C++ Formalism

newer
any body can send me some .spec...

Steven T. Hatton

19 Nov 2005 19 Nov '05

07:37

"In this paper, we develop a general formalism for describing the C++ programming language, and regular enough to cope with proposed extensions (such as concepts) for C++0x that affect its type system. Concepts are a mechanism for checking template arguments currently being developed to help cope with the massive use of templates in modern C++. The main challenges in developing a formalism for C++ are scoping, overriding, overloading, templates, specialization, and the C heritage exposed in the built-in types. Here, we primarily focus on templates and overloading." http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1885.pdf http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1886.pdf,

Show replies by date

Steve Graegert

19 Nov 19 Nov

19:14

New subject: [suse-programming-e] C++ Formalism

On 11/19/05, Steven T. Hatton wrote:

...

"In this paper, we develop a general formalism for describing the C++ programming language, and regular enough to cope with proposed extensions (such as concepts) for C++0x that affect its type system. Concepts are a mechanism for checking template arguments currently being developed to help cope with the massive use of templates in modern C++. The main challenges in developing a formalism for C++ are scoping, overriding, overloading, templates, specialization, and the C heritage exposed in the built-in types. Here, we primarily focus on templates and overloading."

http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1885.pdf

http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1886.pdf,

Steven, Do you expect from the community to comment on that? Since these proposals, among others concerning C++0x, are still under active discussion and only experimental compilers in academic environments have partially implemented some similar techniques, primarily to prove that complete typed ASTs can be created with such a formalism (although C++ is only one example here), it is of no interest for this list. I think that comp.lang.c++ and comp.compilers are better places to get into discussions on this issue (if that is what you wanted). \Steve -- Steve Graegert Software Consultant {C/C++ && Java && .NET} Office: +49 9131 7123988 Mobile: +49 1520 9289212

Randall R Schulz

19:35

New subject: [suse-programming-e] C++ Formalism

Steve, On Saturday 19 November 2005 11:14, Steve Graegert wrote:

...

On 11/19/05, Steven T. Hatton wrote:

...
...

http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1885.pdf http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1886.pdf,

Steven,

Do you expect from the community to comment on that? Since these proposals, among others concerning C++0x, are still under active discussion and only experimental compilers in academic environments have partially implemented some similar techniques, primarily to prove that complete typed ASTs can be created with such a formalism (although C++ is only one example here), it is of no interest for this list. I think that comp.lang.c++ and comp.compilers are better places to get into discussions on this issue (if that is what you wanted).

Well, I for one, as someone interested in formal, mathematical logic and its relationship to programming and its possible application to programming languages and tools, appreciated this pointer.

...

Steve Graegert

Randall Schulz

Steve Graegert

20:02

New subject: [suse-programming-e] C++ Formalism

On 11/19/05, Randall R Schulz wrote:

...

Steve,

On Saturday 19 November 2005 11:14, Steve Graegert wrote:

...
On 11/19/05, Steven T. Hatton wrote:

...
...

http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1885.pdf http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2005/n1886.pdf,

Steven,

Do you expect from the community to comment on that? Since these proposals, among others concerning C++0x, are still under active discussion and only experimental compilers in academic environments have partially implemented some similar techniques, primarily to prove that complete typed ASTs can be created with such a formalism (although C++ is only one example here), it is of no interest for this list. I think that comp.lang.c++ and comp.compilers are better places to get into discussions on this issue (if that is what you wanted).

Well, I for one, as someone interested in formal, mathematical logic and its relationship to programming and its possible application to programming languages and tools, appreciated this pointer.

So have I and your point is indeed valid. There are lots of other mailing lists and newsgroups specifically for discussing advanced compiler techniques, algorithmic approaches to languages and the like. I am willing to discuss general programming topics and provide solutions to certain problems on this list but I think the OP is to far away from the purpose of suse-programming-e. \Steve -- Steve Graegert Software Consultant {C/C++ && Java && .NET} Office: +49 9131 7123988 Mobile: +49 1520 9289212

Steven T. Hatton

22:20

New subject: [suse-programming-e] C++ Formalism

On Saturday 19 November 2005 03:02 pm, Steve Graegert wrote:

...

So have I and your point is indeed valid. There are lots of other mailing lists and newsgroups specifically for discussing advanced compiler techniques, algorithmic approaches to languages and the like. I am willing to discuss general programming topics and provide solutions to certain problems on this list but I think the OP is to far away from the purpose of suse-programming-e.

\Steve Actually, I was just passing the information along. I'm too busy with Mathematica right now to focus on these papers. I did have an extensive discussion about matters related to this topic on comp.lang.c++.moderated several months ago.

news://Message-ID: <8b-dndN94_ni9EzfRVn-uQ@speakeasy.net> Steven

Steve Graegert

23:03

New subject: [suse-programming-e] C++ Formalism

On 11/19/05, Steven T. Hatton wrote:

...

On Saturday 19 November 2005 03:02 pm, Steve Graegert wrote:

...
So have I and your point is indeed valid. There are lots of other mailing lists and newsgroups specifically for discussing advanced compiler techniques, algorithmic approaches to languages and the like. I am willing to discuss general programming topics and provide solutions to certain problems on this list but I think the OP is to far away from the purpose of suse-programming-e.

\Steve Actually, I was just passing the information along. I'm too busy with Mathematica right now to focus on these papers. I did have an extensive discussion about matters related to this topic on comp.lang.c++.moderated several months ago.

Ok, did not know what to do with your post. Thought you would like this topic to be discussed here, which I think would be inappropriate on this list. \Steve -- Steve Graegert Software Consultant {C/C++ && Java && .NET} Office: +49 9131 7123988 Mobile: +49 1520 9289212

Steven T. Hatton

20 Nov 20 Nov

07:43

New subject: [suse-programming-e] C++ Formalism

On Saturday 19 November 2005 06:03 pm, Steve Graegert wrote:

...

On 11/19/05, Steven T. Hatton wrote:

...

...
Actually, I was just passing the information along. I'm too busy with Mathematica right now to focus on these papers. I did have an extensive discussion about matters related to this topic on comp.lang.c++.moderated several months ago.

Ok, did not know what to do with your post. Thought you would like this topic to be discussed here, which I think would be inappropriate on this list.

\Steve

I don't believe it would be inappropriate. It would, however, be unlikely. Skimming the papers, and knowing what they imply is probably of value for anybody serious about C++ programming. Though the formalism is new, the concepts work is not. There are some support libraries for this on the Boost site. Interestingly, I learned about the papers from a post on the /moderated/ Mathematica newsgroup/mailing list where the subject is arguably far less topical than it is here. Since this list tends to be fairly sleepy, I will contend that topicality is defined by the participant (within reason). Steven

6731

Age (days ago)

6732

Last active (days ago)

List overview

Download

6 comments

3 participants

participants (3)

Randall R Schulz
Steve Graegert
Steven T. Hatton

C++ Formalism

Steven T. Hatton

Steve Graegert

Randall R Schulz

Steve Graegert

Steven T. Hatton

Steve Graegert

Steven T. Hatton

tags

participants (3)