Microsoft has published article about needing a safer system programming language.

https://msrc-blog.microsoft.com/2019/07/18/we-need-a-safer-systems-programming-language/

Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?

On Thursday, 18 July 2019 at 23:00:03 UTC, bauss wrote:
> Microsoft has published article about needing a safer system programming language.
>
> https://msrc-blog.microsoft.com/2019/07/18/we-need-a-safer-systems-programming-language/
>
> Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?

I knew this day would come, and while I hoped in my heart it would be D, I understand why it isn't.

I speculate this is an inflection point in the industry.  Walter himself even predicted it with his "memory safety will kill C" prediction at DConf a couple of years ago.  Others will likely follow, and I dread the day my employer comes to me and says "Do you know Rust?".

I think D could meet Microsoft's needs, but only if they forked it and made some fundamental changes to remove the technical debt, remove some of the "weird sh**" (https://youtu.be/TkNep5zHWNw?t=1378), and a number of other things we all could list to make using D a more professional experience.  The D Language Foundation is doing some of that with the copy constructors, the new object hierarchy, templating druntime, etc. but it all happens too slowly.  Regardless, I will continue with my part, at least for the time being, but it is an opportune time for some self-reflection.

Microsoft isn't the industry, so there could potentially be another major player that decides to use D (or fork it), and that could rescue the trend, but it seems unlikely to me, unfortunately.

It's sad news to me, really.  I wanted something like D to be the new trend, but Rust seemed to have its act together from the beginning.

Mike

On Thursday, 18 July 2019 at 23:49:00 UTC, Mike Franklin wrote:

> I think D could meet Microsoft's needs, but only if they forked it and made some fundamental changes to remove the technical debt, remove some of the "weird sh**" (https://youtu.be/TkNep5zHWNw?t=1378), and a number of other things we all could list to make using D a more professional experience.

... and D could complete better with Rust if it had @safe-by-default and statically-check ownership/borrowing mechanism as Walter recently proposed.

I think there is something that Microsoft is overlooking with D.  In the last 2 paragraphs of the blog post they say that programmers should be using managed languages like C#|F# whenever possible, but use Rust when the prior don't scale.

D is unique from Rust and C# in that is scales both up and down.  So, Microsoft is missing an opportunity to have one language for all use cases.  But I still think D has to do something about the technical debt and other "weird sh**" to make it viable.

Mike

On Fri, Jul 19, 2019 at 12:12:06AM +0000, Mike Franklin via Digitalmars-d wrote:
> On Thursday, 18 July 2019 at 23:49:00 UTC, Mike Franklin wrote:
> 
> > I think D could meet Microsoft's needs, but only if they forked it and made some fundamental changes to remove the technical debt, remove some of the "weird sh**" (https://youtu.be/TkNep5zHWNw?t=1378), and a number of other things we all could list to make using D a more professional experience.
[...]
> D is unique from Rust and C# in that is scales both up and down.  So, Microsoft is missing an opportunity to have one language for all use cases.  But I still think D has to do something about the technical debt and other "weird sh**" to make it viable.
[...]

Yeah, D has a really beautiful, marvelous core that works really well and addresses safety concerns, productivity, scalability, and meta-programming needs.  But outside this core is a bunch of "weird sh**", as the above video aptly puts it, of poorly-interacting corner cases, unexpected (often legacy) behaviours, and incompletely- or poorly- implemented features.  These technical debts greatly detract from D's overall value, yet there is no easy way to fix them without causing massive breakage of existing D code (or other D features).

It's sad, but it's reminiscient of the bad ole days when C++ shops would
dictate, via convention/policy to use only a certain, sane(r) subset of
C++ while avoiding the nastier parts.  The parts of the language to be
avoided is much smaller in D than in C++, but still, it's there, and
it detracts from the overall D experience.


T

-- 
What is Matter, what is Mind? Never Mind, it doesn't Matter.

On Thursday, 18 July 2019 at 23:00:03 UTC, bauss wrote:
> Microsoft has published article about needing a safer system programming language.
>
> https://msrc-blog.microsoft.com/2019/07/18/we-need-a-safer-systems-programming-language/
>
> Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?

I don't think D can compete on this with Rust. Yes, D has safety features built-in the language, but they are opt-in. Rust has the safety features opt-out. Now both approaches have their advantages of course. D is more pragmatic and lets you write code without fighting the borrow checker each step of the way.

But from the safety perspective, since Rust mandates safety, you can look at the codebase, and focus on the unsafe blocks for auditing. In case of D it wouldn't be that simple. Sure, you can @safe sections of code, but it's harder to find which sections are safe and which aren't compared to Rust where everything is safe by default and you look for the unsafe ones.

BetterC... it occupies that weird spot between C and D. I think it has some usecases (the no GC/runtime part works wonders for things like WebAssembly), but I don't think it's a production ready solution.

19.07.2019 4:12, H. S. Teoh пишет:
> On Fri, Jul 19, 2019 at 12:12:06AM +0000, Mike Franklin via Digitalmars-d wrote:
>> On Thursday, 18 July 2019 at 23:49:00 UTC, Mike Franklin wrote:
>>
>>> I think D could meet Microsoft's needs, but only if they forked it
>>> and made some fundamental changes to remove the technical debt,
>>> remove some of the "weird sh**"
>>> (https://youtu.be/TkNep5zHWNw?t=1378), and a number of other things
>>> we all could list to make using D a more professional experience.
> [...]
>> D is unique from Rust and C# in that is scales both up and down.  So,
>> Microsoft is missing an opportunity to have one language for all use
>> cases.  But I still think D has to do something about the technical
>> debt and other "weird sh**" to make it viable.
> [...]
> 
> Yeah, D has a really beautiful, marvelous core that works really well
> and addresses safety concerns, productivity, scalability, and
> meta-programming needs.  But outside this core is a bunch of "weird
> sh**", as the above video aptly puts it, of poorly-interacting corner
> cases, unexpected (often legacy) behaviours, and incompletely- or
> poorly- implemented features.  These technical debts greatly detract
> from D's overall value, yet there is no easy way to fix them without
> causing massive breakage of existing D code (or other D features).
> 
> It's sad, but it's reminiscient of the bad ole days when C++ shops would
> dictate, via convention/policy to use only a certain, sane(r) subset of
> C++ while avoiding the nastier parts.  The parts of the language to be
> avoided is much smaller in D than in C++, but still, it's there, and
> it detracts from the overall D experience.
> 
> 
> T
> 
Well, I wouldn't afraid anything while D has this great community, first of all. No worries! Everything has its advantages and disadvantages. I totally agree to Andrei that Rust has "a disharmonic personality" (https://www.quora.com/Which-language-has-the-brightest-future-in-replacement-of-C-between-D-Go-and-Rust-And-Why/answer/Andrei-Alexandrescu). For example you can read a post of one of active Rust user about Rust drawback(s) he has encountered in a real project: http://way-cooler.org/blog/2019/04/29/rewriting-way-cooler-in-c.html
So let's continue and everything will be ok!

On Fri, 2019-07-19 at 00:12 +0000, Mike Franklin via Digitalmars-d wrote:
> […]
> 
> I think there is something that Microsoft is overlooking with D. In the last 2 paragraphs of the blog post they say that programmers should be using managed languages like C#|F# whenever possible, but use Rust when the prior don't scale.
[…]

That is going to make the Java and JVM-based languages communities very happy indeed.

-- 
Russel.
===========================================
Dr Russel Winder      t: +44 20 7585 2200
41 Buckmaster Road    m: +44 7770 465 077
London SW11 1EN, UK   w: www.russel.org.uk

On Thursday, 18 July 2019 at 23:00:03 UTC, bauss wrote:
> ...
> Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?

Question: Have anyone from Dlang PR team tried to send Ryan Levick (Principal Cloud Developer Advocate) or Sebastian Fernandez (Security Software Engineer) an e-mail to talk about D?

I mean maybe "our" language is over their radar (Or not), but this could may increase the chances or at least give them a thought.

I pretty sure that should be nice or maybe engage some sort a relationship in the future.

Matheus.

On Friday, 19 July 2019 at 12:24:10 UTC, matheus wrote:
> On Thursday, 18 July 2019 at 23:00:03 UTC, bauss wrote:
>> ...
>> Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?
>
> Question: Have anyone from Dlang PR team tried to send Ryan Levick (Principal Cloud Developer Advocate) or Sebastian Fernandez (Security Software Engineer) an e-mail to talk about D?
>
> I mean maybe "our" language is over their radar (Or not), but this could may increase the chances or at least give them a thought.
>
> I pretty sure that should be nice or maybe engage some sort a relationship in the future.
>
> Matheus.

They'd never take a language that they don't dominate / create. If something new emerge this will come from their labs / experimental languages. Before .Net the situation was similar, i.e they could have chosen something like Delphi but created .Net in response.

To the best what could happen is that they borrow D features and everyone here then would get mad because of that...

On Friday, 19 July 2019 at 12:59:40 UTC, user1234 wrote:
> On Friday, 19 July 2019 at 12:24:10 UTC, matheus wrote:
>> On Thursday, 18 July 2019 at 23:00:03 UTC, bauss wrote:
>>> ...
>>> Could D (Most likely as betterC) have fit into that domain OR is it not yet safe?
>>
>> Question: Have anyone from Dlang PR team tried to send Ryan Levick (Principal Cloud Developer Advocate) or Sebastian Fernandez (Security Software Engineer) an e-mail to talk about D?
>>
>> I mean maybe "our" language is over their radar (Or not), but this could may increase the chances or at least give them a thought.
>>
>> I pretty sure that should be nice or maybe engage some sort a relationship in the future.
>>
>> Matheus.
>
> They'd never take a language that they don't dominate / create. If something new emerge this will come from their labs / experimental languages. Before .Net the situation was similar, i.e they could have chosen something like Delphi but created .Net in response.
>
> To the best what could happen is that they borrow D features and everyone here then would get mad because of that...

I'm not saying they would jump the ship right the way or use as it is.

But they are changing, just take a look over Microsoft embracing Chromium or adding Linux support on Windows.

And since they're even mentioning Rust, I think they may be open to D.

Matheus.