Re: 'Official' definition of ACID compliance? - Mailing list pgsql-general

From Peter Eisentraut
Subject Re: 'Official' definition of ACID compliance?
Date
Msg-id 200601051920.31686.peter_e@gmx.net
Whole thread Raw
In response to Re: 'Official' definition of ACID compliance?  (Scott Marlowe <smarlowe@g2switchworks.com>)
Responses Re: 'Official' definition of ACID compliance?  (Jaime Casanova <systemguards@gmail.com>)
List pgsql-general
Am Donnerstag, 5. Januar 2006 17:01 schrieb Scott Marlowe:
> The C stands for consistency.  Consistency means that only valid data
> can be written to the database.  MySQL fails this test precisely because
> it does / can write inconsistent data to the database.  Note that even
> the latest version, 5.0.xx, by default, inserts a truncated number on
> overflow.

That's not at all what the C is about.  The C criterion means that a
transaction transfers the database from one consistent state to another.  To
my knowledge, MySQL does that.  On its way there, it silently alters data
that would violate this consistency criterion, but this does not affect the
fulfillment of the ACID criteria.

pgsql-general by date:

Previous
From: "F. Fernandez"
Date:
Subject: java.sql.SQLException: ERROR: invalid byte sequence for encoding "UNICODE": 0xe3936e
Next
From: Jaime Casanova
Date:
Subject: Re: 'Official' definition of ACID compliance?