Doug:
This old code was 100% developed originally in RGB95_64 and now run for
the first time in RGBXE; first in 0627 and then as a check in a
re-installed 0526.
I have quite a number of variable names longer than 18 characters, all
of which are performing flawlessly in their particular contexts, which
may not represent one in which we're banging on system tables.
I'll test several contexts and report back.
Dan/Karen/Doug: thanks very much.
d:|
------ Original Message ------
Sent: 6/28/2016 3:41:57 PM
Subject: Re: [RBASE-L] - ERROR 2038 Referencing Existing Table(s)
From: "Doug Hamilton" <[email protected]>
To: [email protected]
Cc:
"Only difference is the [length of the] variable name..."
Is it just coincidence that the longer variable name is > 18 characters?
Shouldn't make a difference in RBXE, just an observation.
Column name limit is 18 characters in RBGX and 128 in RBGXE.
Doug
On 6/28/2016 4:42 PM, Bruce Chitiea wrote:
Why, yes, Ms. Tellef; yes it did.
That's disturbing. Let's see if it stays that way now.
Bruce
------ Original Message ------
Sent: 6/28/2016 2:35:39 PM
Subject: Re: Re[4]: [RBASE-L] - ERROR 2038 Referencing Existing
Table(s)
From: "karentellef via RBASE-L" <[email protected]>
To: [email protected]
Cc:
Just for giggles, did you try replacing the variable name with my much
shorter name to see if the error message goes away? Otherwise, yeah,
I wouldn't worry about it, set 2038 off...
Karen
-----Original Message-----
From: Bruce Chitiea <[email protected]>
To: rbase-l <[email protected]>
Sent: Tue, Jun 28, 2016 4:31 pm
Subject: Re[4]: [RBASE-L] - ERROR 2038 Referencing Existing Table(s)
Karen:
Thanks. In mine, the variable vLOJoin_Column_LIST is consistently and
correctly populated (with the three column names in the referenced
table); after which the 2038 Unidentified Table ERROR is thrown.
There's the mystery.
Since I'm getting what I want, should I take a walk on the wild side
and wrap the block with SET ERROR MESSAGE 2038 OFF/ON and simply hope
that my luck holds up?
B
------ Original Message ------
Sent: 6/28/2016 2:20:35 PM
Subject: Re: Re[2]: [RBASE-L] - ERROR 2038 Referencing Existing
Table(s)
From: "karentellef via RBASE-L" <[email protected]>
To: [email protected]
Cc:
Okay, let's see... So I just now tested your second select command on
my RBGX database and it worked fine.
Here's the syntax that worked for me:
SELECT (LISTOF(#1)) INTO vText iv1 FROM Sys_Columns t1, Sys_Tables t2
WHERE t1.Sys_Table_ID =
t2.Sys_Table_ID AND t2.Sys_Table_Name = "whycol"
This did NOT work (got "syntax is incorrect for select"):
SELECT (LISTOF(#1)) INTO vLOJoin_Column_LIST IND iv1 FROM Sys_Columns
t1, Sys_Tables t2 WHERE t1.Sys_Table_ID =
t2.Sys_Table_ID AND t2.Sys_Table_Name = "whycol"
Only difference is the variable name...
Karen
-----Original Message-----
From: Bruce Chitiea <[email protected]>
To: rbase-l <[email protected]>
Sent: Tue, Jun 28, 2016 3:55 pm
Subject: Re[2]: [RBASE-L] - ERROR 2038 Referencing Existing Table(s)
Dan/Karen:
Hmm, indeed.
Adding "AND LIMIT = 1" to the first block solves the problem for that
block:
--JOINER Column Name
SELECT #1 INTO vWalk_JLINK_COLUMN INDICATOR iv1 +
FROM Sys_Columns t1 , +
Sys_Tables t2 +
WHERE t1.Sys_Table_ID = t2.Sys_Table_ID +
AND t2.Sys_Table_Name = .vWalk_JLINK_TABLE +
AND LIMIT = 1
That is NOT the solution for the second block, as I need a list of
column names returned:
--Capture Column Names for LOJoin_ColumnSET Assembly
SELECT (LISTOF(#1)) +
INTO vLOJoin_Column_LIST INDICATOR iv1 +
FROM Sys_Columns t1, +
Sys_Tables t2 +
WHERE t1.Sys_Table_ID = t2.Sys_Table_ID +
AND t2.Sys_Table_Name = .vLoop101_TABLE
... adding "AND LIMIT = 1" returns only the first column name. Do I
first need to do a column count and then set the limit to that? Or is
this not the way to snag a list of column names for a table?
Thanks for helping.
Bruce
------ Original Message ------
Sent: 6/28/2016 1:25:49 PM
Subject: RE: [RBASE-L] - ERROR 2038 Referencing Existing Table(s)
From: "Dan Goldberg" <[email protected]>
To: "[email protected]" <[email protected]>
Cc:
Hmm when I run it in 9.5 it says “too many rows returned”, which would
make sense since most tables have more than one row?
Dan Goldberg
From: [email protected] [mailto:[email protected]] On
Behalf Of Bruce Chitiea
Sent: Tuesday, June 28, 2016 11:06 AM
To: [email protected]
Subject: [RBASE-L] - ERROR 2038 Referencing Existing Table(s)
RBGXE v. 0526 or 0627
Can anyone suggest why an Error #2038 ( -ERROR- #21 is an unidentified
table.) ... is thrown by the following code?
--JOINER Column Name
SELECT #1 INTO vWalk_JLINK_COLUMN INDICATOR iv1 +
FROM Sys_Columns t1 , +
Sys_Tables t2 +
WHERE t1.Sys_Table_ID = t2.Sys_Table_ID +
AND t2.Sys_Table_Name = .vWalk_JLINK_TABLE
TRACE confirms that the temp table referenced in variable
vWalk_JLINK_TABLE exists at the time this code runs. BROWSE confirms
that the data is correct.
TRACE confirms that the output variable vWalk_JLINK_COLUMN has been
populated with the correct value before the error is thrown.
The same dynamic occurs with a subsequent block:
--Capture Column Names for LOJoin_ColumnSET Assembly
SELECT (LISTOF(#1)) +
INTO vLOJoin_Column_LIST INDICATOR iv1 +
FROM Sys_Columns t1, +
Sys_Tables t2 +
WHERE t1.Sys_Table_ID = t2.Sys_Table_ID +
AND t2.Sys_Table_Name = .vLoop101_TABLE
The referenced table exists prior, and the columnname list is captured
into the output variable.
Several blocks of identical code - varying only in their variable
assignments - work without issue, before AND after each of these.
This is in code which has worked without modification for over a year.
Thoughts?
Bruce Chitiea
SafeSectors, Inc.
909.238.9012 Mobile
Virus-free. www.avast.com
--
You received this message because you are subscribed to the Google
Groups "RBASE-L" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected].
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to the Google Groups
"RBASE-L" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
