Christopher Browne wrote:
> There is a list of Works In Progress...
> 
> http://slony-wiki.dbitech.ca/index.php/Works_In_Progress
> 
> Most things are addressed, at this point.

I'd like to bring up two old issues around pgAdmin, which will probably 
increasingly be used with slony with 1.2 when it is packaged together in 
the pgsql installer.

There's still no final solution for storing the access information for 
pgadmin to the nodes. pgAdmin still uses "administrative nodes", which 
are normal slony nodes that are not enabled; sl_path is used for storing 
the connect strings as usual. This works with 1.0 and 1.1, because 
apparently a node that never took place in a replication won't make 
events prevent the event queue from cleanup. 1.1 creates listens from 
all existing path information, preferrably only active nodes should be 
addressed.

I asked twice about The Right Way (tm) to failover, with no answer so 
far. Will the algorithm for failover be consistent over versions as 
possible (can I do a 1.1 failover with the 1.2 or later slonik), or do I 
need a specific slonik for each. This boils down to the question: Is it 
feasable to implement the failover algorithm in pgadmin (as most other 
slonik commands), or should it be executed externally.

Regards,
Andreas