A thought, open for debate...

The node numbers are visible, and a bit painful to work with, in that
they are inherently a bit "numerically cryptic."

It would seem an attractive idea to extend nodes to have a node name.
Thus, we might extend sl_node thus:

 alter table sl_node add column no_name text not null;
 create unique index sl_node_name on sl_node(no_name);

And then have:

 STORE NODE (
    ID = 1,
    NAME = 'db1',
    COMMENT = 'Node on host db1'
 );

 STORE NODE (
    ID = 2,
    NAME = 'db2',
    COMMENT = 'Node on host db2'
 );
 STORE NODE (
    ID = 3,
    NAME = 'dbbackup1',
    COMMENT = 'Backup node on host backup1'
 );
 STORE NODE (
    ID = 4,
    NAME = 'dbbackup2',
    COMMENT = 'Backup node on host backup2'
 );


References would no longer be made to a numeric node ID, but rather to
the node name:

 drop node (name = 'dbbackup2');
 restart node 'db2';

and so forth.

Underneath, we would sling around node ID numbers, but the scripts
could use more mnemonic names.

This isn't valuable to people that just have two nodes, but it would
be quite valuable to those of us that have numerous Slony-I clusters,
and quite a bunch of nodes.  It also offers the possibility of
"renaming" a node.

It would definitely be a pain to change over to the new slonik syntax,
although it wouldn't forcibly require many internal changes.
-- 
"cbbrowne","@","ca.afilias.info"
<http://dev6.int.libertyrms.com/>
Christopher Browne
(416) 673-4124 (land)