www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - ISO/IEC standard?

reply Rusty <me rustyx.org> writes:
I've heard many people complain that D is a privately owned 
language (which in my opinion hurts its adoption).
I do not share the same feeling, but was wondering how open the D 
language really is?
Are there any plans to have the D language standardized by 
ISO/IEC?
May 15 2016
next sibling parent reply Ola Fosheim =?UTF-8?B?R3LDuHN0YWQ=?= writes:
On Sunday, 15 May 2016 at 12:41:31 UTC, Rusty wrote:

 Are there any plans to have the D language standardized by 
 ISO/IEC?


ISO is for international industry standards and each nation has a 
representative. That is highly unlikely to ever happen for D.

You could probably get it into ECMA if you worked real hard for 
it, but I doubt Walter would be willing to give up control over 
the design.
May 15 2016
parent reply Rusty <me rustyx.org> writes:
On Sunday, 15 May 2016 at 12:45:12 UTC, Ola Fosheim Grøstad wrote:

 ISO is for international industry standards and each nation has 
 a representative. That is highly unlikely to ever happen for D.


OK but why is it unlikely? :)
May 15 2016
parent Ola Fosheim =?UTF-8?B?R3LDuHN0YWQ=?= writes:
On Sunday, 15 May 2016 at 12:50:15 UTC, Rusty wrote:

 OK but why is it unlikely? :)


Because you have to convince the national bodies that D qualifies 
as an international standard and that businesses around the globe 
will benefit from standardization?

Here are two death traps:

1. D does not have a formal specification.

2. ISO standardization is a global expert driven and consensus 
based process.

Even if D passed point 1), I somehow doubt that it ever would be 
able to move past point 2).

Keep in mind that the national bodies can provide their own 
expert representative. So, the design of D would have to pass a 
panel of experts.

C/C++ were defacto industry standards which made it much easier 
to get a free pass on poor design decisions.
May 15 2016
prev sibling next sibling parent rikki cattermole <rikki cattermole.co.nz> writes:
On 16/05/2016 12:41 AM, Rusty wrote:

 I've heard many people complain that D is a privately owned language
 (which in my opinion hurts its adoption).
 I do not share the same feeling, but was wondering how open the D
 language really is?
 Are there any plans to have the D language standardized by ISO/IEC?


If we stay still for like 5-10 years, I'm sure we can do it.

D is as open as it gets already, want to make your own implementation? 
Go right ahead!
As long as you don't say you're the reference compiler aka the one and 
only that matters, you should be ok :)
May 15 2016
prev sibling next sibling parent reply Russel Winder via Digitalmars-d <digitalmars-d puremagic.com> writes:
On Sun, 2016-05-15 at 12:41 +0000, Rusty via Digitalmars-d wrote:

 I've heard many people complain that D is a privately owned=C2=A0
 language (which in my opinion hurts its adoption).


These people are lying, D is not proprietary. One part of one of the
realizations of a D compiler might have licence restrictions, but the
language is free as are two of the realizations of a D compiler.


there would be a problem, but it isn't it is a FOSS language like Java,

whole ISO standard system is as much a millstone to progress as it is a
guarantee of consistency.


 I do not share the same feeling, but was wondering how open the D=C2=A0
 language really is?


Open is not a scale it is a Boolean. A language is either open or not:
if you make your own implementation will the lawyers come to sue you? D
is open in that you can submit change proposals or fork it for your
own. It is also versioned so there is public consistency for backward
compatibility.


 Are there any plans to have the D language standardized by=C2=A0
 ISO/IEC?


The only real purpose of a standard is to ensure consistency between
multiple realizations of a thing. This usually means proprietary
realizations of a thing so that comparison is impossible. Hence the
Java idea of a TCK to be awarded the label of Java to something
claiming to be Java in the proprietary days. Now the test suite is FOSS
anyway can create a Java realization so no need for a standard.

Just because C, C++ and Fortran have ISO (or any other) standards
doesn't mean any other language should. Python isn't standardized, yet
it has traction in the market. Ditto Groovy. Will Go, Rust, etc. get
ISO standards? Who know, but highly unlikely I suspect.


--=20
Russel.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D
Dr Russel Winder      t: +44 20 7585 2200   voip: sip:russel.winder ekiga.n=
et
41 Buckmaster Road    m: +44 7770 465 077   xmpp: russel winder.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder
May 15 2016
next sibling parent Ola Fosheim =?UTF-8?B?R3LDuHN0YWQ=?= writes:
On Sunday, 15 May 2016 at 14:27:21 UTC, Russel Winder wrote:

 Just because C, C++ and Fortran have ISO (or any other) 
 standards doesn't mean any other language should. Python isn't 
 standardized, yet it has traction in the market. Ditto Groovy. 
 Will Go, Rust, etc. get ISO standards? Who know, but highly 
 unlikely I suspect.


Dart is on an ECMA track which makes it possible later on to get 
on a fast track for ISO AFAIK, but it is probably mostly for 
political reasons (e.g. at some point they probably hoped it 
would replace ECMAScript).

Standardization makes a difference for being selected in big 
government projects, but... eh. Who cares?
May 15 2016
prev sibling parent reply Rusty <me rustyx.org> writes:
On Sunday, 15 May 2016 at 14:27:21 UTC, Russel Winder wrote:


 Open is not a scale it is a Boolean. A language is either open 
 or not


I'm not talking about "open" as in FOSS reference implementation. 
Rather, If I want to add a feature to the language, what are my 
chances of ever succeeding?
In case of C++, if a feature is really useful and gains traction 
among the users, there is the chance that national bodies will 
move and (over the course of a decade) it'll make it into the 
standard.
In case of Swift - forget about it.
In case of D, I really don't know how much 
corporate/political/personal influence is involved. There is 
always some, to some degree, but how much - is the question.

Why do people feel it is risky to develop a large project in D 
but not in C++?

Is there a roadmap or any kind of long-term plan?
May 16 2016
parent Russel Winder via Digitalmars-d <digitalmars-d puremagic.com> writes:
On Mon, 2016-05-16 at 21:18 +0000, Rusty via Digitalmars-d wrote:

 [=E2=80=A6]
=20
 Why do people feel it is risky to develop a large project in D=C2=A0
 but not in C++?
=20


Risk aversion and avoidance, conservatism, size of extant installed
base of application in the programming language, lack of programmer
pool, lack of thinking, "do what we did last time" attitude, sometimes
lack of bindings to dependencies, lack of sizable training content,
lack of courses in training company portfolio, blind panic about
garbage collection, lack of ambition, fear of being different,
assumption that you have to do things in C but C++ is acceptable.

I think that main training companies offer C, C++, Go, and even Rust
training but not D training is actually a very big issue.

[=E2=80=A6]

--
Russel.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D
Dr Russel Winder      t: +44 20 7585 2200   voip: sip:russel.winder ekiga.n=
et
41 Buckmaster Road    m: +44 7770 465 077   xmpp: russel winder.org.uk
London SW11 1EN, UK   w: www.russel.org.uk  skype: russel_winder
May 17 2016
prev sibling parent Bastiaan Veelo <Bastiaan Veelo.net> writes:
On Sunday, 15 May 2016 at 12:41:31 UTC, Rusty wrote:

 Are there any plans to have the D language standardized by 
 ISO/IEC?


ISO is not everything.

I know of an engineering software company that started its 
business in the '80s based on Pascal (ISO 7185), because that was 
the language thought in university. This language has its 
limitations, so when Extended Pascal was standardised in the '90s 
(ISO 10206) this company decided to change to it, anticipating 
abundant support. Yet the world has only seen one single 
commercial Extended Pascal compiler (Prospero) and it is no 
longer available today. The free compilers GPC en FPC support 
only some parts of the language, and not the same parts.

Despite being standardised, Pascal is known for its many 
incompatible dialects. One evolved into Delphi, which is still 
available today and proprietary.

So, standardising a programming language is not guaranteed to do 
any good.

-Bastiaan.
May 15 2016