Date: Mon, 1 Sep 1997 19:25:53 -0400
Reply-To: "Bassett Consulting Services, Inc." <0002395748@MCIMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: "Bassett Consulting Services, Inc." <0002395748@MCIMAIL.COM>
Subject: Re: some FRAME problems
-----------------------------------------------------------------------
CONTENT: re: some FRAME Questions
SUMMARY: encapsulate relevant catalog entries
REL/PLTF: 6.06/n.a.
NAME: Michael L. Davis
INTERNET: Bassett.Consulting@worldnet.att.net
AFFILIATION: Bassett Consulting Services, Inc.
P-ADDR: 10 Pleasant Drive, North Haven, CT 06473
PHONE: (203) 562-0640
FAX: (203) 498-1414
-----------------------------------------------------------------------
My friend and SUGI 23 section co-chair, Phil Mason recently wrote:
> I still get the occasional 80000602 error showing up in various
> places in my frame applications. Tech support want my whole
> application (which is huge) and data before they will be able
> to tell me what the problem is. I have scanned SAS notes and
> tried various solutions proposed there. There seems to be a
> range of causes for this error. Does anyone know of any fixes?
Sometimes, I have a similar situation where part of a large
application will give me trouble. SI Tech Support needs to see
and play with the application in order to pin down what exactly
is the matter. However, only one or two FRAME entries are
relevant to problem. Also, shipping the entire application would
breech the confidentiality agreement in place with the client.
In those situations, I'll extract the FRAME entries that are
affected and copy them to an empty test catalog. I'll re-write
the FRAME to remove the need to supply previously invoked catalog
entries and catalog entries that are invoked subsequently. Then
I compress the catalog into a self-extracting ZIP file and e-mail
it (or FTP it) to the person in SI Tech Support who owns the
tracking number.
This approach has two indirect benefits. First, by encapsulating
the affect FRAME entries, I can code out any confidential references
to the client's data. Second, when I code out the other catalog
entries, the problem sometimes goes away. When this happens,
I gain additional clues as to the true source of my difficulty.