ECMA-372 is stepping in on C++

[Michael Rasmussen]

Hi all,

I write to you because I am concerned about the steps taken by ECMA and MS
to substantially change the specifications for ISO-C++.

I have receive the following post from lois goldthwaite and the UK where
they express their concerns which I also have. Do you fellow C++
programmers have an opinion about that?

Read the full text from this address:
http://www.datanom.net/c++/Objection_to_Fast-track_Ballot_ECMA-372_in_JTC1_N8037.pdf

PS. I sincerelly hope that I am not OT.

 

[Victor Bazarov]

I write to you because I am concerned about the steps taken by ECMA and MS
to substantially change the specifications for ISO-C++.

Where in hell did you get this impression? Nothing in the proposed
C++/CLI specification affects directly what ISO-C++ (I assume you mean
the Standard C++ language) is.

I have receive the following post from lois goldthwaite and the UK where
they express their concerns which I also have. Do you fellow C++
programmers have an opinion about that?

I don't, except to say, 'Live and let live'.

Read the full text from this address:
http://www.datanom.net/c++/Objection_to_Fast-track_Ballot_ECMA-372_in_JTC1_N8037.pdf

I just did. Thank you for the link.

PS. I sincerelly hope that I am not OT.

No, you're not. Since you think (erroneously, IMO) that it somehow
affects the Standard C++ language, your post if topical. It's a knee-
jerk reaction, I think, but it's topical.

V

 

[Bob Hairgrove]

Hi all,

I write to you because I am concerned about the steps taken by ECMA and MS
to substantially change the specifications for ISO-C++.

I have receive the following post from lois goldthwaite and the UK where
they express their concerns which I also have. Do you fellow C++
programmers have an opinion about that?

Read the full text from this address:
http://www.datanom.net/c++/Objection_to_Fast-track_Ballot_ECMA-372_in_JTC1_N8037.pdf

I like this (closing statement):

"This paper should not in any way be taken as suggesting that there is
a sinister plot by Microsoft or anyone else to usurp or subvert the
C++ Standard. (...)"

LOL! If that's not what it is, then what is it? Once CLI is part of
the standard, then standards-conforming compilers will have to
implement it, whether they like it or not. Otherwise, the standard
will fork, and I don't think anyone wants that.

They did it to Java; they did it to XOpen/ODBC; what makes you think
they won't (try to) do it to C++?

PS. I sincerelly hope that I am not OT.

No, not as far as I can see.

 

[Michael Rasmussen]

LOL! If that's not what it is, then what is it? Once CLI is part of the
standard, then standards-conforming compilers will have to implement it,
whether they like it or not. Otherwise, the standard will fork, and I
don't think anyone wants that.

Well, the dont what to put it into the ISO specification, they what to
keep it as a separate standard but MS refers to the C++/CLI standard as
C++. That is my concern!

 

[Michael Rasmussen]

Where in hell did you get this impression? Nothing in the proposed
C++/CLI specification affects directly what ISO-C++ (I assume you mean the
Standard C++ language) is.

No, but what upsets me is that MS refers to the C++/CLI specification as
C++.

There is clear evidence that this confusion is already causing a problem.
Just take a look at the Microsoft online documentation -- start with the
article on "New C++ Language Features" at
http://msdn2.microsoft.com/en-us/library/xey702bw.aspx, which explains
about all the new keywords that have been added to C++ -- NOT "C++/CLI"
or even "C++.Net". Another example is
http://msdn.microsoft.com/visualc/default.aspx?pull=/library/en-us/dndotnet/html/NetFramework.asp,
which has many examples showing parallel code for "C#", "Visual Basic",
and "C++" (without qualifier). None of these examples would compile in a
Standard C++ environment.

 

[Victor Bazarov]

I like this (closing statement):

"This paper should not in any way be taken as suggesting that there is
a sinister plot by Microsoft or anyone else to usurp or subvert the
C++ Standard. (...)"

LOL! If that's not what it is, then what is it? Once CLI is part of
the standard,

Who said that they are trying to make it "part of the standard" (and what
do you mean by it)?

> then standards-conforming compilers will have to
implement it, whether they like it or not. Otherwise, the standard
will fork, and I don't think anyone wants that.

I think you misunderstand something here. It's not an addition or
correction to 14882-2003, it's proposed as its own standard (although
I am not certain whether it refers to 14882 or not). They are not trying
to introduce CLI into the Standard C++.

Essentially, they _are_ "forking" the language definition. Just like C++
has part of 'C' in it (and hence carries the 'C' in its name), they are
proposing to inherit most of the stuff from C++ and thus including those
three letters in the name of their language.

They did it to Java;

They did WHAT to Java? Has Java ever been standardized before they "did
it" (whatever 'it' is)?

> they did it to XOpen/ODBC; what makes you think
they won't (try to) do it to C++?

Why would they? C++ has already been standardized internationally, twice.
What would be the point for them to try "it". Please, open my eyes for
me.

What they are doing is trying to introduce the base for their proprietary
stuff so that others will have more reason to develop something to provide
spreading of their technology. It doesn't affect what C++ is or how it is
going to continue its life. It just gives MS CLI more weight. It also
establishes a more complicated mechanism to make any changes to it. I am
not certain MS folks realise that. But who am I to tell them?

Now, whether in reality people will jump on their band wagon and suddenly
begin developing the "standard" CLI (Common Language Infrastructure) is
yet to be seen. How many platforms does it exist on right now? Well, I
am not asking to continue a discussion on CLI here. I am just asking to
hold your horses a bit before accusing MS of an attack on C++ language.

V

 

[Victor Bazarov]

No, but what upsets me is that MS refers to the C++/CLI specification as
C++.

There is clear evidence that this confusion is already causing a problem.
Just take a look at the Microsoft online documentation -- start with the
article on "New C++ Language Features" at
http://msdn2.microsoft.com/en-us/library/xey702bw.aspx, which explains
about all the new keywords that have been added to C++ -- NOT "C++/CLI"
or even "C++.Net". Another example is
http://msdn.microsoft.com/visualc/default.aspx?pull=/library/en-us/dndotnet/html/NetFramework.asp,
which has many examples showing parallel code for "C#", "Visual Basic",
and "C++" (without qualifier). None of these examples would compile in a
Standard C++ environment.

I think MS opponents should be _happy_ about all those things. MS f***ed
up once again! As much as they might try to diminish the language, they
most likely won't be able to, if the language is strong and continues to
develop and improve, independently of what _other_ languages pop up here
and there. It all depends on the market. As soon as they gain support,
just because the number of programmers in the world is a constant, the
fewer will do _real_ C++. But that's market. There are no absolute
things here. Even if they rename it and call it F++, people will follow
if they see profit in it. And that means fewer will continue working in
C++. It's inevitable.

Now, if you're afraid that's what's going to happen, I cannot help you.
Nobody can. Fear of change is incurable.

V

 

[Shark]

No, but what upsets me is that MS refers to the C++/CLI specification as
C++.

There is clear evidence that this confusion is already causing a problem.
Just take a look at the Microsoft online documentation -- start with the
article on "New C++ Language Features" at
http://msdn2.microsoft.com/en-us/library/xey702bw.aspx, which explains
about all the new keywords that have been added to C++ -- NOT "C++/CLI"
or even "C++.Net". Another example is

I think this explains why many people post to this group asking
questions about Visual Studio windows programming/dlls/mfc etc. You
have a good point, imho......

 

[P.J. Plauger]

I think this explains why many people post to this group asking
questions about Visual Studio windows programming/dlls/mfc etc. You
have a good point, imho......

Uh, no. There is, perforce, in every programmer's career a time when
s/he knows only one implementation of a programming language (the
first one encountered, of course). It is *extremely* hard to tease
apart that which is specific to an implementation from that which is
portable across most/all systems. Quite a few programmers never get
beyond this point, simply because they don't have to. So this ng
gets naive questions about system-specific issues. They give the
kinder folk a chance to provide some basic education, and the Off
Topic Police yet another opportunity to be rude.

It has *always* been that way, folks. Whether the dominant vendor in
some enclave was IBM, DEC, AT&T, or whatever, the leading brand at
the time looked to many like the One True Implementation. And that
dominant vendor is more concerned with making happy customers than
explaining on a daily basis which bits are easily moved to a
competitor's platform and which are helpful additions supplied by
Yours Truly. You can accuse them of indifference, sloppiness, or
the evil crime of Vendor Lockin, but the forces at work are natural
and practically inevitable.

Today, Microsoft rules the roost with upwards of a billion
programmable machines as a potential marketplace and a cadre of
millions of programmers -- of varying abilities -- eager to pursue
that marketplace. What is actually unique this time around is that
Microsoft has made a concerted effort to make C++/CLI play well
with Standard C++. Even further, Microsoft has made a point of
ceding at least nominal control of their new dialect to a standards
organization.

So if the folks at the Evil Empire don't always distinguish their
innovations from the vanilla stuff, what the heck. At least they're
innovating.

P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com

 

[Bo Persson]

Well, the dont what to put it into the ISO specification, they what
to
keep it as a separate standard

Sure, it is not a change to The ISO C++, it is another language
standard with the proposed name of ISO C++/CLI.

So who's confused?
but MS refers to the C++/CLI standard as

C++. That is my concern!

That's the problem.

And if we get two standards, ISO C++, and ISO C++/CLI, which one is
then the real C++?

Why do we want two of them?


Bo Persson

 

[Chris Hills]

Where in hell did you get this impression? Nothing in the proposed
C++/CLI specification affects directly what ISO-C++ (I assume you mean
the Standard C++ language) is.

Yes it does, and it greatly effects it.

 

[Chris Hills]

372_in_JTC1_N8037.pdf

I like this (closing statement):

"This paper should not in any way be taken as suggesting that there is
a sinister plot by Microsoft or anyone else to usurp or subvert the
C++ Standard. (...)"

LOL! If that's not what it is, then what is it? Once CLI is part of
the standard, then standards-conforming compilers will have to
implement it, whether they like it or not. Otherwise, the standard
will fork, and I don't think anyone wants that.

It has happened in C already and it will happen in C++(/CLI) The
majority of C++ compilers in the world do not target a windows platform.

 

[Chris Hills]

Sure, it is not a change to The ISO C++, it is another language
standard with the proposed name of ISO C++/CLI.

So who's confused?

Al the MS manuals and documentation and some new books coming out on
C++(/CLI).

but MS refers to the C++/CLI standard as

That's the problem.

And if we get two standards, ISO C++, and ISO C++/CLI, which one is
then the real C++?

Why do we want two of them?

that is the problem. IT will by default mean that the MS platform is the
standard and anything else is not standard...