On Thu, 6 May 2010, Tom Lane wrote:
Devrim =?ISO-8859-1?Q?G=DCND=DCZ?= <[email protected]> writes:
On Tue, 2010-05-04 at 18:18 -0400, Tom Lane wrote:
In any case it was ignoring the elephant in the room, namely the
thousands of lines of PL/SQL code that could not by any stretch of the
imagination be handled that way.
Has anyone tested Spacewalk on EDB AS?
Dunno; but if the goal is to not depend on a proprietary database,
that path wouldn't get us there.
(Having gotten pretty up close and personal with the PL/SQL code,
I would bet that EDB could only offer about a 90% solution anyway.
There is some weird, not sanely automatically translatable, stuff
in there --- like use of autonomous subtransactions, not because
the code logic requires it, but to work around deadlocks in Oracle's
foreign key checking.)
It wouldn't be the worst idea to take this opportunity to beat the
Oracle-isms out of the code base. Many of those choices were made almost
ten years ago.
That refactoring would be incredibly expensive, of course, and I'm sure
that Red Hat engineers don't have the time for it. But it might be a
useful exercise to go through the worst Oracle-isms in the codebase and at
least point them out.
--g
--
Educational materials should be high-quality, collaborative, and free.
Visit http://opensource.com/education and join the conversation.
_______________________________________________
Spacewalk-devel mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/spacewalk-devel
