ditionally, whether one might personally think that this behaviour is
important or not, it is what most commercial RDBMS do. So it would make
porting of applications from e. g. Oracle to PostgreSQL dramatically
easier. Until now, this has been one of the pitfalls. That's why I
. (if there is a practical reason it does not necessarily
need a technical reason, I believe.)
Sincerely,
Holger
--
[EMAIL PROTECTED], Bergisch Gladbach, Germany
Telefon (0 22 02) 5 99 91 oder (01 77) 7 92 24 66
---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]
one part
failed, (s)he can always program a rollback in case of at least one
error. But there should always be a difference between a rollback and a
commit, after at least one statement changing data has reported a
success.
Hopefully this can be cleared and perhaps improved within PostgreSQL.
Otherwise,
been able to find an explanation why PostgreSQL
behaves like this and why all other RDBMS I tried behave differently. In
this case the others make more sense to me.
Additionally, I have discovered that phantom reads occur in PostgreSQL
even if isolation mode serializable is used. Also not so nice!
Sincere
