Statistics standard discussion

J

jacob navia

Last year, Paul Bristow proposed a series of mathematical functions
to be added to the C and C++ standard libraries.

I have implemented this proposal, and I would be interested in
knowing what happened to this proposal in the standards discussion.

References:
Reference: ISO/IEC IS 14882:1998(E)

For the implementation just download the lcc-win32
compiler system and use the stats.lib library, with the
header file "stats.h"

http://www.cs.virginia.edu/~lcc-win32
 
J

John Smith

jacob said:
Last year, Paul Bristow proposed a series of mathematical functions
to be added to the C and C++ standard libraries.

I have implemented this proposal, and I would be interested in
knowing what happened to this proposal in the standards discussion.

References:
Reference: ISO/IEC IS 14882:1998(E)

For the implementation just download the lcc-win32
compiler system and use the stats.lib library, with the
header file "stats.h"

http://www.cs.virginia.edu/~lcc-win32

I'm writing a stats library as a hobby project and would be
interested in looking at your code. Can the stats.lib source be
downloaded individually?

JS
 
F

Fred J. Tydeman

jacob said:
I have implemented this proposal, and I would be interested in
knowing what happened to this proposal in the standards discussion.

As I recall, there was not enough interest to do work on a
Technical Report. The C committee needs support from 5
countries for the work to progress.
---
Fred J. Tydeman Tydeman Consulting
(e-mail address removed) Testing, numerics, programming
+1 (775) 358-9748 Vice-chair of J11 (ANSI "C")
Sample C99+FPCE tests: http://www.tybor.com
Savers sleep well, investors eat well, spenders work forever.
 
J

jacob navia

John said:
I'm writing a stats library as a hobby project and would be interested
in looking at your code. Can the stats.lib source be downloaded
individually?

JS
I can send you the source code if you send me an email address.
The one in your message bounces.
jacob
 
J

jacob navia

Fred said:
As I recall, there was not enough interest to do work on a
Technical Report. The C committee needs support from 5
countries for the work to progress.
---
Fred J. Tydeman Tydeman Consulting
(e-mail address removed) Testing, numerics, programming
+1 (775) 358-9748 Vice-chair of J11 (ANSI "C")
Sample C99+FPCE tests: http://www.tybor.com
Savers sleep well, investors eat well, spenders work forever.

A pity. Thanks for answering Mr Tydeman.

jacob
 
A

algrant

Fred said:
As I recall, there was not enough interest to do work on a
Technical Report. The C committee needs support from 5
countries for the work to progress.

It would be wrong to deduce a mere lack of interest.

Some people may feel it is more important to deal with
issues in the core language, or in library functions that
cannot efficiently be implemented in C (e.g. SIMD and
saturating intrinsics, unaligned data, endianness,
synchronization, memory barriers), than having the committee
consider a potentially huge number of domain-specific
libraries which need domain-specific expertise to evaluate.

If the library can be written in C then just sell it or
give it away. If it's any good people will get it.
If the API is great but the implementation sucks, someone
will do it better. If the API sucks then let's be glad
it wasn't standardized.
 
D

Douglas A. Gwyn

If the library can be written in C then just sell it or
give it away. ...

Good point. One hardly needs to involve a standards committee
in order to distribute a library, unless the library critically
depends on some linguistic feature that isn't adequately
supported by the standard language (such as with <tgmath.h>).
Note that the type-generic features of <tgmath.h> aren't
available for any other code (library or not) in a standard
way, so nearly any numerical library might "need" to be
sanctioned by the C standard in order to be widely
implemented/implementable. That is a regrettable situation.
 
J

jacob navia

Douglas A. Gwyn a écrit :
Good point. One hardly needs to involve a standards committee
in order to distribute a library, unless the library critically
depends on some linguistic feature that isn't adequately
supported by the standard language (such as with <tgmath.h>).
Note that the type-generic features of <tgmath.h> aren't
available for any other code (library or not) in a standard
way, so nearly any numerical library might "need" to be
sanctioned by the C standard in order to be widely
implemented/implementable. That is a regrettable situation.

The argument for the proposal was that a standard notation
and a standard interface would allow statistics people to
interchange more easily their software.

I am giving away the library anyway, as you may know,
lcc-win32 can be downloaded at no charge (including the
statistics library and the documentation) at:

http://www.cs.virginia.edu/~lcc-win32
 
K

Keith Thompson

jacob navia said:
Douglas A. Gwyn a écrit :

The argument for the proposal was that a standard notation
and a standard interface would allow statistics people to
interchange more easily their software.

I am giving away the library anyway, as you may know,
lcc-win32 can be downloaded at no charge (including the
statistics library and the documentation) at:

http://www.cs.virginia.edu/~lcc-win32

Is the library implemented in portable C, or does it depend on
lcc-win32?
 
J

jacob navia

Keith Thompson a écrit :
Is the library implemented in portable C, or does it depend on
lcc-win32?

The library conforms to the interface proposed as standard. Some
parts are in assembler, others in C.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

No members online now.

Forum statistics

Threads
473,744
Messages
2,569,484
Members
44,903
Latest member
orderPeak8CBDGummies

Latest Threads

Top