> I wouldn't limit it to 10 layers deep. That should be all user configurable.
> Some implementations would need full history audits, etc.
There's a few implementations to be found on gborg (?), using
C or plpgsql, respectively. Another one (which I wrote for
GnuMed) is in the GnuMed CVS below
Ask for details if interested.
> My skill with
> triggers and plpgsql is not up to this task currently, but this is a
> suggestion for the PostgreSQL developers.
I should hope the developers spend their time on less trivial
(as long as the auditing isn't mucking with the MVCC
properties, that is) tasks. A big thanks to them for making
PostgreSQL what it is.
GPG key ID E4071346 @ wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD 4537 78B9 A9F9 E407 1346
In response to
pgsql-general by date
|Next:||From: Chris Stokes||Date: 2003-10-31 00:54:55|
|Subject: database speed|
|Previous:||From: Claudio Natoli||Date: 2003-10-30 23:47:58|
|Subject: Re: Info Required!!|