Re: Cursors and backwards scans and SCROLL

"Zeugswetter Andreas SB SD" <ZeugswetterA(at)spardat(dot)at> writes:
>> but just give a warning and then run the underlying
>> query _again_, this toime with materialize on top and also do a Move to
>> reposition the cursor. This will probably not work correctly for all
>> tranasaction isolation levels though but it will penalize only these
>> cases that absolutely need it. The penalty will of course be
>> heavier ;(

> rescan can only work in serializable isolation, no?

I had thought about this and concluded it was not worth the trouble.
It could be made to work if we copy the snapshot data from old plan to
new, but realistically there's no value in it. Existing applications
that are successfully using backwards-fetch are using queries that don't
need materialization; therefore there's no compatibility gain from
adding this extra code.

regards, tom lane