Got it, sort of!
Looks like it *does* work as long as the preparedstatement is not the very first query sent. As long as a Connection_executeQuery is sent first everything is peachy.
Smells like a small initialization bug in libzdb. But you guys are better suited to find out.
Paul J Stevens wrote:
I've tried both with ZBDEBUG=TRUE and of course with getLastError, but both tell me nothing.
And I didn't miss the getConnection part. Normal queries using executeQuery run just fine, as do prepared statements with mysql.
I'm attaching my testzdb.c which doesnt work.
My git tree with zdb related work is visible at:
http://git.dbmail.eu/?p=paul/dbmail;a=shortlog;h=zdb
the calls that talk to zdb are all in dm_db.c starting around like 150.
thanks,
Jan-Henrik Haukeland wrote:
Ps. Also run with the debug flag set, ZBDEBUG = TRUE; Not that we output much debug info, but you may be lucky. Since postgres work with pool.c this sounds strange, though pool.c of course may not test all cases. Please double check how you call libzdb if you have not already done so first.
Maybe you have a dbmail repository we can look at for the implementation?
-- To unsubscribe: http://www.tildeslash.com/mailman/listinfo/libzdb-general
-- To unsubscribe: http://www.tildeslash.com/mailman/listinfo/libzdb-general