Re: sort --random-sort

[Frederik Eaton]

> >> - The second patch changes a constant ISAAC_LOG in isaac.h from 8 to
> >> 3. The code explicitly says that this can be done, so, going by that
> 
> I'm leery of this change.
> Won't such a reduction in the size of the state array
> have a detrimental effect on quality of the random numbers?
> 
> Maybe we need an option to trade off speed for quality,
> if it makes such a big difference.

Hmm. Maybe there will be a difference. Well, why don't I make
ISAAC_LOG or ISAAC_WORDS part of isaac_state so that shred and sort
can use different values? I don't think it will be important for
'sort' to use a very strong hash, since crackers only have the order
of the hash values to go by, and not the values themselves. 
Alternatively I could just seed a simpler checksum with an ISAAC
random value (i.e. just using ISAAC for the part that accesses
/dev/random etc.).

> I've included some comments below, mostly for improved
> maintainability/readability, and to keep this code consistent
> with existing style and practices used in coreutils.

Done.

> isaac.c should eventually move to lib/, but for now
> it's ok to leave it in src/.  Even in src/, the file name
> could be more evocative of its function.  How about rand-isaac.c?

I've updated my copy to use rand-isaac.c.

Frederik