Re: Unlimited size result sets

Walden wrote:

> She wants to throw around 100Meg data structures in Java?
> Tell me this isn't a web application where there will be multiple
> users doing this on a single server, just think, 10 users would
> be a Gigabyte of memory just for the data structure.

It is indeed a web application, and that is exactly what the proposal on the
table is.  There are only a hand ful of end users whose bills run this
large; the first standard deviation is around 30 calls.  So the average
workload on the server wouldn't be nearly this extravagant.

The whole idea was to be able to handle very small and very large data sets
without extravagant programming.  In her view, this means avoiding the 'page
at a time' approach like we do with green screen subfiles.  As I mentioned
earlier, it's really a proof of concept more than anything else, although
Java might surprise us.

> Next you'll tell me she wants to hold these structures in session objects.

I hear memory is cheap these days.  One never knows how things may go, but
having an open mind (and an architecture that is reasonably flexible) helps
adapt to change.  We're doing OK so far (touch wood).
  --buck