[fpc-devel] Some Dataset and Variants patches

[Joost van der Sluis]

> > Any reason to re-implement the tempbuffer? Or were you just using an
> > older version?
> Please give me more datails (history) about "tempbuffer".

Well, it was an empty procedure. But no-one had a clue where it was used
for. Probably it had something to do with some very old fpc-TDataset-
descendents. So it got removed.

But if you can find a reason to keep it there... Maybe that ZEOS uses
it? 

> I can look deeply into ZeosLib and find out why it needs tempbuffer. 
> Patch with constraints and tempbuffer is really not important.