Could D have fit Microsoft's needs?

[Laeeth Isharc]

On Monday, 22 July 2019 at 23:16:26 UTC, Margo wrote:
> On Sunday, 21 July 2019 at 00:22:11 UTC, Walter Bright wrote:
>> On 7/18/2019 5:12 PM, Mike Franklin wrote:
>>> ... and D could complete better with Rust if it had 
>>> @safe-by-default and statically-check ownership/borrowing 
>>> mechanism as Walter recently proposed.
>>
>> D has some huge advantages over Rust.
>>
>> For example, D has a familiar syntax and jargon. For another, 
>> you'll be able to move to memory safety incrementally with D, 
>> you won't have to rewrite your app from the start.
>
> If you can not convince a company like Jetbrain, to support a 
> official D language plugin, then those few advantages do not 
> rack up to much. Rust language plugins, be it vscode, jetbrain 
> products or other ides are well developed allowing for more 
> developers choice.

Not everyone uses an IDE.  It's obviously an impediment to 
adoption in the games industry judging by what Manu says, but 
there are other industries besides games and in some places 
people don't use IDEs much at all.

> There is a reason why some major companies look at Rust for 
> replacing C++ and not D. No matter how much advantages you 
> think D has, it can be said, that without the actual support 
> system to convince the companies and developers, D will simply 
> keep trampling on the same spot.

Why so concerned about major companies when small and medium 
sized enterprises are in many places (the US being one) 
responsible for the majority of job creation.

The Mittelstand in Germany has a bunch of world leading companies 
I have mostly never heard of.  I think it would be much more 
interesting to know what the impediments are for them to adopt D 
than large companies.  The moment you have a committee involved 
formed of people who are not owners then the facts become much 
less important than social perceptions.  So why bother playing 
that game when you don't have to.

> Recently i needed some specific crates for a Rust project, i 
> looked over at D. More then half the stuff simply did not even 
> exist for D.
>
> D package growth is like 1 per day, with Rust growing by 25 per 
> day. Its easy to see not just momentum diference but also how 
> easier it is to get going with a project in Rust ( and half a 
> dozen other languages ) compared to D.
>
> https://www.youtube.com/embed/TkNep5zHWNw?start=1496
>
> Shows the exact reason why D has issues. Stand alone D is 
> usable but the moment you need specific packages or tools, D 
> can not be considered a main application language.

Using DPP you have all of C libraries and progressively the C++ 
capabilities are being extended although it's not there yet.

D can easily call python and Lua libraries.  I'm working on C# 
though it's not the number one priority for me right now.


> And no, we do not have the time to write packages for D, when 
> we can find perfectly good quality and supported packages for 
> Rust ( and half a dozen other GC languages! ).
>

Well then maybe you shouldn't use D if that's your own calculus.  
Others might have a different calculus.  My own experience has 
been that wrapping libraries takes time but it's a rounding error 
you can amortise in my context.

The world is a very big place and it's amazing how different 
situations can be.

> Stop confusing your language with its eco system. Nobody gives 
> a darn about a language if everything around it is limited and 
> frustrating. Its like having a 15" dong but having the face of 
> a mull. Sure, you can be proud of that one thing but its not 
> going to get you the girls ( or guys, we do not discriminate :) 
> ) if the rest is lacking.

That's the thing - there's tremendous variation in what's 
important to people or even what they find appealing and not.

I think whether D is right for you also depends on time 
preference of the people involved and the culture.  I don't think 
it's a coincidence that D is surprisingly popular in Germany and 
Japan and is less appealing to Americans than one might have 
guessed. Academic work finds quite a big difference in time 
preference between the US and Germany.

If you decide to use D, better be prepared for discomfort and 
suffering and the pain is front-loaded too.  That may or may not 
be a problem and different people and groups within organisations 
will have a different degree of pain tolerance.  My own view is 
it's just pain and discomfort and to try to focus on enterprise 
value added, and from that point of view these things are 
relevant but not that important.


> In other words: Stop being so darn proud of the language and 
> realize that the language without the rest is simply not 
> accepted for mainstream support. Its 2019 and beyond a few 
> companies and some side project, nobody really gives a darn 
> about D. Rust simply moves faster, has a better Eco system, 
> better support, ... So unless you plan on trowing a truckload 
> of money into hiring developers to speed up the development of 
> those issues point, your simply no match for Rust.

Why is mainstream adoption so important to you ?  And I wonder if 
you have considered that the only way to change something about 
the world that bugs you is to act.  Complaining by itself, I 
wonder how you think that will change anything.

> Its frankly pathetic whenever a popular blog comes up about 
> Rust as a potential choice for a big company, that people here 
> think "but we are also a choice, why does nobody really look at 
> us".

Big companies again.  Can you explain in logical chain of 
causation terms why people should be particularly concerned about 
big companies?  It reminds me of Americans planning to sell 
washing machines to China a few years back.  But if we could get 
just 2% share of this enormous market...  Yeah but you are 
thinking about it all wrong.

That's simply not the way waves of the adoption of technologies 
tend to unfold and it definitely isn't going to be the story for 
D.

> Walter, you remind me why some companies never become a 
> success. Great ideas, genius, hard working but geniuses do not 
> become fortune 500's without the marketing sales pitch to 
> convince people to put the money into your company, for a 
> (potential) payout years down the line.

Study the history of the German Mittelstand.  It's a much bigger 
world than just Anglo culture and in other places things can work 
quite differently.  Marketing in particular is also in flux 
because people wised up.  And btw plenty of companies are very 
successful not doing any marketing at all.  It depends on your 
context.  High trust cultures with a high future time orientation 
can work a bit differently.

Have you ever started a company and sold it for a decent result?  
You might remember when speaking to Walter that he has, competing 
for years as a one man band with Microsoft and eventually being 
acquired.

I think myself that yes dub could be improved and the registry 
also.  Maybe we can make it an SAOC project.   And that a start 
has been made on getting the merits of the language across and 
that from here what's necessary is just to build on that 
beginning.

Success stories are the best form of marketing really I think.  
That and just educational posts and people quietly doing great 
work.

Seb Wilzbach has achieved by far more in that respect than the 
total of all the naysayers and people complaining laid end to end.

If you want to change reality then unless you have a very unusual 
mind with insights people are ready to hear,then nothing is going 
to happen unless you take action.  And if one isn't interested in 
doing so,I wonder why it is that there would be an expectation 
that there will be a receptiveness to words.

Without skin in the game, talk by itself is cheap.