Writing huge Sets() to disk

[Martin MOKREJŠ]

Batista, Facundo wrote:
> [Martin MOKREJŠ]
> 
> #- > At least you'll need a disk of 34694 EXABYTES!!!
> #- 
> #- Hmm, you are right. So 20E15 then? I definitely need to be
> 
> Right. Now you only need 355 PETABytes. Nowadays disk is cheap, but...
> 
> 
> #- in range 1-14. ;-)
> 
> Why?

I need to test for occurence every such combination in some real-world
examples. Many many will be missing, and those I have to keep.
I've no clue what size will be the drop, so I rather expect full size.
So, the generated, theoretical lexicon I could generate on the fly,
but the results I have to keep.

Even if I give up on this, can I write Sets onto a disk while keeping
their nice, built-in methods?
M.