On Sat, 06 Jul 2002 17:48:11 -0400, Tom Lane <tgl@sss.pgh.pa.us>
wrote:
>My objection to this remains the same as before: heap_getsysattr
>now has a core-dump condition built into it, because it can't tell
>whether it's safe to do HeapTupleGetOid().
HeapTupleHeader format is still unchanged. The change in the struct
has been proposed, not yet done.
>You really need to propagate relhasoids into tuple descriptors,
This is addressed in a patch I'm currently working on. It is harder
than I have thought (hoped), because tuple descriptors are manipulated
in many places in the executor and that's an area, where I have never
been before. I don't trust my changes enough to post the patch now,
I'm still testing...
>rather than having to write
>
>> + /* no Relation here to AssertRelationHasOids(relation); */
Yeah, in my working copy of the code (e.g. in GetSysCacheOid) I now
have
/* no tuple descriptor here to Assert(...->hasoid); */ :-)
You're right, these places are much less frequent than those where we
don't have a relation handy.
Servus
Manfred
