>Hi.
>I've hit the 2 gig lim on a file in an app
>It would be a huge job to start splitting up the file , or going to SQL SERVER so i'm trying to identify other options,
>There arent that many characyer felds that I can reduce, but there are a few numeric fields I could optimise.
>I have these now set to N 17 2 and would need to allow for 999,999,999.99
>Whats the Least otimum numeric data type for this ??
>Also, if there are any other suggestions ?

I'ld go right now for changing numeric fileds into integer, currency and double fields - just to get some breathing space till you implement your "true" strategy.

If there is no memo file as of now, you can put non-indexed char fields into memo file (for perf reasons leave fields often acessed in searches/filters in the dbf even if not indexed.)

Main routes to consider:
vertical split - attainable with probably the least amount of work. Split every record of your large table into to tables - should be "only" routine work fixing all the locations - based on the assumption that you have a PK. Drawback: you have only gained about 50% free space, and if your tables are growing quickly or exponentially this is not the best way to jump.

horizontal split: divide the table according to a consecutive dimension - best is probably time. If your current table has reached the 2 gig after 10 years, splitting into yearly tables will give you much more time than strategy 1. It also gives you nice options for backup strategies <g>.

use the time given by changing field types for moving into a database capable of handling more than 2 gig. Depneding on the way the current app is written this can amount to anything between some simple changes and a nearly total rewrite <bg>.

But if you are getting near the 2 Gig, make sure your backup strategy is in order!

my 0.02 EUR

thomas