Carl:
> > In my experience, you can speed things up with more memory.
>  
>  Or, perhaps, by using less?

When files get to above a certain size (depending on lots of variables), it 
can make a lot of sense not to loft them straight into series, so something 
like:

  file: open/lines %myfile.txt

may well be faster than:

   file: read/lines %myfile.txt

(Unless, of course, %myfile.txt is is being accessed via FTP or some other 
distance access method).

It also, in Louis's case, probably makes a great deal of sense to sort as you 
suggest: build an index to the file and then sort the index. That way, the 
'parse in the original need only be done once; and the volume of data being 
physically moved around inside the sort is well reduced.

You've also neatly bypassed the need for a sort comparison function as your 
method is intrinsically stable.


But when it comes to working out what is actually faster none of us has much 
of an internal model of how Rebol goes about doing things.  All we can do is 
speculate and experiment.

If Louis has the time to try some of these suggestions for improvement and 
the let us know, it'd be a valuable start for understanding how to throw 
large chunks of data at Rebol. And he might even get his sort down to under 
twenty minutes.

Sunanda.
-- 
To unsubscribe from this list, please send an email to
[EMAIL PROTECTED] with "unsubscribe" in the 
subject, without the quotes.

Reply via email to