Thanks for all the suggestions everyone. My thinking on these is as
follows..

There are generic database replication tools. These tend to work by
polling and comparing tables. This is quite performance intensive,
especially if the replica is to be kept reasonably up to date.

Similarly, continually taking small incremental backups sounds a bit
fiddly and performance intensive. Also rather a slow recovery,
reassembling all the incremental backups.

There is also the approach of putting triggers on all tables, then
using a UDF to write details to another machine, then replaying the
changes. This is rather invasive on the database structure and hard
to keep up to date.

I'd hoped that Firebird's shadowing would provide a much easier and
more performant option. I'm not actually after a full replication
solution, just a way to quickly recover a database on a second
standby machine, which should be as up to date as possible comapred
with the primary database.

Shadow databases are SO close, but as things stand I have the problem
that I'll have to find NFS client & server software for Windows (yes,
I'm on a Microsoft platform here - sorry!) and I'll have to hope I
don't get a network glitch which would invalidate my shadow. This
might be less of an issue if I could detect the glitch and initiate
the rebuild of the shadow - but that's overkill compared with
queueing the changes until the link is restored.

I'm new to firebird, so I can't comment on the internals, but the
current shadowing seems so close to being really useful, but just
falls short. Is there any prospect of a modest tweak to make it
truly brilliant?

Cheers,

Tim