Red Hat Bugzilla – Bug 474415
no rows shown for a particular table using OO Base
Last modified: 2013-07-02 23:20:12 EDT
Created attachment 325571 [details]
database dump of problem table
Description of problem: I use Openoffice Base + postgresql-jdbc to edit a particular database. It mostly works, except on one table. If I try to load the table in Base, no rows are shown. If I try to re-create the table (using pgadmin3) and copy the data into the new table, I can briefly access the table. After a few accesses however, the new table becomes inaccessible.
I will attach a database dump for the table in question.
It might work for you a couple of times, but try loading the table in Base repeatedly until it stops loading.
Version-Release number of selected component (if applicable):
How reproducible: Always, for me.
Steps to Reproduce:
1. Use the provided database dump to create the basemodels table.
2. Try accessing this table repeatedly through OO base, using the postgresql-jdbc driver.
Actual results: Table becomes inaccessible. No rows are shown.
Expected results: Should show 407 rows.
Created attachment 325572 [details]
full database dump
This is a full database dump, because the basemodels table has some foreign keys that I forgot about. All tables in the database work fine except for the "basemodels" table.
Hm, can you say whether this problem is new in F-10, or was it there before?
The problem was also present in F9.
OK, I'll try to reproduce on F9. I won't be able to look at it till next week though --- I'm actually out of town right now and don't have access to my F9 machine.
Have you had time to look at this sample data? Nudge, nudge...
No, not really. It would save me some time if you explained what "accessing the table through OO base" means, bearing in mind that I know absolutely zero about OO's database features.
I tried to write an easy-to-follow sequence to reproduce the problem, but I can't reproduce the problem when I create a new OO database file... which leads me to believe that it is strictly an OO problem, rather than postgresql-jdbc.
I still experience the problem in my original OO database file, but recreating that file from scratch has made the problem go away. So there is a real bug somewhere, but I can't document it well enough for developers to fix it...
Closing bug for now.