"Will Carbon Replace C++?" by Manuel Rubio

[Lynn McGuire]

"Will Carbon Replace C++?" by Manuel Rubio
https://semaphoreci.com/blog/carbon

"The last CppNorth 2022 was announced with Chandler Carruth scheduled to
give a keynote, where he showed the results of a new scientific
experiment. The keynote was titled: Carbon Language: An experimental
successor to C++. But why do we need a successor and where did this idea
come from?"

Here we go again.

Lynn

[Sam]

Lynn McGuire writes:

> "Will Carbon Replace C++?" by Manuel Rubio
> https://semaphoreci.com/blog/carbon

Carbon will replace C++ as soon as C++ gets replaced by Java, dot-Nyet,
Python, and Rust. Then it will be Carbon's turn.

[Mut...@dastardlyhq.com]

I'm fairly sure thats what Go was supposed to be. Presumably it'll soon be Gone.

[Öö Tiib]

Answer to the question is "Certainly No". There are better candidates
(like D, Rust and Swift). Meanwhile C++ is no way dying despite being
dangerously obese ... its popularity is high.

If a language tries to be as performant and to interoperate with C++
(so to gradually replace it) then it has first either to make compiler
frontend for some toolchain that compiles C++ or at least transpiler
to C++ (or some other language in such a toolchain). Without that the
talks can't be taken seriously.

As of syntax sugar ... the proposed improvements / changes
compared to C++ look about same or even less impressive than
others. So Carbon looks like a list of unfulfilled wishes and
buzzwords. Now if Google pushes it ... then it can perhaps reach
questionable state of Golang and Dart. If it does not then it will
remain a vapourware.

[Alf P. Steinbach]

With the NSA recommendation to not use C or C++ the succession might
happen swifter than one would think. Or not. Things are happening.

NSA, November 2022:
¹❝NSA advises organizations to consider making a strategic shift from
programming languages that provide little or no inherent memory
protection, such as C/C++, to a memory safe language when possible. Some
examples of memory safe languages are C#, Go, Java, Ruby™, and Swift®❞

The New Stack, February 2023 (linking to Bjarne's NSA response):
²❝In mid-January, the official C++ “direction group” — which makes
recommendations for the programming language’s evolution — issued a
statement addressing concerns about C++ safety. While many languages now
support “basic type safety” — that is, ensuring that variables access
only sections of memory that are clearly defined by their data types —
C++ has struggled to offer similar guarantees.

This new statement, co-authored by C++ creator Bjarne Stroustrup, now
appears to call for changing the C++ programming language itself to
address safety concerns. “We now support the idea that the changes for
safety need to be not just in tooling, but visible in the
language/compiler, and library.”

The group still also supports its long-preferred use of debugging tools
to ensure safety (and “pushing tooling to enable more global analysis in
identifying hard for humans to identify safety concerns”). But that
January statement emphasizes its recommendation for changes within C++.

Specifically, it proposes “packaging several features into profiles”
(with profiles defined later as “a collection of restrictions and
requirements that defines a property to be enforced” by, for example,
triggering an automatic analysis.) In this way the new changes for
safety “should be visible such that the Safe code section can be named
(possibly using profiles), and can mix with normal code.”❞

¹
https://media.defense.gov/2022/Nov/10/2003112742/-1/-1/0/CSI_SOFTWARE_MEMORY_SAFETY.PDF
²
https://thenewstack.io/can-c-be-saved-bjarne-stroustrup-on-ensuring-memory-safety/


- Alf

[Alf P. Steinbach]

On 2023-03-01 12:02 PM, Alf P. Steinbach wrote:
> On 2023-02-28 10:35 PM, Lynn McGuire wrote:
>> "Will Carbon Replace C++?" by Manuel Rubio
>>     https://semaphoreci.com/blog/carbon
>>
>> "The last CppNorth 2022 was announced with Chandler Carruth scheduled
>> to give a keynote, where he showed the results of a new scientific
>> experiment. The keynote was titled: Carbon Language: An experimental
>> successor to C++. But why do we need a successor and where did this
>> idea come from?"
>>
>> Here we go again.
>
> With the NSA recommendation to not use C or C++ the succession might
> happen swifter than one would think. Or not. Things are happening.
>
> NSA, November 2022:
> ¹❝NSA advises organizations to consider making a strategic shift from
> programming languages that provide little or no inherent memory
> protection, such as C/C++, to a memory safe language when possible. Some
> examples of memory safe languages are C#, Go, Java, Ruby™, and Swift®❞
>

[snip]
>
> ¹
> https://media.defense.gov/2022/Nov/10/2003112742/-1/-1/0/CSI_SOFTWARE_MEMORY_SAFETY.PDF

Sorry, the NSA apparently didn't use those words.

The quote is from Bjarne's response, where it appeared to be a quote of
the NSA article.

Either the NSA has changed the article, or else Bjarne needs to train on
how to present summaries and paraphrases so that they do not look
exactly like literal quotes.

But as a summary it's correct.

Bjarne's NSA response:
https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2023/p2739r0.pdf

- Alf

[Mut...@dastardlyhq.com]

It doesn't help that Carbon is also an Apple C API so Google clearly didn't
bother to check if the name was already used in the IT world.

[Mut...@dastardlyhq.com]

On Wed, 1 Mar 2023 12:02:16 +0100
"Alf P. Steinbach" <alf.p.s...@gmail.com> wrote:
>On 2023-02-28 10:35 PM, Lynn McGuire wrote:
>> "Will Carbon Replace C++?" by Manuel Rubio
>>    https://semaphoreci.com/blog/carbon
>>
>> "The last CppNorth 2022 was announced with Chandler Carruth scheduled to
>> give a keynote, where he showed the results of a new scientific
>> experiment. The keynote was titled: Carbon Language: An experimental
>> successor to C++. But why do we need a successor and where did this idea
>> come from?"
>>
>> Here we go again.
>
>With the NSA recommendation to not use C or C++ the succession might
>happen swifter than one would think. Or not. Things are happening.
>
>NSA, November 2022:
>¹❝NSA advises organizations to consider making a strategic shift from
>programming languages that provide little or no inherent memory
>protection, such as C/C++, to a memory safe language when possible. Some
>examples of memory safe languages are C#, Go, Java, Ruby™, and Swift®❞

Apples and oranges. The use case of C and up to a point C++ is rather different
to those languages. You'd think the NSA might know this.

>identifying hard for humans to identify safety concerns”). But that
>January statement emphasizes its recommendation for changes within C++.

Just leave C++ alone FFS.

[Bonita Montero]

Am 01.03.2023 um 10:44 schrieb Mut...@dastardlyhq.com:

> I'm fairly sure thats what Go was supposed to be. Presumably it'll soon be Gone.

Go wasn't designed as a C++-replacement but as a cloud-computing
language for the backend. Go has a performance similar to Java and
Java is on average half as fast as C++ in the computer languages
benchmark game.

[Bo Persson]

Not only that, it is really hard to Google for Carbon and find something
related. :-)

https://carbondesignsystem.com/
https://carbon.website/
https://carbon.nesbot.com/docs/
https://www.carbonads.net/
https://www.carbonbrief.org/
https://www.carbontrust.com/
https://www.carbonfootprint.com/
https://carbontracker.org/
https://carbon-pulse.com/
https://www.globalcarbonproject.org/
https://www.carbonclean.com/
https://carbonengineering.com/
https://www.getcarbon.co/
https://carbonmarketwatch.org/
https://www.carbonrecycling.is/
https://www.carbonneutral.com/
https://www.carbon-direct.com/
https://joincarbon.com/
https://carbonfund.org/
https://carbon.coop/
https://www.carbon.tv/

[Dan Cross]

In article <ttnbbo$3u603$1...@dont-email.me>,

Alf P. Steinbach <alf.p.s...@gmail.com> wrote:

>On 2023-02-28 10:35 PM, Lynn McGuire wrote:
>> "Will Carbon Replace C++?" by Manuel Rubio
>>    https://semaphoreci.com/blog/carbon
>>
>> "The last CppNorth 2022 was announced with Chandler Carruth scheduled to
>> give a keynote, where he showed the results of a new scientific
>> experiment. The keynote was titled: Carbon Language: An experimental
>> successor to C++. But why do we need a successor and where did this idea
>> come from?"
>>
>> Here we go again.
>
>With the NSA recommendation to not use C or C++ the succession might
>happen swifter than one would think. Or not. Things are happening.

Too bad Carbon isn't memory safe. :-/

- Dan C.

[Öö Tiib]

On Wednesday, 1 March 2023 at 13:02:33 UTC+2, Alf P. Steinbach wrote:
> On 2023-02-28 10:35 PM, Lynn McGuire wrote:
> > "Will Carbon Replace C++?" by Manuel Rubio
> > https://semaphoreci.com/blog/carbon
> >
> > "The last CppNorth 2022 was announced with Chandler Carruth scheduled to
> > give a keynote, where he showed the results of a new scientific
> > experiment. The keynote was titled: Carbon Language: An experimental
> > successor to C++. But why do we need a successor and where did this idea
> > come from?"
> >
> > Here we go again.
>
> With the NSA recommendation to not use C or C++ the succession might
> happen swifter than one would think. Or not. Things are happening.
>

Nice. NSA is evil organisation so its advise should be ignored but
good news are that Bjarne seems to realise that specifying language
to be "UB" multiple times per every page and then writing "core
guidelines" how to navigate in that forest of bear traps is not enough.

That is unlikely to make C++ worse or to help vapourware like Carbon.

[Vir Campestris]

Nothing is going to replace C++ - at least not completely.

Fortran and Cobol haven't gone away, and we all know C is pretty
healthy. I've written assembler for odd bits quite recently - and that's
pretty old. Heck, it's not that long since I've been forced to write
actual machine code...

It's possible that C++ will get less popular. It isn't going away.

Andy

[Daniel]

On Wednesday, March 1, 2023 at 9:30:36 AM UTC-5, Bo Persson wrote:

> Not only that, it is really hard to Google for Carbon and find something
> related. :-)
>

Try googling for carbon c++ :-)

Daniel

[Richard]

[Please do not mail me a copy of your followup]

Lynn McGuire <lynnmc...@gmail.com> spake the secret code
<ttls3k$3msuf$3...@dont-email.me> thusly:

>"Will Carbon Replace C++?" by Manuel Rubio

>[...]
>
>Here we go again.

Betteridge's law of headlines

Betteridge's law of headlines is an adage that states: "Any headline
that ends in a question mark can be answered by the word no."
<https://en.wikipedia.org/wiki/Betteridge%27s_law_of_headlines>
--
"The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline>
The Terminals Wiki <http://terminals-wiki.org>
The Computer Graphics Museum <http://computergraphicsmuseum.org>
Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

[Lynn McGuire]

I think that Fortran is in the process of converting from a development
language to a hobbyist language. The recent shutdown of most of the
Fortran Compiler companies after the freewareing of Intel Fortran was a
sign of the times.

Lynn

[Scott Lurndal]

Lynn McGuire <lynnmc...@gmail.com> writes:
>On 3/1/2023 10:50 AM, Vir Campestris wrote:
>> On 28/02/2023 21:35, Lynn McGuire wrote:
>>> "Will Carbon Replace C++?" by Manuel Rubio
>>>     https://semaphoreci.com/blog/carbon
>>>
>>> "The last CppNorth 2022 was announced with Chandler Carruth scheduled
>>> to give a keynote, where he showed the results of a new scientific
>>> experiment. The keynote was titled: Carbon Language: An experimental
>>> successor to C++. But why do we need a successor and where did this
>>> idea come from?"
>>>
>>> Here we go again.
>>>
>>> Lynn
>>
>> Nothing is going to replace C++ - at least not completely.
>>
>> Fortran and Cobol haven't gone away, and we all know C is pretty
>> healthy. I've written assembler for odd bits quite recently - and that's
>> pretty old. Heck, it's not that long since I've been forced to write
>> actual machine code...
>>
>> It's possible that C++ will get less popular. It isn't going away.
>>
>> Andy
>
>I think that Fortran is in the process of converting from a development
>language to a hobbyist language.

Now you're being absurd.

> The recent shutdown of most of the
>Fortran Compiler companies after the freewareing of Intel Fortran was a
>sign of the times.

I'd suggest that the existance of gcc and clang haven't relegated
C and C++ to "hobbyist languages". They have relegated a whole bunch
of commercial toolchains to niche or oblivion.

[Chris M. Thomasson]

On 3/1/2023 3:02 AM, Alf P. Steinbach wrote:
> On 2023-02-28 10:35 PM, Lynn McGuire wrote:
>> "Will Carbon Replace C++?" by Manuel Rubio
>>     https://semaphoreci.com/blog/carbon
>>
>> "The last CppNorth 2022 was announced with Chandler Carruth scheduled
>> to give a keynote, where he showed the results of a new scientific
>> experiment. The keynote was titled: Carbon Language: An experimental
>> successor to C++. But why do we need a successor and where did this
>> idea come from?"
>>
>> Here we go again.
>
> With the NSA recommendation to not use C or C++ the succession might
> happen swifter than one would think. Or not. Things are happening.

> [...]

Reminds me of people saying ADA is the only safe language.

[Lynn McGuire]

Dude, have you commercially worked with any computer software in the
last ten years ? Commercially means that you got paid by somebody for
your work.

Lynn

[David Brown]

I get paid for all my programming work - and I use gcc for the solid
majority of it. I used to make much more use of commercial compilers,
and for the most part, I am glad to be done with them. It is not a
matter of the price of the tools - we have on several occasions paid for
commercial packaging of gcc toolchains. It is simply that I find them
to be better tools for helping me do my job.

If you are talking about getting paid for work on compilers, then the
solid majority of gcc development (and presumably also clang) is done by
people who are paid to do the job.

[Scott Lurndal]

Yes, of course. And paid quite well. And for the last quarter
of a century, exclusively using gcc and sometimes clang or systemC.

[Mut...@dastardlyhq.com]

Every single company I've worked at both big and small that wrote software
for linux in C/C++ used gcc so I've no idea what that guy is talking about.
Seems to me he's the one who doesn't have any commercial experience, or if
he does its exclusively working in Windows.

[Lynn McGuire]

So none of these projects included Fortran ?

Lynn

[Keith Thompson]

Your assertion is that Fortran is largely a hobby language, not a
development language.

The rise of freeware Fortran compilers does not support that
assertion. C and C++ are mostly compiled with freeware compilers,
even by professional C and C++ programmers.

The fact that someone posting to a C++ newsgroup hasn't used
Fortran doesn't support that assertion. The sets of professional
C++ programmers and professional Fortran programmers are likely to
be mostly disjoint.

Taking a very quick look, I see Fortran questions on Stack Overflow
and moderate activity in comp.lang.fortran.

I believe that Fortran is still used professionally in some fields.
I don't have much direct evidence for that. What evidence do
you have?

--
Keith Thompson (The_Other_Keith) Keith.S.T...@gmail.com
Working, but not speaking, for XCOM Labs
void Void(void) { Void(); } /* The recursive call of the void */

[Scott Lurndal]

Would you find that surprising? Fortran has been niche for almost
four decades now and is still heavily used in certain fields.

[Malcolm McLean]

On Thursday, 2 March 2023 at 21:18:37 UTC, Keith Thompson wrote:
>
> I believe that Fortran is still used professionally in some fields.
> I don't have much direct evidence for that. What evidence do
> you have?
>

I did my PhD mostly in Fortran.
Whilst Fortran is frequently used for serious, non-hobby purposes, it
tends not to be used by professional programmers, but by scientists,
mathematicians, and engineers who do a certain amount of programming
as part of their job, but whose main expertise lies elsewhere.

[Chris M. Thomasson]

Windows has a fairly decent sub system.

https://learn.microsoft.com/en-us/windows/wsl/install

[Lynn McGuire]

A long time ago in a far away USA, Fortran IV (66) was used to build
operating systems for minicomputers such as Prime and DEC. Maybe
others, I don't know. After the Fortran 77 debacle where data
structures were not added to the language definition, Prime moved to the
PL/1 language for their operating system and system utilities like Edit.
I don't know what DEC did as their Fortran data structures were used
extensively throughout their VMS operating system and became a semi
industry standard.

We bought a Prime 450 in 1978 ??? and the source code came with it for
everything so we modified the system utilities of course. I added a
system command named POP written in Fortran that allowed you to jump to
the directory above your current directory. Otherwise, you had to type
in the entire directory path to move anywhere. Pre Unix and MS-DOS
directory structures sucked.

Lynn

[Scott Lurndal]

Lynn McGuire <lynnmc...@gmail.com> writes:
>On 3/2/2023 4:28 PM, Malcolm McLean wrote:
>> On Thursday, 2 March 2023 at 21:18:37 UTC, Keith Thompson wrote:
>>>
>>> I believe that Fortran is still used professionally in some fields.
>>> I don't have much direct evidence for that. What evidence do
>>> you have?
>>>
>> I did my PhD mostly in Fortran.
>> Whilst Fortran is frequently used for serious, non-hobby purposes, it
>> tends not to be used by professional programmers, but by scientists,
>> mathematicians, and engineers who do a certain amount of programming
>> as part of their job, but whose main expertise lies elsewhere.
>
>A long time ago in a far away USA, Fortran IV (66) was used to build
>operating systems for minicomputers such as Prime and DEC. Maybe
>others, I don't know. After the Fortran 77 debacle where data
>structures were not added to the language definition, Prime moved to the
>PL/1 language for their operating system and system utilities like Edit.
> I don't know what DEC did as their Fortran data structures were used
>extensively throughout their VMS operating system and became a semi
>industry standard.

I worked with the VMS sources for three years. There weren't any
"Fortran" data structures at all. The data structures were defined
in a language called MDL and there was a program to generate "header"
files for both BLISS-32 and MACRO-32 in which the bulk of the operating
system was written. MDL could also generate headers for C, Pascal,
Fortran, COBOL and probably PL/1 from the same description language.

In the first releases of VMS, most of the RSX-11 utilities were
used, being replace by native utilities in VMS2.0.

I'm not aware of any DEC operating systems written in Fortran; the
PDP-8 operating systems were written in various assemblers, the PDP-11
operating systems in Macro-16 and Bliss-16. PDP-10 TOPS-10 was written
in assembler and perhaps later in Bliss-10/36.

[David Brown]

My understanding (based on "things I read somewhere" rather than
personal experience) is that Fortran is very rarely the language of
choice for /new/ software, but is still very much in use in the
maintenance and further development of existing software. In its
heyday, it was the prime choice of language for engineering, scientific
and numeric work. (In its early days it was used for much more than
that, simply because there was little choice for alternative general
purpose languages at the time.)

There are still programmers who feel Fortran is the best choice when you
need the absolute fastest results for numerical programming. (I don't
know how true that might be.) So they still choose Fortran for new code.

The main Fortran compilers, as I understand it, are Intel's, gfortran
(part of GCC), and flang (part of LLVM). flang is a relatively new
project (replacing an older Fortran front end for LLVM). The fact that
people have developed a new Fortran front end for LLVM is, I think, a
good indication that the language is /not/ just for hobby use. Making
such a tool is a massive undertaking, and without knowing any details, I
expect there is commercial backing - it's not a hobby project. No
commercial entity would pay for it unless there was serious professional
use of the language, and an expectation that this would continue for
some time to come.

Like Cobol, Fortran is usually viewed as outdated but there is still
massive amounts of code in use, and still serious professional work with
it. And like Cobol, Fortran is not a language many people choose for
hobby use! (gcobol is in the process of being integrated into the main
GCC project - Cobol is not dead either.)

[David Brown]

For new stuff, these folks now often use Python and numpy or scipy, with
the backend written in C (or perhaps Fortran!).

[Mut...@dastardlyhq.com]

Or alternatively don't bother using a crippled spyware OS in order to run a
linux compat layer when you can just run linux directly.

[Chris M. Thomasson]

;^)

[James Kuyper]

You've lost the thread of the conversation. He was comparing Fortran
with C and C++. His point was that the creation of freeware C and C++
compilers didn't mark the conversion of either of those languages to
hobbyist status, so the creation of freeware Fortran compilers shouldn't
be expected to mark the conversion of Fortran to hobbyist status either.
Particularly since freeware Fortran compilers have already been around
for a long time. A small portion of all the new code that I've seen
people develop in the last decade or so was in Fortran, and GNU Fortran
was the only compiler I've seen any of them use.