not yet. There's a toolchain in the making that has been demonstrated to compile small C# programs down to standalone executables with size less than 1MB. Going from there to bare metal should be quite doable, but footprint would very likely still be an issue. Unfortunately, I don't have a link to the source anymore.

I think most of that work is supposed to be formally released in .NET 6.0.

On Thursday, 12 November 2020 at 17:05:52 UTC, Gregor Mückl wrote:
>
> I fail to parse that question. I'm afraid. Do you mean to ask whether .NET users are porting the runtime to other operating systems themselves?
>

Yes that is what I meant and if .NET has a MIT license that helps.

>
> Most of .NET is licensed under MIT. It is my understanding that the entirety of the cross-platform code is released under that license. I do not know whether the Windows Desktop version of the runtime includes extra components that are still proprietary.
>

You have to be careful with this one. Let's say you some proprietary system which which is using .NET and is MIT. Then you have unittests that can run on your x86 Windows PC target and those link in proprietary MS components, then you must buy a license.

On Thursday, 12 November 2020 at 15:57:44 UTC, Jack wrote:
> .NET 5.0 is going to be multiplatform[1] and getting alot more sugax-syntax[2]. Can't this make people pick C# rather D? I'm a bit afraid D can gain less popularity due to that or even make people switch. What are your thoughts on that?
>
>
> [1]: https://devblogs.microsoft.com/dotnet/announcing-net-5-0/
> [2]: https://devblogs.microsoft.com/dotnet/c-9-0-on-the-record/

That only threat to D is itself.
There are features that need to finished and implemented.
There are language design issues that needs to be address.
The standard library needs a reviewer to prevent Pull requested from getting stale.
D needs to focus on its consumers instead of worrying other language development.

- Alex

On Thursday, 12 November 2020 at 18:08:48 UTC, IGotD- wrote:
> On Thursday, 12 November 2020 at 17:05:52 UTC, Gregor Mückl wrote:
>>
>> I fail to parse that question. I'm afraid. Do you mean to ask whether .NET users are porting the runtime to other operating systems themselves?
>>
>
> Yes that is what I meant and if .NET has a MIT license that helps.
>

Then again, the .NET CLR implementations that I know of are massive beasts. Porting them is no easy tasks and I am not aware of anyone who is doing that.

This is (part of?) the single static binary compilation project I was referring to earlier: https://github.com/dotnet/corert

I believe that an approach like that can be retargeted to different operating systems more easily.

>>
>> Most of .NET is licensed under MIT. It is my understanding that the entirety of the cross-platform code is released under that license. I do not know whether the Windows Desktop version of the runtime includes extra components that are still proprietary.
>>
>
> You have to be careful with this one. Let's say you some proprietary system which which is using .NET and is MIT. Then you have unittests that can run on your x86 Windows PC target and those link in proprietary MS components, then you must buy a license.

This is an overview over the licenses for the .NET 5.0 Windows Desktop Runtime:

https://github.com/dotnet/core/blob/master/license-information-windows.md

So this is quite a mix and the most worrying part are the two libraries under the Visual Studio License. My understanding is that these are not free (as in beer) unless you qualify for a Visual Studio Community license.

My understanding is that the Linux version is free of these constraints.

On Thursday, 12 November 2020 at 19:08:37 UTC, Gregor Mückl wrote:
>
> This is an overview over the licenses for the .NET 5.0 Windows Desktop Runtime:
>
> https://github.com/dotnet/core/blob/master/license-information-windows.md
>
> So this is quite a mix and the most worrying part are the two libraries under the Visual Studio License. My understanding is that these are not free (as in beer) unless you qualify for a Visual Studio Community license.
>
> My understanding is that the Linux version is free of these constraints.

The question is what "free" means here. If that part contains GPL or LGPL parts then it might be even more useless than a commercial license. You can always buy a commercial license if you want to and if you think the price is fair and then you can do what you want. GPL and LGPL puts restrictions on the distribution including your product. What is usually the most deterring is that you must open up your system for tampering which is a no no for many products.

D wins here as it is using the Boost license which is a huge benefit.

On Thursday, 12 November 2020 at 18:32:59 UTC, 12345swordy wrote:
> D needs to focus on its consumers instead of worrying other language development.

It's been said before and it'll be said again but it'd be nice to have more visible and more frequent (i.e. regular) communication about language direction and developments from the core maintainers. I feel like Atila is even less visible than Andrei was when everyone was complaining about this before.

On Thursday, 12 November 2020 at 19:24:17 UTC, IGotD- wrote:
> On Thursday, 12 November 2020 at 19:08:37 UTC, Gregor Mückl wrote:
>> [...]
>
> The question is what "free" means here. If that part contains GPL or LGPL parts then it might be even more useless than a commercial license. You can always buy a commercial license if you want to and if you think the price is fair and then you can do what you want. GPL and LGPL puts restrictions on the distribution including your product. What is usually the most deterring is that you must open up your system for tampering which is a no no for many products.
>
> D wins here as it is using the Boost license which is a huge benefit.

I wrote "free (as in beer)" for a reason above. The basic .NET framework components don't contain anything that is licensed under the GPL/LGPL. It's all open source, but not copyleft. The MIT license is very liberal. You only need to reproduce the license text somewhere in your documentation and you're good.

On Thursday, 12 November 2020 at 21:45:16 UTC, rinfz wrote:
> On Thursday, 12 November 2020 at 18:32:59 UTC, 12345swordy wrote:
>> D needs to focus on its consumers instead of worrying other language development.
>
> It's been said before and it'll be said again but it'd be nice to have more visible and more frequent (i.e. regular) communication about language direction and developments from the core maintainers. I feel like Atila is even less visible than Andrei was when everyone was complaining about this before.

A heads down working attitude is very commendable, but - watching from the sidelines, mind you - I get the impression that D could profit from some more aggressive outreach. That means not just talking to the existing community, but also more active advocacy. Reaching out to other communities and organizations, pointing out that D exists, listening to them to figure out their needs, pointing out what it is capable of doing for them etc... To be blunt: this part is sales, but different.

It takes a certain kind of character to do this. Technically minded people often tend to dislike doing that, in my experience*. But let's be honest, we all like to listen once someone has started a pitch in a way that made us genuinely curious about some aspect of the product.

I wonder now what would happen if the D Foundation could actually hire a person who can present such a pitch convincingly, especially to the corporate world.