Well it was going to be FTP into the IFS...
But that may change to an HTTP GET.
Regardless the data will be in a pipe delimited ASCII text file in the IFS.
I agree on the 5 digit idea, so that rules out small int. But not integer
or Packed/Zoned. So we are back to my original question :)
On Wed, Jan 22, 2014 at 3:09 PM, <rob@xxxxxxxxx> wrote:
Name, Type/Length, Picture
MyFld, N/5, 9(5)
"Picture" seems to imply COBOL origins. Like the data is either coming
from a COBOL based package or the analyst still refers to data in terms
that come from their roots.
Since it's supposed to be data exchanged between disparate systems (PC or
mainframe to IBM i) then the odds that it's packed are probably still slim
to none. Therefore I think you have little worry that if you change from
packed then you're going to get a performance hit changing from sent
packed data to an alternative format.
What we all don't know is how the data is being exchanged.
Is it ftp exchange with your stream file system?
Is it ftp exchange with your Qsys.lib (aka DB2) file system?
Is it soap/rest exchange?
Does knowing any of the above really affect the decision?
Since they say 5 digits, then you had better, by golly, allow for 5 digits
and do not select some format that doesn't allow for that. Even if they
currently only use 3 digits for everything.
IBM Certified System Administrator - IBM i 6.1
Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
Kendallville, IN 46755
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives