Re: [Treebase-devel] we may have to give up on PK auto-generation by postgres

[Vladimir G. <vla...@du...>]

On Jan 27, 2010, at 3:53 PM, youjun guo wrote:

> Every table in TB was created with a piece of script incuding a line  
> like this:
>  phylotreenode_id bigint NOT NULL DEFAULT  
> nextval('hibernate_sequence'::regclass),

Just to make sure we do not speak past each other...

Column declarations like the above are what one can see in the  
treebasedev instance since last Friday, when I announced the "switch  
to hibernate_sequence".  The declarations are exactly the effect of  
that switch.  Till Friday, the columns were declared as

  phylotreenode_id bigint NOT NULL DEFAULT  
nextval('phylotreenode_id_sequence'::regclass),

For reference, the treebasestage instance still has the old  
declarations.

I would guess individual-sequence declarations were in effect since at  
least mid-Nov 2009, when Rutger added them to treebase-core/src/main/ 
resources/TBASE2_POSTGRES_CREATION.sql  in SVN.

The original version of this schema (in SVN since mid-June-2009) did  
not have any auto-increment declarations on PKs, as far as I can see.

I would not yet guess one way or another how the columns were declared  
in the actual database at creation time, or when the first import from  
TB1 was done.

--VG