🚀 go-pugleaf

And can someone explain what National Information Standards Organisation
is? (as in NISO Z39.50, the information retrieval standard)

In article <1992Oct05.171911.9981@nuustak.csir.co.za> acooper@nuustak.csir.co.za (Antony Cooper) writes:
>Why is there a need for both ANSI (American National Standards Institute)
>and NIST (National Institute for Standards and Technology) in the USA?
>
>I understand that ANSI is a private institution that does not actually
>develop standards, but that it verifies that standards have been developed
>properly, and that NIST is responsible for Federal standards in the USA,
>such as FIPS (Federal Information Processing Standards).
>
>Is there not an overlap between the two? Are ANSI standards automatically
>accepted by NIST, and vice versa?

(I decided to post this after emailing it to Mr. Cooper, and it bounced.)

ANSI does develop standards. It is the U.S. national representative to
I.S.O. (International Standards Organization). As such it participates
in developing I.S.O. standards. It also develops standards for the
U.S.  (e.g., the standard is meant to be just for the U.S., and is not
intended to be an international standard). In some cases ANSI uses
"accredited bodies" (I think that is the correct term) to develop
standards as a "sub-contractor". As an example, the I.E.E.E. did the
work for POSIX.1 for ANSI. As an example of ANSI doing standards
itself, ANSI did the ANSI C language standard. ANSI is supposed to be
THE (capitals for emphasis) standards setting body for the US. It does
much more than computer-related standards (just as ISO does). There are
standards for paper sized, electrical connections, keyboards, bolt (as
in machine screw) sizes, cable densities, etc.

NIST (among other things) "advises" the US government (or parts of it,
I've never been quite sure) for procurement specifications. As an
example, there is a NIST document that describes specifications for
"POSIX-type" operating systems environments that should be met when
bidding on government contracts that require said specifications. This
is not exactly the same thing as specifying conformance to a particular
standard. In the POSIX example :
 (1) POSIX.1 has many, many options. The NIST spec says which options
     have to be present
 (2) NIST sometimes will not wait for a standard. [NIST made specs
     using draft versions of POSIX.1 and POSIX.2.]

To resume, (and to grossly simplify) ANSI is the standards organization
for the US. NIST writes specifications to be used/met for government
contract bidding.

--

  --- Matthew Caprile    || Hey, I only speak for myself ! Even I  ---
  ---                    || won't admit to the opinions expressed  ---
  --- caprile@apple.com  || here, so don't expect my employer to.  ---

Not quite right --

  NIST develops OFFICIAL US GOVERNMENT STANDARDS.  Sometimes it does
this in conjunction with ANSI work and sometimes not.  However, NIST
standards are very much official and are not just "advisory" or
"recommendations."  The acronym FIPS stands for Federal Information
Processing STANDARD, for example.

  The relationship between the IEEE and ANSI wasn't described very
clearly either.  The IEEE develops standards that are used world-wide
on its own, independent of ANSI.  Occasionally some of those IEEE
standards will be propagated to ISO via ANSI (e.g. the POSIX work of
IEEE 1003.*/ISO9945-* and most of the LAN standards in IEEE 802.*/ISO
8802-*).  However many IEEE standards are never offered to ANSI at
all.  Also, IEEE standards are world-wide rather than just being for
the US, though the power transmission specs do tend to be North
American conventions of 120 VAC, 60 Hz rather than the European
conventions of 220 VAC, 50 Hz.

In article <3865@ra.nrl.navy.mil> atkinson@itd.nrl.navy.mil (Randall Atkinson) writes:
>  NIST develops OFFICIAL US GOVERNMENT STANDARDS.  Sometimes it does
>this in conjunction with ANSI work and sometimes not.  However, NIST
>standards are very much official and are not just "advisory" or
>"recommendations."  The acronym FIPS stands for Federal Information
>Processing STANDARD, for example.

Or to put it another way, ISO/ANSI says "All ISO 123 widgets are 10mm long"
and FIPS says "All widgets used by the U.S. government must conform to 
ISO 123". In particular, NIST produces the US GOSIP, which is basically a 
list of ISO standards.

Matt Caprile <mfc@medoc.aux.apple.com> writes:
|
|   ANSI does develop standards. It is the U.S. national representative to
|   I.S.O. (International Standards Organization). As such it participates
|   in developing I.S.O. standards.

Just a very minor correction.  "ISO" is not an abbreviation, and it
doesn't stand for "International Standards Organization".  It's sort of
an acronym, but it also comes from Greek, where it's a prefix which
means "same".  The full name of ISO is "the International Organization
for Standardization".  "ISO" is never used with interspersed periods.

Best regards,
</Erik>
--
Erik Naggum             :  ISO  8879 SGML     :      +47 295 0313
                        :  ISO 10744 HyTime   :
<erik@naggum.no>        :  ISO 10646 UCS      :      Memento, terrigena.
<enag@ifi.uio.no>       :  ISO  9899 C        :      Memento, vita brevis.

In <1992Oct7.015834.16314@times.aux.apple.com> mfc@medoc.aux.apple.com (Matt Caprile) writes:

Mathew is essentially correct with the exception of two small nits. 
[Hey. I'm a standards person. I'm supposed to be retentive :-) ]

>ANSI does develop standards. 
ANSI develops *no* standards. 
It accredits other organizations to develop standards. 
The ANSI C standard was developed by a 
CBEMA organization, X3, hence the standard's designation X3.159. 

> (2) NIST sometimes will not wait for a standard. [NIST made specs
>     using draft versions of POSIX.1 and POSIX.2.]
Exactly correct, except there was never a real FIPS for POSIX.2 in draft form.
[FIPS == Federal Information Processing Standard]
There were a couple of attempts, but they failed due to the real instability
of POSIX.2 in those days. I point this out only because NIST is just now 
putting together the real draft FIPS of POSIX.2 as POSIX.2 became a full use 
IEEE standard in September. (IEEE Std. 1003.2-1992.) The NIST announcement
has just gone out this week to begin the process. 

>To resume, (and to grossly simplify) ANSI is the standards organization
>for the US. NIST writes specifications to be used/met for government
>contract bidding.
Absolutely. 

cheers,
stephe
-- 
---------------- So there is a God. What's your point? ----------------------
Stephen R. Walli      Mortice Kern Systems Inc.      phone: +1 (519) 884-2251 
stephe@mks.com        35 King Street N., Waterloo, Ontario, Canada, N2J 2W9

atkinson@itd.nrl.navy.mil (Randall Atkinson) writes:

>Not quite right --
[...]
>      Also, IEEE standards are world-wide rather than just being for
>the US, though the power transmission specs do tend to be North
>American conventions of 120 VAC, 60 Hz rather than the European
>conventions of 220 VAC, 50 Hz.

Not quite right -- :-)

The new power standard for Europe (I forgot the precise name of the document)
is now 230 VAC. Previously, there have been several voltages around 220-235 V.

Germany is changing at the moment from 220 V to 230 V and we have been told that
this won't have any consequences for our hardware ...

Markus

-- 
Markus Kuhn, Computer Science student -=-=- University of Erlangen, Germany
Internet: mskuhn@immd4.informatik.uni-erlangen.de  |  X.500 entry available
-A distributed system is one in which the failure of a computer you didn't-
-even know existed can render your own computer unusable. (Leslie Lamport)-