Thinking things through was not a core Borland concept. Let use not
forget the "politically" "correct" disaster, their version 5 that never
worked, their destruction of the journalling system in favor a scheme
introduced a single point of failure for which there was no recovery,
etc. etc. etc.

They are the only culprits. MySQL error codes change every point
release. But then so does the storage engine interface.

On 10/14/2010 10:56 AM, Dimitry Sibiryakov wrote:
> 14.10.2010 16:36, Jim Starkey wrote:
>> It really don't make any bloody difference what the text of the
>> message is -- a database client should check the status code, not the text.
> You could leave this instruction to Borland guys who used isc_random for almost all
> DROP errors.
>


--
Jim Starkey
Founder, NimbusDB, Inc.
978 526-1376