RT 2.0 and Postgres 7.3

At this time, RT 2.0.x is not compatible with Postgres 7.3.

Postgres’ developers have made several incompatible SQL sand server
changes, requiring a number of changes to RT’s core.

At this time, it is strongly recommended that sites running RT on
Postgresql stick with v7.2

-j

rt-announce mailing list
rt-announce@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-announce

Jesse Vincent jesse@bestpractical.com writes:

At this time, RT 2.0.x is not compatible with Postgres 7.3.

Postgres’ developers have made several incompatible SQL sand server
changes, requiring a number of changes to RT’s core.

We are already migrating our databases to PostgreSQL 7.3. Can you
estimate how much work the required changes would be, and indicate
which areas of RT are affected?

Florian Weimer Weimer@CERT.Uni-Stuttgart.DE
University of Stuttgart http://CERT.Uni-Stuttgart.DE/people/fw/
RUS-CERT fax +49-711-685-5898

Jesse Vincent jesse@bestpractical.com writes:

At this time, RT 2.0.x is not compatible with Postgres 7.3.

Postgres’ developers have made several incompatible SQL sand server
changes, requiring a number of changes to RT’s core.

We are already migrating our databases to PostgreSQL 7.3. Can you
estimate how much work the required changes would be, and indicate
which areas of RT are affected?

There are a couple of things:

Any create or update tha can pass '' instead of '0' to the
database needs to be changed, as postgres has decided to start
sticking to the ANSI standard, which prohibits atoi from
converting '' to '0', as postgres < 7.3 does and as mysql does.
It is possible to patch your postgres sources to change this
behaviour, if you want. I don't expect to audit RT2's code to
make this happen, but would take a contributed change. RT3
should be better behaved, but I need testers.

DBIx::SearchBuilder needs to be updated to use postgres' new
LIMIT behaviour.  I've started the work on this, but it's not a
super-high priority.


Florian Weimer Weimer@CERT.Uni-Stuttgart.DE
University of Stuttgart http://CERT.Uni-Stuttgart.DE/people/fw/
RUS-CERT fax +49-711-685-5898


rt-devel mailing list
rt-devel@lists.fsck.com
http://lists.fsck.com/mailman/listinfo/rt-devel

»|« http://www.bestpractical.com/rt – Trouble Ticketing. Free.

Jesse Vincent jesse@bestpractical.com writes:

RT3 should be better behaved, but I need testers.

Hmm, in this case, we will try to port our RT extensions to RT3
instead. This way seems to be more future-proof. :slight_smile:

Florian Weimer Weimer@CERT.Uni-Stuttgart.DE
University of Stuttgart http://CERT.Uni-Stuttgart.DE/people/fw/
RUS-CERT fax +49-711-685-5898