Missing gncGUIDCache table in PG Backend

[Linas Vepstas]

On Tue, Aug 12, 2003 at 12:09:45AM -0500, Matthew Vanecek was heard to remark:
> Linas can probably answer this.  I was doing a little regression
> testing, and discovered that for the 1.0.0 -> 1.1.1 PG backend upgrade,
> the upgrade referenced a gncGUIDCache table that does not exist.  I
> cannot find a reference to creating this table anywhere.  Is this
> perhaps something that never got committed to CVS, or is in another
> branch and didn't make it to HEAD?  I did a search via the CVS Web
> interface, and all it gave me was the UPDATES in upgrade.c and a couple
> of DROPs.  Any ideas, or is this a removable upgrade?

I think you still misunderstand what 'upgrade' does.

Version 1.0.0 used to have a table called gncGUIDCache.  It was a bad
idea. So I got rid of it.  The way to do this was with upgrade.c,
which added a new column to an existing table, copied from gncGUIDcache
to the new column, and then totally destroyed the gncGUIDCache table.
Which is why you will not find it anywhere else in the source code. 

You are not supposed to be able to find it anywhere else. Working
as designed.  The only place where you will find this table is in
old databases that were created with gnucash 1.4 or 1.6, and the 
user still has not run gnucash-1.8.x  If the user had run gnucash-1.8, 
then the upgrade process would have deleted this table.

The easiest (only rational way) to test the upgrade is to get a
copy of gnucash-1.4 or 1.6, create a postgres account, and then see
if gnucash-1.8.4 can open and work with it.   

The whole point of the upgrade is to create, alter and/or drop tables.  
When a table is dropped, (e.g. gncEntry), it will not appear anywhere
else, ever again, in the source code.  (You did drop gncentry, didn't
you?) 


--linas


-- 
pub  1024D/01045933 2001-02-01 Linas Vepstas (Labas!) <linas at linas.org>
PGP Key fingerprint = 8305 2521 6000 0B5E 8984  3F54 64A9 9A82 0104 5933