To get a successful standard you need a core team who work
hard, who are technically highly competent, and who
understand the needs of the users as well as (a more common
reason for failure) the needs of potential vendors. You need
a consensus on the general principles and objectives, an
aversion to introducing unproven innovations, and an absence
of people with an interest in obstructing the process. You
don't need consultation or democracy or legal authority;
these can sometimes help to achieve the necessary consensus
but can also slow things down or send things off in the
wrong direction.
But I don't think it serves any purpose to be secretive. I
have certainly always believed that the more people knew
what was going on, the greater the chance of success.
Publishing work in progress will enable the user and vendor
community to respond more rapidly when the thing is finally
published, and will harness the resources of a wider group
of people to spot the errors. I find it a little
disappointing, now that there is no cost argument to prevent
open dissemination, that W3C should (apparently) have a
policy of secrecy which goes beyond anything I ever
encountered in ISO or ANSI or X/Open or OMG committees.
Perhaps the problem is that they would be deluged by
feedback, but I doubt it.
Michael Kay