Peter Fein wrote:
> Hi-
>
> Any general tips on using version control (CVS, SVN) while doing
> database design? My thought was to do a text-mode dump (including
> populated code tables) from PGAdmin.
>
> How do people do this?
>
Currently we just store a dump of the data structure. However, what I
think is really needed is a specialist diff tool which works out the
commands needed to move from one schema to another. That would be
*extremely* useful, but would also probably require a separate
implementation for each database backend.
There's also the problem of configuration data: you might want some rows
to be version-controlled too because they contain lookup validation
data which is vital to the operation of the system, rather than being
user data.
It's an interesting problem which I've not yet seen a usable solution
to. I've been looking for an OSS project to either start or contribute
to. Maybe that's one itch that I might consider scratching...
--
Russ.
