Re: [forum] cooperation with ISO

[<Jim.Gettys@hp.com>]

As far as I know, all the national and international standards bodies 
(e.g. Ansi, ISO, etc), do variations of voting schemes for adopting standards 
based on membership at a national and/or international level (e.g. to 
get an ISO standard, Gabon's vote is equal to Germany's or the US).  And 
usually, ISO is more blessing standards  developed in other venues than 
generating new ones).

There is therefore extensive politicing involved in adoption of ISO 
standards, and no easy way for individuals to be involved, and any standards 
need development and probably some previous approval before they get there.  
It has been really more than 3 years since I was last involved with people 
who were participating in the ISO process, so this is somewhat stale memory.

The X consortium was company membership based voting of adotption, and 
had no provision for outside individual contributions.  It is somewhat 
less painful with the Web consortium  for two reasons (which has a membership 
agreement similar to the old X Consortium; it was patterned very much 
after it). 

In the Web consortium case there are so many member organization 
that you can usually dream up affiliation with one of them, and W3C has 
an "invited expert" exception to allow participation of individuals in 
their process; this was not true for the X Consortium).  It's been quite 
a while since I read the X.org contracts, but my memory is that they were 
very similar to the X Consortium; member company based voting, with little 
or no provision for contribution by non-affilliated individuals.

Keith Packard can tell you the previous sad tale of attempted ANSI 
standardization of X.  The rules of that organization and the other standards 
organizations have evolved since that date, of course, but it isn't 
encouraging.


So the issue before the community is 
	o how we develop "standards" in the first place
	o how/who can participate, on what grounds
	o the process for adoption by the community
	o the venue for this formal standardization, in circumstances
	when it is deemed good and necessry by the community.

Understand, I believe there are many circumstances where formal 
standardization processes are necessary or very worthwhile (heck, I'm 
the editor of the IETF HTTP/1.1 spec, and would not have spent some years 
of my life on that task if I hadn't believed it worthwhile), but I'm also 
leery of standardization before its time (something the IETF processes 
try very hard to avoid, while ensuring individual contributions are 
possible). 

Some of the processes the IETF uses are, at least right now, may or may 
not be infeasible to us: requirements to progress beyond proposed standard 
require interoperable implementations of the specification where the 
implementations are pretty independently developed, so even they routinely 
standardized API's, this requirement would be difficult to meet.

				- Jim

--
Jim Gettys
Cambridge Research Laboratory
HP Labs, Hewlett-Packard Company
Jim.Gettys@hp.com