On Fri, Jan 30, 2009 at 08:26:58AM -0500, Bruce Momjian wrote: > > > > Any change which requires #2 will need to be QA'ed in Oracle / > > Satellite as well. While it is a noble goal to have the codebase in > > pure ANSI syntax, milestone-wise it will be much more feasible to > > start with compatibility layer (#3) first and not depend on #2. > > > > Including for DECODE. > > Well, the wiki states we should try to make queries be the lowest > numbered item possible: > > https://fedorahosted.org/spacewalk/wiki/PostgresTechnicalApproach > > If we want to avoid #2 and make make more #3 and #4s, fine, but we need > to decide this as a group.
That's why I've sent it as a proposal to the mailing list. > Keep in mind if we decide we want no #2, we > are going to have a lot more #4 duplicate queries, which I thought we > wanted to avoid. My proposed rule is: If #2 can be also done via #3, start with #3, before doing #1. > Basically, if we make any #2s, the Oracle port is going to have to be > retested, so why try to minimize #2 changes? It's retest and it's retest. Retesting one query and retesting wholesale change of syntax are different matters, I assume. -- Jan Pazdziora Satellite Engineering, Red Hat _______________________________________________ Spacewalk-devel mailing list Spacewalk-devel@redhat.com https://www.redhat.com/mailman/listinfo/spacewalk-devel
