On Thu, Oct 23, 2008 at 01:25:47PM -0400, Collin Kidder wrote:
> that that makes Thunderbird crippled but I see it more as a sign that
> nobody outside of a few fussy RFC worshipping types would ever want the
> behavior of the Postgre list.
Indeed. And PostgreSQL not interpreting '' as NULL, or '2008-02-31'
as a date, or other such silly strictness is just the imposition on
you of the personal views of a few fussy ANSI worshipping types.
Nobody would ever want such behaviour.
Of course, you could consider that the behaviour as defined in the
standards, which are there to ensure good interoperability, were
written over many years by painstaking standards weenie types who
spent a great amount of time thinking about the advantages and
disadvantages of these various options.[1] Or perhaps you think that
you're the only person to whom it ever occurred that some different
behaviour might be desirable?
Sorry, but pointing and laughing at people with whom you disagree
doesn't constitute an argument. Relying on an appeal to popularity
is, in fact, a well-known fallacy (sometimes known as _ad populum_).
If you want to argue that the standard is wrong, you need something
better than this. Alternatively, please go stand in the corner with
the people who think that MySQL version 3.x is the pinnacle of correct
database behaviour.
A
[1] To pick an example I can think of off the top of my head, you
would not believe just how much wrangling has gone on this year alone
over whether the "ß" character (LATIN SMALL LETTER SHARP S) should or
should not be allowed into internationalized domain names.
--
Andrew Sullivan
ajs@commandprompt.com
+1 503 667 4564 x104
http://www.commandprompt.com/
