I'm still in the "run it up a flagpole" stage of development. I had hoped to do some performance testing tomorrow.

The only things I can see gaining so far by using the File System object are:

1. The ability to work with very large files.
2. The ability to work with non-ASCII files and read a file byte-by-byte.

Let me look some more at the file object methods and see if there are any more major gains to be had. I'm pretty sure the "pointer" location in the file can be returned. However, any program logic should keep track of this independently. There may be one or two more areas that aren't handled (or handled well for large or binary files) in KiXtart file handling.

As for the return codes, I'm still pretty new with KiXtart, and I didn't realize you could raise an error code by "exit"ing the function with an exit code. That solves the problem of return codes conflicting with returned data. Thanks! I'll get some modified code out there tomorrow.

New Mexico Mark

[ 15 January 2002: Message edited by: New Mexico Mark ]