MISRA-C++

[E. Robert Tisdale]

see

www.MISRA-C.com

The page cannot be displayed
Explanation: There is a problem with the page you are trying to reach
and it cannot be displayed.

Try the following:

* Refresh page: Search for the page again by clicking the Refresh
button. The timeout may have occurred due to Internet congestion.
* Check spelling: Check that you typed the Web page address
correctly. The address may have been mistyped.
* Access from a link: If there is a link to the page you are
looking for, try accessing the page from that link.

Technical Information (for support personnel)

* Error Code: 403 Forbidden. The server denied the specified
Uniform Resource Locator (URL). Contact the server administrator. (12202)

 

[E. Robert Tisdale]

Which would mean that your assumption would have been correct for a
newsgroup dedicated to the development of embedded systems. This is a
newsgroup dedicated to discussion of the ISO standard C++ language.

Hi Chris,

Please ignore the off-topic cops.
All aspects of MISRA-C++ that address standard C++
are on-topic in the comp.lang.c++ newsgroup.

 

[P.J. Plauger]

Hi Chris,

Please ignore the off-topic cops.
All aspects of MISRA-C++ that address standard C++
are on-topic in the comp.lang.c++ newsgroup.

I'm basically inclined to agree. Nevertheless, anybody rude
enough to toss out a zinger like:

: On the other hand you are probably right. C++ language "experts"
: probably want nothing to do with actually using the language.

deserves to get zinged in return.

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

 

[Zara]

The page cannot be displayed

You may try http://www.misra.org.uk/

BTW, I think this is not OT. MISRA tries to define a Coding Style
oriented to make programs more reliable, as I see it. And that is one of
the objectives in this list: C++ style

 

[Jules Dubois]

Sorry I thought everyone knew of MISRA-C

Certainly not.

and therefore would know what MISRA-C++ would be.

Regardonly of what "MISRA-C++" would be, only that portion which is
compliant with the ISO C++ standard is on-topic. Anything else is not part
of the C++ language and therefore obviously belongs elsewhere.

 

[Chris Hills]

I'm basically inclined to agree. Nevertheless, anybody rude
enough to toss out a zinger like:

: On the other hand you are probably right. C++ language "experts"
: probably want nothing to do with actually using the language.

deserves to get zinged in return.

That was responding in kind to the unprovoked abuse thrown at me in the
in the first instance. The impression I have from several members of
the NG was we only want to talk about the ISO C++ standard [not the use
of the language]

I find it a strange reaction when some one posts a note of a group
working on a language subset for industrial use. Give the world wide
influence MISRA-C has had I would have thought C++ experts would have
wanted to keep track of the C++ version or even help guide it.

It is not use complaining after the event that it got it wrong if you
turned down the opportunity to guide it.

 

[Chris Hills]

You may try http://www.misra.org.uk/

BTW, I think this is not OT. MISRA tries to define a Coding Style
oriented to make programs more reliable, as I see it. And that is one of
the objectives in this list: C++ style

It is a coding guide NOT a style guide. Yes it aims to make C++ safer in
use. As did EC++

The C version started life as a UK automotive industry guide in 1997.
Now it is in world wide use by AFAIK ALL automotive manufacturers as
well as many aerospace and medical users. It has also been acepted by
the SAe and JSAE among others. Apart from a whole host of "ordinary"
users. BTW no one claimed it was perfect, a second version came out in
2004.

Because of the widespread take up of MISRA-C MISRA was asked to do a
similar guide for C++ so people have been mentioning it in various
places to give interested parties the opportunity to contribute.

 

[msalters]

Chris Hills schreef:

MISRA is looking at doing a MISRA-C++ This is because they were asked to
do it by a lot of people in industry.

They are looking for some people to make up the team (based in the UK)
The team will be made up of experienced industrial high-integrity C++ sw
engineers. They are looking at meeting once a month for about a year.

Will this be a publicly available document? If so, it might benefit
from being also offered as an ISO C++ Technical Report (like the C++
TR on Performance).

This will also shortcircuit any issues about being off-topic. True,
an ISO C++ TR is no ISO C++ Standard, but it is part of the "concept"
ISO C++

HTH,
Michiel Salters

 

[Chris Hills]

Chris Hills schreef:


Will this be a publicly available document?

Yes. It will be sold, as are most MISRA (and ISO, ANSI, BSI) documents.

If so, it might benefit
from being also offered as an ISO C++ Technical Report (like the C++
TR on Performance).

It could do if it goes in with out any editing by ISO or it's panels.
IE published as delivered.

This will also shortcircuit any issues about being off-topic. True,
an ISO C++ TR is no ISO C++ Standard, but it is part of the "concept"
ISO C++

However given the large response and numbers of people involved
internationally in MISRA-C I don't MISRA has any need of support from
ISO. The reason they are starting MISRA-C++ is that they were asked to
do it by a lot of people, companies and organisations.


The posting here was just to see if any C++ people wanted to get
involved.

Some do (thanks for the emails) and some don't.

 

[msalters]

Chris Hills schreef:

Yes. It will be sold, as are most MISRA (and ISO, ANSI, BSI) documents.

I see that "publicly" means something else to you. I actually meant
free-as-in-beer, to use the informal term.

It could do if it goes in with out any editing by ISO or it's panels.
IE published as delivered.

Couldn't say that for sure. Of course, /if/ it's edited, the resulting
TR would be different from the MISRA document and MISRA can still
publish
the original version. After all, ISO did change C when they got the
ANSI
C Standard (and most national organizations add a coverpage with
national
Standard IDs).

However, if MISRA would join WG21 for the purpose of this doc, it's
members can attend WG21 meetings. In that case, they would have direct
influence over the report. ISO doesn't willy-nilly edit documents.
There
may be a request by e.g. BSI to edit the docs (with the implied threat
of a NO vote). If the authors decide to ignore that, the result may be
that the ISO TR is never voted in or voted in despite BSI objections.
In both cases MISRA would have the document they wanted, allthough in
the first case it would be only a MISRA doc.

However given the large response and numbers of people involved
internationally in MISRA-C I don't MISRA has any need of support from
ISO.

IIRC, there may be other (legal) reasons. I'm not sure about the legal
form of MISRA, but ISO has the advantage that competing companies can
join forces. In some countries it would be an anti-trust violation, but
official Standards Bodies have exemptions. ISO is about as official as
it gets. E.g. the WG21 members together control pretty much 100% of the
C++ compiler market, but this is legal because it's ISO.

HTH,
Michiel Salters

 

[Howard Hinnant]

However, if MISRA would join WG21 for the purpose of this doc, it's
members can attend WG21 meetings. In that case, they would have direct
influence over the report. ISO doesn't willy-nilly edit documents.
There
may be a request by e.g. BSI to edit the docs (with the implied threat
of a NO vote). If the authors decide to ignore that, the result may be
that the ISO TR is never voted in or voted in despite BSI objections.
In both cases MISRA would have the document they wanted, allthough in
the first case it would be only a MISRA doc.

That's a very interesting suggestion. That would put embedded
developer's concerns directly in plain view of the committee, with the
possible end result that the C++ langauge / library itself evolves in a
direction that is at the very least, not embedded-hostile, and perhaps
even embedded-friendly.

Indeed this is the approach that was adopted by the EC++ consortium:

http://www.caravan.net/ec2plus/

While I do regularly attend the standards meetings, I was not part of
the EC++ consortium, and so can't speak as to the success or failure of
their efforts. However as a C++ vendor which regularly targets embedded
systems, I can assure you that there is interest already in the
committee on ensuring that C++ is a viable development language in the
embedded markets.

I believe I can say with some authority (as can Michiel Salters as well)
that MIRA's participation in WG21 would be welcomed.

-Howard

Senior Library Engineer
Freescale
Library Working Group Chairman, ANSI (J16) and ISO (WG21)

 

[Chris Hills]

Chris Hills schreef:


I see that "publicly" means something else to you. I actually meant
free-as-in-beer, to use the informal term.

You mean like the freely available ISO standard?

Couldn't say that for sure.

Then it is a non-starter.

However, if MISRA would join WG21 for the purpose of this doc, it's
members can attend WG21 meetings. In that case, they would have direct
influence over the report. ISO doesn't willy-nilly edit documents.

It has no need to do that. In industry the MISRA name is OK on it's own.
However at least two members of the MISRA team are on the UK C++ panel

There
may be a request by e.g. BSI to edit the docs

Then it would definitely not get anywhere near ISO.

(with the implied threat
of a NO vote). If the authors decide to ignore that, the result may be
that the ISO TR is never voted in or voted in despite BSI objections.
In both cases MISRA would have the document they wanted, allthough in
the first case it would be only a MISRA doc.

"only be a MISRA doc" is all it needs. MISRA has a VERY strong brand.
That is why it was approached to do it in the first place.

IIRC, there may be other (legal) reasons. I'm not sure about the legal
form of MISRA, but ISO has the advantage that competing companies can
join forces.

They do for MISRA ie ALL the worlds major automotive companies for the C
one. The net is a bit wider for the C++ version.

 

[Chris Hills]

That's a very interesting suggestion. That would put embedded
developer's concerns directly in plain view of the committee, with the
possible end result that the C++ langauge / library itself evolves in a
direction that is at the very least, not embedded-hostile, and perhaps
even embedded-friendly.

Which is why I mentioned MISRA C++ in this NG but I got told to take it
away!

I think that the C++ community should be involved especially the
standards people.

While I do regularly attend the standards meetings, I was not part of
the EC++ consortium, and so can't speak as to the success or failure of
their efforts. However as a C++ vendor which regularly targets embedded
systems, I can assure you that there is interest already in the
committee on ensuring that C++ is a viable development language in the
embedded markets.

Perhaps you may want to talk to the MISRA C++ team? They need reviewers
as well as a working party. Whilst the working party is UK based the
reviewers can be anywhere. For the C version reviewers were across the
globe. Several hundreds in fact.

I believe I can say with some authority (as can Michiel Salters as well)
that MIRA's participation in WG21 would be welcomed.

Then perhaps you need to talk to The MISRA-C++ team?

 

[Howard Hinnant]

That's a very interesting suggestion. That would put embedded
developer's concerns directly in plain view of the committee, with the
possible end result that the C++ langauge / library itself evolves in a
direction that is at the very least, not embedded-hostile, and perhaps
even embedded-friendly.

Which is why I mentioned MISRA C++ in this NG but I got told to take it
away![/QUOTE]

If you bait the topic police, they will come. Please stop. You do not
need nor should you want the last word in that debate.

I think that the C++ community should be involved especially the
standards people.


Perhaps you may want to talk to the MISRA C++ team? They need reviewers
as well as a working party. Whilst the working party is UK based the
reviewers can be anywhere. For the C version reviewers were across the
globe. Several hundreds in fact.

My organization is still evaluating your proposal. The reviewer status
you mention here for the first time is an interesting proposition. The
working party participation for non-Europe based personnel seems
excessively expensive. But a participation that didn't require long
travel once a month for twelve months may be much more palatable. It's
a shame I couldn't have informed my organization of this option a week
ago. Quite frankly I fear the milk has already soured. But I will make
an effort...

Additionally, perhaps the MISRA C++ team may want to talk to WG21? If
MISRA C++ ends up with the same spec that EC++ did, it will surely have
the same effect.

-Howard

 

[Chris Hills]

[...]
On the other hand you are probably right. C++ language "experts"
probably want nothing to do with actually using the language.

You know, I am offended by this remark. Not only have you no idea
where to post or how to post, you also apparently have a very strange
sense of how to behave when your mistakes are pointed out.

MISRA-C++ has NOTHING to do with standardisation of C++, a simple
search on 'comp.std.c++' shows that.

Which probably explains why WG21 (the ISO C++ panel) has appointed a
three person liaison team to MISRA C++. Several other WG21 people have
asked to be involved. SO now it is definitely on topic.

BTW until I mentioned MISRA-C++ there would have been no other mentions
of it to find in a search because it had not actually started at that
time.