Re: [xml-dev] Avoding a repeat of W3C XSD - was Re: [xml-dev] Is Web 2.0 the new XML?

[Rick Jelliffe <ricko@allette.com.au>]

Michael Kay wrote:

>I didn't say people should stay on because they're the only people who
>understand the thing. 
>
I didn't say you did.

>I said that if you bring new people in half way, they
>will bring in yet more "good ideas" and the spec will get even bigger as a
>result, in fact it will grow indefinitely. You did a classic - arguing
>against a point I didn't make, and failing to address the point I did make
>  
>
If your point is that if the gods depart before creation is finished, 
the suns of dust who
follow shall never finish and the specs shall grow inifinitely, as a 
natural law,
it is a ludicrous assertion that does not require addressing, except to 
reference Chicken Little.

All that will happen will be a refactoring: the gods may document their 
compelling rationales
before twilight so that the new blood (and the outside world) will 
understand what the
justifications are. During changeover periods, the outgoing members will 
take extra time to
make sure their POVs are understood. Corporate POVs should outlast 
indivual member's
POVs anyway. Newbies in groups have no choice but to be idiots if 
decisions are
not adequately documented and justified.  Standards will still get made: 
just different
ones. 

Everyone may trim back their "good ideas".  (In any case, why is a 
sprinkling of two
generations of "good ideas" necessarily worse than wholesale 
implementation all of
a single long-lived generation's "good ideas"?)

Specs may be factored out into frameworks and layers, dependencies will be
fewer, attempts at universal solutions may be discounted in favour of 
more modular designs,
groups may be smaller so that people can contribute in their areas of 
best interest
(e.g. for XSD there should have been one WG for dates only, etc.) 

People would have a very strong impetus to get their specs finished 
during their
term, and that its product would set up the next stage to go in their 
favoured direction.
This would work strongly against premature standardization and encourage
alternatives (even if the form of experiments which finally get 
standardized.)

Of course the people who initiated a project might not like the way it 
ends up: but that
is the case for many people who work in WGs for several years but not 
its entire life.

If 2 years is too short, I am happy to substitute 2.5 years. But less 
than 3 or 4 :-)

>Occasionally one wishes that certain individuals would retire, so that the
>rest of the group can get rid of the hobby-horse that's only there because
>that one person wouldn't shut up about it. But new people will bring their
>own hobby-horses, so it's not really a solution.
>  
>
If everyone's hobbey-horses were equal, perhaps. The topic is how to 
avoid a repeat of
W3C XSD. 

Cheers
Rick Jelliffe

P.S. I probably cannot respond to other points for a week or so in the 
forensic detail
that may be polite; I am recuperating from my surgery and I have to sit 
in the sun on a
banana farm from the next few months on doctor's orders (err, the farm 
and the particular
fruit are not specifically indicated) and the Internet access and data 
rates
up north in Australia is pretty frustrating.