🚀 go-pugleaf

J.Crowcroft@cs.ucl.ac.uk (Jon Crowcroft) writes:


>I'm interested in why the whole area of standards is so gothic
>anyhow...

Gothic is a nice description of it ...

[...]

>contrast with the far more distributed approach of de facto standards
>(e.g. internet, dare i say...)

Wellll, then almost everybody could implement/comment on it, that would
be open --- or even worse, people could point out shortcomings/bugs/
fuzziness, we don't want that, right? Also, people would notice that not even
members of the same committee are not sure what they're talking about/writing
about (one of the best example for obfuscated standard writing is ISO 8348-2 on
NSAP address, shudder)

Maybe we should propose an International Standard's Obfuscated Standard 
Writing Contest (ISOSWC), but then we'd be swamped with far too many 
eligible contestants, right? 8=)/8=(

[...]

>in fact, i believe that the "internatinal standards" in general and "open
>systems" in particular have almost exact opposite effects from the
>original supposed intentions - 

>they result in deliberate obfuscation by any companies with good technical
>input in an area, as who wants to subvert their own sales

Or by companies with *bad* technical input who want to protect their already
established equipment and not spent money on new stuff (X.25 for example 
could need quite some improvements, X.25/LLC2 likewise 8=(

[...]

>as for document availability, i don't object to copyright and so forth 
>to protect costs, but there is legal recourse rather than simple
>witholding of electronic versions

Hmm, shouldn't standards designed to provide OPEN SYSTEMS be available to 
everyone who want's to have a look (I'd agree to handling & printing charge,
but currently getting ISO docs is almost as expensive as getting drugs - or so
they say, I'm not on drugs, rather I'm on ISO docs 8=)

[...]

>i spose its too much to ask for - iso is based on levelling
>down/lowest common denominator almost by defn.

yes, sigh ...


	Dirk
---

In article <BtHJrB.H50@immd4.informatik.uni-erlangen.de> husemann@camelot.informatik.uni-erlangen.de (Dirk Husemann) writes:
>Hmm, shouldn't standards designed to provide OPEN SYSTEMS be available to
>everyone who want's to have a look (I'd agree to handling & printing charge,
>but currently getting ISO docs is almost as expensive as getting drugs - or so
>they say, I'm not on drugs, rather I'm on ISO docs 8=)

My local public library has copies of all ISO documents which have made
it to BSI standards (which includes all the OSI ones, SGML, esoteric 
things like DOA, etc. etc.). Note, _public_ library not university library.

They don't have RFCs.

Jon Crowcroft <J.Crowcroft@cs.ucl.ac.uk> writes:
|
|   most the western world is enganged in a move away from monopolies,
|   state intervention etc and yet ISO represents one of the most
|   central repositories for standards around

It's called "copyright", not a monopoly, although a copyright is indeed
a monopoly on the say over who gets to copy something.  You can,
however, enter an agreement with ISO or your national member body to
sell copies of international standards.  You don't need to make a
profit, or anything, all you need to do is pay ISO royalties.  This way,
you're free to undercut the market for the national member bodies who
you seem to think monopolizes their markets.  Not so.

|   contrast with the far more distributed approach of de facto standards
|   (e.g. internet, dare i say...)

Most soi disant "standards" are proprietary and have gained unfortunate
spread.  This is not true of ISO standards.

|   i think until proper competition is introduced, and open peer
|   technical review, and discursive definition of standards as opposed
|   to the current hierarchical approach and dry post facto presentation
|   style of most full standard documents...

You seem to know very little about what the working group procedures in
ISO are, and make the dangerous mistake of confusing it with the higher-
level political procedures.

As for the "dry post facto" thing, ISO was meant to embrace "prior art"
and existing standards.  OSI is a departure from that intention, but
mainly because CCITT was the inventing party, and ISO and CCITT have a
very cozy relationship.

|   i know we can put anyone up to attend standards meetyings (and myu
|   community, UK academics, does) - however, trying to pursuade anyone
|   really good from the technical world to go for any length of time
|   results in cries of "oh mi gawd, no, anything to avoid that political
|   quagmire"

Maybe you should attend a (special) working group meeting and get some
real life experience to relate your prejudice to?  Just because the top
level and the political structure of something is complex, the lower
levels who actually do the work need not be.  In fact, they are more
like the "open peer technical review" you're talking about.

|   in fact, i believe that the "internatinal standards" in general and
|   "open systems" in particular have almost exact opposite effects from
|   the original supposed intentions -

Please don't confuse international standards in general with OSI.  You
seem really hung up in trying to have people believe that all of ISO is
as bad as OSI might make people think if they didn't know better.  You
don't know better, and you should refrain from spewing your ignorance
around like you do.

|   they result in deliberate obfuscation by any companies with good
|   technical input in an area, as who wants to subvert their own sales

Yeah, right.

|   they result in zero competition where there is a good standard, and
|   semantic free standards where competition is likely (e.g. just look at
|   MPEG - deliberately avoids defining strict coding rules, since thats
|   where different companes will achieve an edge in
|   performance/quality...ends up underspecifying things...)

The automobile industry got competition all right when nuts and bolts
got standardized dimensions, and interchangeable parts became possible.
Why would the automobile industry want to "subvert their own sales" of
parts?  Because it cost a hell of a lot of money to make small volumes.

This is the best and most concrete example I have, but it's not the most
recent success story for standards.

|   as for document availability, i don't object to copyright and so forth
|   to protect costs, but there is legal recourse rather than simple
|   witholding of electronic versions

There's also the cost of legal recourse to consider.

|   how about some OPTIMISIM in how to do things, with legal redress for
|   when it goes wrong later

Substitute your "optimism" with "naivite".  Look around at all the
people who claim conformance to ISO xxxx, and don't.  Try to see this
from ISO's perspective, because it's ultimately ISO's perspective which
rules, not yours.

|   i spose its too much to ask for - iso is based on levelling
|   down/lowest common denominator almost by defn.

By your definition, perhaps, but to those of us who actually have some
experience in this game (_outside_ of the OSI quagmire), you're full of
hot air, Jon.

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.