Horst has this absolutely correct. Take a look at the Internet, which we are using to carry on this conversation and do other useful things like file transfers, remote control and web stuff. These were all developed by using the early IETF process, which involved starting with small simple standards, putting out an open reference implementation and then finding the 'usefullness' point, and then re-implementing and re-defining (as a further standards) as usefullness increased and inter-operability was demanded.
Standards that work nowadays and make everybodies life easier have arisen out of somebody actually "doing something", and the process of becoming a de-facto standard has been helped by either sheer commercial market domination or complete openness. Complex domain specific "standards" developed on the white board and then imposed onto humanity have not often worked well AFAIK.
Marshal Rose has a very perceptive analysis of what standards worked and why for applications. http://mappa.mundi.net/features/mtr/
The problem that the IETF faces today, one that the OSI groups faced earlier, has been the complete turnover of membership in working groups to private companies. As we all should know, engineers, when presented a problem, come up solutions. Most of them are different and the standards process now merges them all together because no company/engineer wishes to back down. Recent IETF work looks like OSI work. That standards are huge, technically complicated (if not 'correct') and lack any practical implementation upon first release.
In case anyone's experience of the Internet is recent, let me say that there was a time when TCP/IP and it's application protocols of SMTP, Telnet, FTP, IMAP, etc. were considered technically inferior to OSI protocols and were thought to be doomed. DEC, for one company, completely ignored TCP/IP in favor of ISO networking.
I just threw away a loose leaf binder I had on the merits of OSI networking, a presentation given to a group of us at the University of Michigan in the late 80's. Fortuneately, most of us stuck with TCP/IP and that's what we recommended to the Universities managment.
I still remember the huge fight we had with the Hospital in the early 90's over this decision. At one time, their top level managment was advised that a small group of us at the University had invented TCP/IP and that's why it shouldn't be used, instead the OSI stacks should be used.
-- Wayne Wilson An attachment containing my pgp-signature is included. My public key fingerprint is: 9325 05AD 866B BCCB 45BF E86A 63E1 C6ED 4130 5461 My public key can be downloaded from wwwkeys.us.pgp.net
pgp00000.pgp
Description: PGP signature
