Jerry,
You're saying that the DM is in 3 of the D-descriptor? I
think the problem is that the real data is an internal date, the
external data in your TCL statements is an integer month number, and
there is no way for the DM conversion to make meta-data that would
match. What would 12
.L RELLEVEL
That will display the item RELLEVEL stored in the VOC of the account
you're logged to. Or at least the version used to *create* that
account, if the VOC hasn't been updated after subsequent
updates/upgrades.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL
Try Savedlists do not contain data. They are collections of record
keys selected for processing as a group (reports, postings, etc.), and
deleting them does not change or delete the actual data. The records
pointed to by the keys still exist as-is in their original files/tables.
Keeping them
Chuck,
Thanks for extending the usefulness of that i-descr. I tend to
favor i-descr's without @n references so that they can be used as a
reference in another i-descr. The @n references and semicolons sure
make it more readable, but you are then limited to replicating the logic
in
Bob,
Hey, that was fun! I got this one to work:
U2UG.TEST
0001 I
0002
SUM(MULS(EQS(RAISE(TRANS(FILE2,OCONV(@ID,G1*1),1,X)),REUSE(OCONV(@
ID,G3*1))),RAISE(TRANS(FILE2,OCONV(@ID,G1*1),2,X
0003
0004
0005 10L
0006 M
It gets 1 from file2, using G1*1 of the key from file
Bob,
Went brain dead on the 5 and 9 in your FILE2. The corrected
I-descr should read:
0001 I
0002
SUM(MULS(EQS(RAISE(TRANS(FILE2,OCONV(@ID,G1*1),5,X)),REUSE(OCONV(@
ID,G3*1))),RAISE(TRANS(FILE2,OCONV(@ID,G1*1),9,X
0003
0004
0005 10L
0006 M
--Ron P.
---
u2-users mailing list
Bill,
After you did the mkdir in Unix, did you put an F-pointer in the
VOC of the account from which you want to access the new folder/program
source code directory?
0001 F
0002 /u2/SOFTWARE/NEW.BP
0003 make a new one with CREATE.FILE DICT NEW.BP -or- use the same one
as AP.BP?
Jeff,
For the something.else use @RECORDattrnumber. You could
even use OCONV(@RECORDnnn,correlative) if the something.else
requires formatting before being added to the output string.
list thisfile eval prod.num:'|':pck.qty:'|':@RECORD### fmt 255l
-OR-
list thisfile eval
Bob,
Does the called PROC have an RI in it by any chance? That
would clear the input buffers and leave nothing for the original PROC to
use in the IF A statement.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Bob Woodward
Sent:
Bill,
Whenever I'm doing output capture like this, I always issue TERM
32767,32767 first. Gives me the max term size UV will support, and only
1 set of HEADING lines (unless the report's REALLY long), and very
little chance the report will line-wrap. You could do a TERM 80,24 when
Tom,
READLIST sounds like the BASIC statement you're looking for:
READLIST dynarray FROM listnumber THEN/ELSE
But it does seem to require a list number rather than a list
variable name. If you leave out the FROM listnumber clause, it will use
the active list #0. If your
Try SET.INDEX to update the file's header to point to where the I_ file
is on the new W2K box.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Mark Waldron
Sent: Thursday, December 01, 2005 10:23 AM
To: u2-users@listserver.u2ug.org
Subject:
Mick,
UniVerse has ASSOC and ASSOC.WITH, look for them in UniData, or
something with a similar name... Both can be used on-the-fly.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Gahan, Mick
Sent: Monday, November 28, 2005 2:29 PM
Bill,
I think you're going to have to transpose your array (swap
columns for rows), or create a stand-alone array to hold just the dollar
amount, in order to make this work. The thing you're trying to LOCATE
is a 3-value dynamic array itself (1 attribute, 3 values), so it's not a
number,
Don,
Try using single quotes or backslashes around your command:
SH -c \mpack -s this is the subject -d body.file attachement.pdf
emailname\
Works on UniVerse, not sure about UD...
--Ron P.
-Original Message-
From: Don Kibbey
Sent: Wednesday, November 09, 2005 11:16
Peter,
I didn't find any grand total suppress keywords available on
our UniVerse system, UD may have one.
But if all else fails, try
GRAND.TOTAL 'P'
That will make the grand total blurb print on the next page to
be thrown away.
--Ron P.
-Original
Bill,
UniVerse/SQL -- there's your issue. SQL is enforcing an
INTEGER data type on the SETUP field. You'd have to convert that DICT
item to NUMERIC[1] (I think?) in 8 of an I/D-type or 6 of an
A/S-type. That would define the column as a 1-decimal-place number for
SQL purposes. If it
Ed,
Try HUSH. Here's the help for it:
HUSH statement
_
Jeff,
UV/UD do not keep track of which multi-value was last accessed.
Only the attribute. Each iteration of the For-Next had to start
counting from the beginning of attribute 1 according to your notes
below. If you re-run your test with FOR/NEXT/I (instead of 1,I) you
should see a
John,
Does the RECORD_A field in FILE_A contain item-id's from FILE_B?
A foreign key into FILE_B in SQL parlance?
I'm on UV, so sorry if this doesn't work for UD
SELECT FILE_A WITH NOT EACH RECORD_A AABBCC AND WITH A_IN_B
Where the Dictionary item A_IN_B would look
Does import in field 16 mean this record was imported from some other
DB? Mayhap that field got some funky characters in it (like a
left-arrow, or backspace) on that other system, and was not cleaned up
during import to your U2 system.
Turn on up-arrow mode in ED, and then look at attribute 14
The compiler is a C-routine, right? Maybe if the text editor doesn't
get the right end-of-file character in there, then the compiler never
sees the end-of-string \0 character in the C string array, so it just
keeps parsing in the next character off into the never-ending ether
At least until
David,
If you added a new part to an existing distributed file, and the
new part already had records in it, you could potentially have records
in this new part whose item-id should be in another part based on the
partitioning algorithm. Then you could create that item-id in the
proper
Or leaving off the quotes?
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Kevin King
Sent: Friday, March 04, 2005 5:17 PM
To: u2-users@listserver.u2ug.org
Subject: RE: [U2] Getting CR instead of @ at colon prompt
You might try pressing Ctrl-X instead
Folks,
snip refer to themselves in the third person (ie. the
brain). Or... Did I misunderstand yet another e-mail. Haha JK
Did ya notice the Haha JK (as in just kidding)? He was
intentionally misunderstanding so as to be funny.
--Ron P.
P.S. an End-of-thread is probably
I've seen:
(0:VARNAME)[5] on UniVerse to do that. Ugly, but works.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of
[EMAIL PROTECTED]
Sent: Wednesday, March 02, 2005 3:30 PM
To: u2-Users
Subject: [U2] [UV] making 1 our of 1
Hi
John,
Per the Unidata 6.1 manual on the IBM web site:
EVAL
Syntax
EVAL expression
Description
---
The UniQuery EVAL keyword allows you to define a virtual attribute
expression for the current execution of a UniQuery statement only.
expression can be any expression valid in a
Brenda,
Suggestion: put
IF NOT(NUM(SEL)) OR SEL= THEN .
Before your CASE block. That should trap any interesting user
input before you get to the comparisons.
--Ron P.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On
Harold,
You can include PCL escape codes directly in the print stream.
PRINT This will be in the normal font. :
PRINT CHAR(27): codes-for-font-and-pitch:
PRINT This will be in BIG letters
PRINT CHAR(27):
I used to experience lots of core dumps on UV 8 and 9 (DG/ux) when running TCL
selects/sorts that had I-descriptors. I think it was someone on this list that
suggested we look to see if the I-desc's had COMMON blocks in them. Sure
enough, they did. We removed the COMMON blocks, recompiled
Perry,
You probably won't be able to have multiple END TRANSACTION statements.
One BEGIN TRANSACTION where you have it, and one END TRANSACTION at the bottom
of the LOOP structure, should suffice.
Then just put a ROLLBACK and CONTINUE within each IF-, READ-, and
One BIG caveat on that one: MATCHES expects the 2nd operand to be a pattern
matching string. If the contents of VAL2 are a valid pattern matching
expression, it will be treated as such, not as just any old string. For
example:
RPTEST
0001 CRT TIMEDATE()
0002 VAL1 = 1234567
0003 VAL2 =
Bill,
You can use the backslash (\) as a delimiter within BASIC:
execute \SORT SALES.ANALYSIS BY CUST HEADING 'Year-To-Date Sales
':This.Date.F \
note the backslashes surround the entire command to be executed.
Or you could build the command in a variable, piece by
Peter,
Make sure your emulator is sending the same ASCII sequence your term
type is expecting. Even if you change your terminfo's, and use TERM/PTERM/etc.
to use ASCII 8 (backspace), your emulator may still be sending DEL (which often
shows up as ^? when it's not interpreted and
The UV keyword REQUIRE.SELECT makes this a possibility. On the Mentor and
Ultimate systems I cut my teeth on, if the 1st select returned zero records,
then the second select would select against the entire file. Once I moved to
UV, and learned about REQUIRE.SELECT, then breaking out SELECT
Bruce,
Did you try an ascii transfer? That probably won't work on a hashed UD file.
Try a binary ftp. UV has a tool called format.conv for byte-swapping. Not sure what
UD calls it. You may have to use that tool on the file after it's transferred over.
--Ron P.
-Original
I've seen this sort of thing a lot when emulators don't emulate the original
*exactly*. We used to get $20 after a @(-1) in ProComm Plus doing a Wy50 emulation.
When we looked in the terminfo library, sure enough, there was the clear-screen
escape sequence followed by $20. An actual Wy50
Mark,
How about allowing 2 usage forms:
PRINT ; stmts {; morestmts}
-or-
PRINT
If no stmts are supplied at the command line, then LOOP for input, building
a dynamic array, then write that out as the temporary program. That would allow you
to build something more complex
David,
A type-1 directory file splits item-id's at 14 characters, which I believe was
based on the old Unix file name limit. If a record's ID is longer than 13 characters,
a new subdirectory is created to contain the rest of the name. If the rest of the
name is also longer than 13
Shawn,
I just started using gVim last week, UV 10.1 on AIX. I have the Windows
client side stuff loaded running on my PC, and use FTP to get the source code and
write it back to the UV server. Nothing running on the UV side.
I do a :edit ftp://server/path/program.name to get
Mark,
I'm on UV 10.1.2.
The 1st value for Y didn't change after uncommenting the 2nd set
of statements. It came back 1235 both times. What did you observe, and
did you run this on UV 10.1.0?
--Ron Pingilley
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL
Allen,
From what I've seen in UV literature, they tend to officially call
CHAR(253) @FM for field mark, but use @AM attribute mark for
compatibility with other PICK-like implementations.
--Ron P.
-Original Message-
From: Allen E. Elwood (CA) [mailto:[EMAIL PROTECTED]
Sent:
Kevin,
Tried this on my UV system, and it works just fine:
LIST INVOICE 16 17 WITH 16 0.00 AND WITH 17 16 EVAL (@RECORD16 /
@RECORD17) * 1 CONV MD2 FMT 14R COL.HDG Tax % of Gross
05:28:17pm 16 Aug 2004 PAGE1
INVOICE... Tax Amount Net Invoice... Tax % of Gross
394407
What are the attribute numbers in the ORDERS file where the MFG.NO and
PART.NO are stored?
Let's say it's 4 and 18, respectively.
: SELECT ORDERS SAVING UNIQUE EVAL @RECORD4:'*':@RECORD18
should do it.
The Illegal attribute name for sort: MFG.NO:'*':PART.NO error message may
be saying that you
AFAIK = acronym for As Far As I Know
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Sent: Monday, August 09, 2004 11:30 AM
To: [EMAIL PROTECTED]
Subject: Re: [U2] Question on Unidata tables access speeds
I am doing select with BRANCH = 0616 0844 AND LEASE_TYPE
Joe,
UFD, as to my understanding, points to the home directory of the
account you are in, where the VOC resides. So you are actually looking at
the Windows/Unix level information. A type-18 file, for example, is just
one big file/record to Unix or Windows, and would look mammoth!
Sara,
That error message looks like there is a:
STOP 201,CONTRACT
or
STOP 201,FILE.NAME.VAR
at that point in the program, probably in the ELSE clause of an OPEN
statement. Can you post the top 20 lines or so of source code for that
47 matches
Mail list logo