[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Reject then reinvent..?

• To: Rick Jelliffe <ricko@a...>
• Subject: Re: Reject then reinvent..?
• From: Michael Champion <michaelc.champion@g...>
• Date: Mon, 27 Jun 2005 09:08:35 -0700
• Cc: "xml-dev@l..." <xml-dev@l...>
• Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=QFg/btbljwAbA5HmMOeZEEuyij53hnE+FrC4+8nfUzZPYn7GtWwFPs6eVaYxAYz5sY3jYtPWvlmXw2OUCmqwfasxn3PGBDqTmaT5Y1Zm3Uh7z4A5y7wbafqyqxBsDhNhPdI7EjcjmDTqTTesz/IbgFErbSOiq4lq0as6y0FSt/A=
• In-reply-to: <42BF96F9.2010504@a...>
• References: <1f2ed5cd05062508106a9d1305@m...> <42BF96F9.2010504@a...>
• Reply-to: Michael Champion <michaelc.champion@g...>

On 6/26/05, Rick Jelliffe <ricko@a...> wrote:

> Surely Danny means "I was wondering if there were any other cases around
> where a spec
> added specific elements over time as the pressure|demand for them become
> unstoppable, rather than buying into YALA (Yet Another Layer of Abstraction)."

Maybe Danny didn't mean that, but that's the best answer IMHO.  It's a
bit like asking why we keep inventing programming languages when we
have LISP, or why we keep fooling around with new data paradigms when
we have the relational model. Both these have formally sound
abstractions for all possible programs or data, but have proven to be
too abstract for ordinary mortals to use to get the job done.

I'm not involved in WinFS but I have given some thought to whether it
would make sense for them to use RDF/OWL for what they are doing.  I
came to the personal conclusion that it would not -- RDF/OWL are
certainly abstract enough to cover everything that WinFS is supposed
to so do, but it's really hard to imagine actual developers getting
used to the style of thinking: "This is 343-98-7612 ... 343-98-7612 
has the first name "John" .... 343-98-7612  has the last name "Smith"
.... 343-98-7612  has a wife "343-98-7733".   343-98-7733 has the
first name "Mary".  There are a lot of corner cases where that works
better than conventional structured data (e.g. when John and Mary have
a child of their own, John has a child with someone else, and Mary has
a child who was adopted), but most business apps don't care about the
biological details, they just need to know who the kids are in John
and Mary's family.

It all illustrates what I think is an inescapable point:  Industry
Standards are defined by industries, not standards committees.  For
whatever reasons, good and bad, RSS  has become adopted by the
industry, and that makes it in some fuzzy sense a "standard."  I'm not
happy about that myself, but I have long since accepted that the
worse-is-better stuff wins almost every time and there is little point
in whining about that unpleasant fact.

Hmm, "nothing is certain but death, taxes, and worse-is-better technologies."

• References:
  • Reject then reinvent..?
    • From: Danny Ayers <danny.ayers@g...>
  • Re: Reject then reinvent..?
    • From: Rick Jelliffe <ricko@a...>

• Prev by Date: Re: Reject then reinvent..?
• Next by Date: RE: Reject then reinvent..?
• Previous by thread: Re: Reject then reinvent..?
• Next by thread: Re: Reject then reinvent..?
• Index(es):
  • Date
  • Thread