www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - Still not D standard yet ?

reply "Ledd" <a3826447 trbvm.com> writes:
I would like to know if something has changed in the plans for 
the future of D, because I really think it needs some kind of 
formalization and standardization to be a good investment, 
especially for medium/large/commercial projects .

Thanks.
Nov 26 2014
parent reply "Kagamin" <spam here.lot> writes:
What is missing?
Nov 28 2014
parent reply "Ledd" <a3826447 trbvm.com> writes:
On Friday, 28 November 2014 at 12:35:28 UTC, Kagamin wrote:

 What is missing?


an ISO standard ?
Nov 28 2014
parent reply Jonathan M Davis via Digitalmars-d-learn writes:
On Saturday, November 29, 2014 01:30:55 Ledd via Digitalmars-d-learn wrote:

 On Friday, 28 November 2014 at 12:35:28 UTC, Kagamin wrote:

 What is missing?


 an ISO standard ?


Someday, maybe, but most languages don't have an ISO standard, and I really
on't see what it would buy us. What we're generally missing most is
manpower. Putting a bunch of effort into formalizing it in a standard
wouldn't really help us. If anything, it would just take away manpower from
actually getting code written, getting bugs fixed, etc. And even if getting
an ISO standard for D were a goal, C++ was something like 20 years old
before it got an ISO standard, so even those languages that do have
standards didn't generally get them very early in their development, meaning
that we're not necessarily slow about getting a standard in comparison to
those languages that do.

- Jonathan M Davis
Nov 28 2014
parent reply "Ledd" <a3826447 trbvm.com> writes:
On Saturday, 29 November 2014 at 02:43:14 UTC, Jonathan M Davis 
via Digitalmars-d-learn wrote:

 On Saturday, November 29, 2014 01:30:55 Ledd via 
 Digitalmars-d-learn wrote:

 On Friday, 28 November 2014 at 12:35:28 UTC, Kagamin wrote:

 What is missing?


 an ISO standard ?


 Someday, maybe, but most languages don't have an ISO standard, 
 and I really
 on't see what it would buy us. What we're generally missing 
 most is
 manpower. Putting a bunch of effort into formalizing it in a 
 standard
 wouldn't really help us. If anything, it would just take away 
 manpower from
 actually getting code written, getting bugs fixed, etc. And 
 even if getting
 an ISO standard for D were a goal, C++ was something like 20 
 years old
 before it got an ISO standard, so even those languages that do 
 have
 standards didn't generally get them very early in their 
 development, meaning
 that we're not necessarily slow about getting a standard in 
 comparison to
 those languages that do.

 - Jonathan M Davis


It depends on what kind of languages you are talking about .

There are de facto standards that basically don't need any 
standard mostly because there isn't even a real competition so 
the users that want to code and solve a certain problem can't 
even look at real alternatives, for example what are the 
alternatives when it comes to Postscript or TeX/LaTeX ? They are 
basically de facto standards .

There are also languages that are linking their lifetime to a 
"main" language, for example languages that transcode to other 
languages don't really need a standard because they are just an 
extra layer on top of another language. There are examples of 
languages that have source-to-source compilers to C, Javascript 
and Lua for example .

Given the ambitions of D I can't see how you can pretend to be a 
relevant language without a standard, it also boils down to 
creating a reliable ecosystem and make a contract with the 
community. Do you think that this situation is doing any good to 
D ? For example there is a significant lack of tools in D where 
C/C++ have plenty of tools for anything since forever, especially 
in the last years with llvm .

Name even just 1 tool in D that is comparable with the 
counterpart in C/C++, or try to rate the ecosystem for D after 13 
years of existence .

Do you really think that a "system language", or just a language 
that aims to be popular, can possibly discard the idea of getting 
into an international standard ?

I still can't recall any major language that doesn't have a 
standard, what is the language/s you are thinking about ?
Nov 29 2014
parent reply "bachmeier" <no spam.net> writes:
On Saturday, 29 November 2014 at 09:03:13 UTC, Ledd wrote:

 Do you really think that a "system language", or just a 
 language that aims to be popular, can possibly discard the idea 
 of getting into an international standard ?

 I still can't recall any major language that doesn't have a 
 standard, what is the language/s you are thinking about ?


C was standardized in 1989. C++ was standardized in 1998. I'm 
unaware of ISO (or any other) standardization for Go, Python, 
Perl, Objective C, or PHP. And as I recall there is no ISO 


ISO standardization is very expensive. D is not even close to the 
popularity level necessary to begin thinking about that.
Nov 29 2014
next sibling parent reply Jonathan M Davis via Digitalmars-d-learn writes:
On Saturday, November 29, 2014 10:35:32 bachmeier via Digitalmars-d-learn wrote:

 On Saturday, 29 November 2014 at 09:03:13 UTC, Ledd wrote:

 Do you really think that a "system language", or just a
 language that aims to be popular, can possibly discard the idea
 of getting into an international standard ?

 I still can't recall any major language that doesn't have a
 standard, what is the language/s you are thinking about ?


 C was standardized in 1989. C++ was standardized in 1998. I'm
 unaware of ISO (or any other) standardization for Go, Python,
 Perl, Objective C, or PHP. And as I recall there is no ISO


 ISO standardization is very expensive. D is not even close to the
 popularity level necessary to begin thinking about that.


According to wikipedia, this is the list:

http://en.wikipedia.org/wiki/Category:Programming_languages_with_an_ISO_standard

There aren't very many on the list, and most of them aren't used in serious
production at this point. An ISO standard isn't even vaguely necessary for a
language to succeed. It might be nice to have, but it's not required. Maybe
someday, D will have an ISO standard, but at this point, our energies are
best directed elsewhere. And not all aspects of the language are set in
stone anyway. Yes, it's far more stable than it used to be, and we're trying
very hard to avoid breaking existing code, but some features may require
reworking to work the way we need them to, which could break code (e.g.
shared is on the list of things that we need to rework on some level).
Standarizing the language at this stage would harm those efforts, and we'd
end up with a worse language as a result.

- Jonathan M Davis
Nov 29 2014
parent "Ola Fosheim =?UTF-8?B?R3LDuHN0YWQi?= writes:
On Saturday, 29 November 2014 at 17:48:31 UTC, Jonathan M Davis 
via Digitalmars-d-learn wrote:

 On Saturday, November 29, 2014 10:35:32 bachmeier via 
 Digitalmars-d-learn wrote:

 C was standardized in 1989. C++ was standardized in 1998. I'm
 unaware of ISO (or any other) standardization for Go, Python,
 Perl, Objective C, or PHP.




C and C++ have improved by being faced with a standardization 
process, but they also had many implementations before they 
started. I don't think D qualifies for ISO standardization.

But Python is a bad example. Several incompatible versions of 
Python are being used, this is bad for the Python community. 
Python 2.7 carries so much weight today that the commercial 
sector are developing JITs for it, and ignoring Python 3… It 
might have been easier to move forward with formal 
standardization since that would have given Python 3 more weight.

Go is not standardized yet, probably because they aren't done?

Google did standardize Dart with ECMA, so they clearly see the 
value. The ECMAScript standardization has been very important IMO.


 production at this point. An ISO standard isn't even vaguely 
 necessary for a
 language to succeed. It might be nice to have, but it's not 
 required.


Standardization might be a requirement for use in larger 
governmental projects. Having a standard makes the language 
"electible" from an evaluation point of view.



 Standarizing the language at this stage would harm those 
 efforts, and we'd
 end up with a worse language as a result.


Yes, D is not ready for standardization, but a formal write up is 
needed.

Having to write up a formal specification will put light on 
special cases and inconsistencies, so having a formal write up is 
probably more important than formal standardization at this point.
Nov 29 2014
prev sibling parent ketmar via Digitalmars-d-learn <digitalmars-d-learn puremagic.com> writes:
On Sat, 29 Nov 2014 10:35:32 +0000
bachmeier via Digitalmars-d-learn <digitalmars-d-learn puremagic.com>
wrote:


 On Saturday, 29 November 2014 at 09:03:13 UTC, Ledd wrote:

 Do you really think that a "system language", or just a=20
 language that aims to be popular, can possibly discard the idea=20
 of getting into an international standard ?

 I still can't recall any major language that doesn't have a=20
 standard, what is the language/s you are thinking about ?


=20
 C was standardized in 1989. C++ was standardized in 1998.


and both "standards" sux. i won't buy anything that contains "undefined
behavior" in it as any kind of standard. this is pile of shit.

i mean, so-called "standardization by committee" does nothing valuable.
Nov 29 2014