How did this story end? Or is it still a problem?
From: [EMAIL PROTECTED]
Sent: Tuesday, December 13, 2005 9:06 AM
To: u2-Users
Subject: [U2] Secondary Index Problem
Hi All,
UV9.6.1.16
HPUX 11
I have an index that is no longer working and am not sure
why. The whole purpose of
David Wolverton
As a 'security risk', has IBM explicitly been asked to fix
this item and said they'd prefer just to leave a gaping hole?
Or is it like many things, everyone knows it, but everyone
thinks someone else has followed up on it, and it must just
be 'the way it must be'...
I think you have the answer. Someone would have to gain access into
your system, and having done so, what would be the point of
sabotaging something within universe to do something malicious? They
already have access to your system. If it's someone internal, then I
would imagine your hiring
It is a security hole, well-known and by design.
From: john reid
I notice that an ls -lt in the u1 /uv /catdir directory
indicates that the *PROGRAM.NAME is updated apparently each
time an execution happens, at least that is what it looks
like to me. Anyone know if or why that is
Try using OCONVS (with an S) instead of OCONV.
The problem has nothing to do with I-descriptors.
OCONV treats the 1st argument as a string.
OCONVS will process each delimited substring individually.
Notice the FD (hex value mark) in the middle of the 1st output line
below,
but the }, which is an
From: Stevenson, Charles
Try using OCONVS (with an S) instead of OCONV.
After seeing Ray W's response, I think I misread your purpose. What I
said was true, but irrelevant.
The formatting got screwed up on the email. Be gracious let me blame
it on that.
I'll second Ray's answer instead
I have not had occasion to use the src4uv.com decompiler service.
Probably worth the price when you really need it.
I am thankful for the wonderful www.src4uv.com freebie download,
SRC.UV.HEADER, that allows access to info about the object file from an
I-descriptor subroutine.
Besides VLIST, but
Everyone's going to jump in with the answer, but here's a way to figure
it out yourself (on UV, but not UD): Use VLIST.
Compile the program, then VBLIST will decompile the object (pcode) into
an assembler-kinda-looking output you can see step by step what's
happening.
CT CDS.BP CALEB
From: Perry Taylor
Hrm. Good info.
How would one go about finding a hung port.status semaphore?
Red Herring:
---
One would *expect* smat -s to indicate it, but no.
Toward the bottom of a smat -s it says:
Port status State Netnode Owner Collisions Retries
Semaphore
link with the list_readu executable to run
UNLOCK commands from UNIX :
# cd `cat /.uvhome`
# ln bin/list_readu bin/unlock
# bin/unlock LOGINLOCK SEMAPHORE
From: Stevenson, Charles
A more dangerous similar example is locking
@LOGNAME = 'anthonylaptop\Anthony Dzikiewicz' it is awkward
because of the space. Keyvalues with spaces are awkward to deal with.
Anthony
My dear Mr. Dzikiewicz,
The space is the least of the awkwardness. Shoot, I have a hard
enough time spelling Stevenson! (We have this security
Tim,
Was transaction logging or data replication used?
Can you share any interesting performance stats comparing w/ vs w/o TX
or DR?
cds
---
u2-users mailing list
u2-users@listserver.u2ug.org
To unsubscribe please visit http://listserver.u2ug.org/
).
cds
From: Timothy Snyder
Stevenson, Charles [EMAIL PROTECTED] wrote on
11/22/2005 11:44:26 AM:
Tim,
Was transaction logging or data replication used?
You're not the first person to ask that. Unfortunately these
were not part of the goals of this load test.
---
u2-users mailing
LIST PARTS EVAL @NI FMT 3R 04:36:00pm 22 Nov 2005 PAGE1
PARTS. @NI
TARP-GHZPDT-CFG 1
42H1128 2
2108503 3
0511045219-008 4
7258940 5
3769333 6
NAI-DAE50036S7
PS1520-108
3C8228C 9
2657
yeahbut, be careful not to leave the semaphore set via miss-typing:
For examplem if you did:
$port.status | asdfasf (where asdfasf is garbage.)
You'll need superuser (or uv administrator?) to do
UNLOCK PSTATLOCK SEMAPHORE
from TCL.
From: Rod Hills
If you display the VOC entry for
report_lic /tmp/uvlicr.txt ; cat
/tmp/uvlicr.txt
This is kinda rinkydink too, because if userA runs uvlicr, the /tmp file
gets created and is not overwritten when userB runs it. uvlicc is not a
problem since only superuser can run it.
From: Stevenson, Charles
yeahbut, be careful
2 processes can write to the same COMO, too (and ED it simultaneously
from a 3rd).
Looks like sometimes the 2nd overwrites what the 1st writes, sometimes
they interweave.
I dunno, I just avoid it rather than keep[ing] them dogies rollin' and
sort[ing] 'em at the end of the ride. (Rawhide).
And
From: Marc Harbeson
What about an index on an I-DESC? (Does universe do that?)
Yes.
Virtual attribute is another name for I-descriptor, UniData's term
mostly.
See Scott Ballinger's earlier post on thist thread here is a repost of
part of my own answer:
From: Stevenson, Charles
Sent
On UV I believe there were some optimizations at rel 10.0 and it will
use both indexes and take the intersection.
Use EXPLAIN keyword to see what's going on.
Before then, and maybe still on UD(??) it is as you described.
A good workaround is:
SELECT FILE WITH DATE = 10.18 TO 1
SELECT FILE
TRANS( file, mv.ids, mv.target.fld, 'X'); CONVERT( @SM, '~', @); SUBR(
'-FIELDS', @, '~',1,1)
or
SUBR( '-FIELDS', CONVERT( @SM, '~', TRANS( file, mv.ids, mv.target.fld,
'X' )), '~', 1, 1 )
Explanation: Since FIELDS the other multivalued handling subroutines
break things out by and other
And just for jollies here are some others that are sorta like what you
want:
DICT 3. DICT 5... Result.
DWAMADY 30L WEDNESDAY SEPTEMBER 14 2005
DWAMADY}MCT 30L Wednesday September 14 2005
DWAMADY[3,3,2,4]
No, I did not know these tricks! Thank-you, Richard Stuart.
oconv( 13801, DWAMADY[', ',' ',2', ',4] ) = THURSDAY, OCTOBER
13, 2005
oconv( 13801, DWAMADYL[', ',' ',2', ',4] )= Thursday, October
13, 2005
oconv( 13801, DWAMADY[3'., ',3'. ',2', ',4] ) = THU., OCT. 13,
2005
oconv(
Bruce,
There has been several responses, but I don't think you have an answer
yet, do you?
I haven't followed the closely don't have the original problem, and
it is unclear to me whether you have multi- or single-valued date/part
data.
At any rate, you'll need to make 2 passes on the file,
I prefer to write what essentially becomes the final User Documentation
during the Requirement phase of a project. It is the spec the users
sign off on and I design and program to that.
Sure, there were a few mods along the way, but the approach works very
well. And yes, there are often
Vectoring (is 'vector' a verb?) is built into TRANS already.
No need to abandon in favour of OCONVS T-correlative.
You can specify a dynamic array of IDs for the target file to read.
I have used the Translate code with OCONVS() function when it
is stylistically in keeping with the use of
oo! OO! I know! I KNOW! :
Don't use GOSUB.
Only program with GOTOs.
Problem solved.
If you insist on using GOSUBs instead of GOTOs, then what Ray said.
Don't specify it until after main routine is done and gosubs begin:
01: FUNCTION ANTHONY( INARG1, INARG2 )
02: GOSUB ABC
This is a bit disconcerting.
BASIC SELECT should be faster than EXECUTE SELECT...
Maybe the smart people can weigh in on this:
From: Louis Windsor
A few years ago we used the BASIC SELECT FILE as opposed to
the EXECUTE SELECT FILE.
We updated UniVerse (don't ask from what version to
yeahbut: basic's SSELECT does an alpha sort, even if keys are numeric.
i.e.: 1,10,100,11,2 instead of 1,2,10,11,100.
And the important gains for basic SELECT filevar over EXECUTE SELECT
filename... are lost, (namely traversing the file only once), since the
file needs to be entirely traversed to
Do I detect an enhancement suggestion?
Whether this is trivial or hard is beyond me, but I tend to believe Mr.
Phillips.
(On this subject, anyway. I don't know his opinion of the British
Royals, for example.)
cds
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On
One beauty of using a source control system like RCS or SCCS is that you
do not need to maintain the mods in the source, the current running
version of the source looks clean, but you can also view or print out
the incremental differences.
When I first introduced using RCS at one shop I got
From: Hari T. K. Varma
Is there any direct one step method to locate the position of
last occurrence of a Sub-string in an existing string.
For ex:- to locate the last occurrence of AC in the string
X=AC2BC2CD2AC2AC2BC2SS for this case the position should be
returned 5.
Your valuable
Can you replace your final
MERGE.LIST 1 UNION 2 TO 1
with
MERGE.LIST 1 UNION 2 TO 0
?
From: [EMAIL PROTECTED]
SELECT TRX ..limited branches.. TO 1
SELECT TRX ..limited branches.. TO 2
MERGE.LIST 1 UNION 2 TO 1
SELECT TRX ..limited branches.. TO 2
MERGE.LIST 1 UNION 2 TO 1
Which
yeah,, and HTH and CTDSL are initials not acronymns.
Acronyms are initials that are pronounced as a single word.
NASA NATO are acronyms.
UN is not.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Don Verhagen
Sent: Wednesday, September 14, 2005
I have UV/RCS, a UV frontend to GNU RCS.
It allows RCS functionality and security with the comfort of never
having to leave your own home in UV:
- Universe menu,
- TCL commands,
- honours and builds select lists,
- and you can control hashed records dictionaries as well as
programs.
PH ids are usually [jobname]_[itime]_[idate].
The time-date suffices generally suffice to insure a unique id
(filename).
But what should happen when the same jobname is spawned more than one
time in one second? E.g., 25 instances of SQL.DATA.EXTRACT at exactly
00:02:59?
Instead of adding a
btw named common isn't quite safe either - it's possible
that another routine uses the same name for it's common ...
I hit that using Pace at a previous job ...
What's really needed is a 'local static' modifier.
Would be nice :-)
What's really needed is a good software mapping
No one mentioned the ENUM keyword?
Here is an example using a file everyone knows:
SORT VOC BY TYPE SAMPLE 10 ID.SUP BREAK.ON TYPE ENUM @ID
TYPE VOC.
K=
K=
K
K=
** 4
VCLEAR.LOCKS
VDIVX
VHASH.TEST
VMENU.DOC
VT.LOAD
** 5
XQUIT.KEY
** 1
ENUM works with multivalues too.
Here is an example, again using a file everyone should know, Pick
Flavour NEWACC (NEWACC is used to build VOCs for new accounts) verbs
have multivalues in 8.
The following shows how many NEWACC verbs allow the various ...-SUPP
keywords.
F8.MV is a D-item with 6 set
From: Kevin King
How about...
READU REC FROM FILE,Z ELSE
PRINT 'DUSTIN IS AN A**HOLE'
STOP
END
Real code with the exception of the asterisks.
And this one:
READ ... ELSE
CRT ERR: 'Direct access to nonexistent data not allowed.'
which regularly displayed on the users'
If a guy maintains existing systems that someone else wrote, you can't
fault him for not running into a construct that doesn't show up in said
systems.
I learn new stuff on this list, too. I suppose the DI conversion is in
the manual, but I learned it from Neil Charrington on this list
yesterday.
On another note, I've yet to come across any apps that use
the NAMED COMMON concept. Is that a recent addition to the MV
world or just not that popular.
Thanks in advance.
Mark Johnson
wow.
[Pause. It slowly sinks in.]
WOW!!!
Named (aka labelled) common has been common on aps with
I did not know that! Works like a charm on UV.
Mark, how about D3?
I just did DI not wrapped in an F- or A-correlative:
CT VOC CDS
CDS
0001 X
0002 12/15/05
LIST VOC CDS F2 CDS.DATE CONV CDS.DATE
LIST VOC CDS F2 CDS.DATE CONV CDS.DATE 02:53:48pm 02 Sep 2005
PAGE1
VOC.
Clif Oliver Tim Snyder already said it better than I, but I cannot be
silent.
This is NOT a workaround.
Your fix is the correct way of doing it.
Richard Kryka's initialization example is good, too.
I don't know why the sudden difference in performance, though.
It is curious that the sme
Would the new improved 10.1.2 fixtool work on 10.0 files?
I'm on 10.0.16, but if file structure wasn't changed - and (w/o reading
the fine manual) I doubt it was - then the new fixtool should be a nice
thing to have in my back pocket.
cds
From: Clifton Oliver
The word I got was that the first
Ray,
Excellent point, as always, about field- vs value-level -1 difference.
One question:
Since each X-1 = ... adds to the size of memory allocated for the
variable X, every so often X runs out of room and there has to be some
sort of linking or remapping. (Recall we're talking about very
0017: EXECUTE 'LIST.READU' CAPTURING REPLY
OrEXECUTE 'LIST.READU USER ':STATUS() CAPTURING REPLY
or even EXEUCTE 'LIST.READU INTERNAL USER ':STATUS() CAPTURING
REPLY
If INTERNAL keyword is used, login id is probably REPLY2,8.
(This, for UV.)
0018: REPLY =
(I forgot to add delimiters to the X array being built)
My original question about memory allocation and overhead while
incrementally appending, increasing string length still stands.
note 2 changed lines below:
From: Stevenson, Charles
X = SPACE( 100 ) ;* reserve a million bytes
I'm surprised there are no answers yet.
The WHEN keyword is what you're looking for.
And you need to make sure the dictionary is set up right.
If ITEM.ID, STATUS.FLAG, CUBES, TOTE.ITEM, UOM are all D- or I- items,
then for each of them set
6 = M
7 = BOB.ASSOC
The same for any other fields
can't find one.
For UV SYSTEM(9001) and I'll dust off this old post from a couple years
ago:
-Original Message-
Subject: [UV] SYSTEM(9001) tells you what program you are in where you
came from!
From:Stevenson, Charles ([EMAIL PROTECTED])
Date:03/20/03
After whining about
VLIST the program and make sure source matches object
Any include files that would confuse the debuggers line numbering?
Anyone see why I cannot set a breakpoint at line 60?
SIS907: 59:IF BAL.LIST1,ARI.PTR # 0 THEN
SIS907: 60: ARI.ID =
...use $OPTIONS EXT to get the REMOVE to work on D3...
I recall Sequioa had REMOVE but the delimiter flag variable had
different values.
I think 1 meant end of array, for example.
e.g. REMOVE A FROM X SETTING D D=1 meant you reached the end.
---
Back to U2. 4 comments:
1. As a
A lot of specifics seem to be lost flushed by that time. SYSTEM(9001)
is useless.
It's frustrating.
cds
From: Nick Cipollina
Is there a way to capture the specifics of the error using ON.ABORT?
Is there a way to override error handling in Universe to do
something
other than display a
Noah,
Not directly, but *if* File2 is indexed by its FILE1_ID attribute there
are a couple ways.
Assume that I have 2 files, such as customer and invoice
File1: has 1 attribute, ID
File2: has 2 attributes, ID and FILE1_ID
I have a savedlist with ID numbers from file1.
I want to select
I thought you wanted a list of File2 IDs.
Won't the MERGE.LIST INTERSECTION give you a list of File1 IDs?
Great! That is exactly what I needed.
I forgot about the INTERSECTION option.
Noah
-Original Message-
(If BLORP is your savedlist from File1) GET.LIST BLORP TO 1 SELECT
READNEXT ID FROM LIST ELSE ID = @AM
But why do you use an odd convention like ID = @AM ?
because it would be impossible for ID to be an attribute mark via
readnext.
If you later test for ([EMAIL PROTECTED]) there's only one way that could have
happened.
I'll admit to occasionally doing
A known cause of UV error 040019 (my guess for Anmol is
Variation-on-the-Theme #2):
When processA is holding the update lock on a record,
and processB issues a WRITE to that record,
then processB will wait for the lock for 20 minutes (non-configurable),
then abort or take ON ERROR clause with
George,
It should be there, but it ain't.
But for this one it is easier to roll your own verb, than any other
course of action.
Just flesh out the following with error handling, @SYSTEM.RETURN.CODE,
and maybe a get(arg.) for FROM [listno] for numbered select lists,
@SYSTEM.RE
CT CDS.BP
I'll second Mats, and move to add an amendment:
iii)
SELECT PRODS
and then
READNEXT PROD ELSE DONE = TRUE
iii) extracts fields ( attributes ) .
READNEXT PROD ELSE ...
extracts the ENTIRE 1st VALUE of each field (i.e., including ALL
SUBvalues of that 1st value) but ignores any other
REQUIRE.SELECT (or its synonym SELECT.ONLY) is the keyword you are
looking for.
QUERY = \SELECT PO.GROUP = 'Group Reference'\
DATA \QSELECT PO.GROUP (2)\
DATA \SSELECT POF BY GROUP BY-DSND PO.DATE\
EXECUTE QUERY
Well, this works absolutely fine as long as the user enters
in a group
In my opinion...
NO : Retrofitting TRANSACTIONs into existing programs would be a
nightmare.
YES: Definitely use them when writing a new application from scratch and
even a new sub-system enhancement if it is pretty much self-contained
with limited, well-defined interfaces that write into the
From: Les Hewkin
What would you call a really big transaction?
A transaction should encompass all updates that should happen (or be
abandoned) as a group in order to insure logical data integrity at the
application layer.
Let me use an example, keeping the preceding statement in mind.
There are 10 kinds of people in this world: those who understand base-2
and those who don't.
(I think I heard that from someone on this list. Can't remember who.)
BTW, why did microsoft assign (-1) as true. I don't think
(1) works as
true and, while being 'not zero', is hard to use
OK, I'll bite.
How does one find a situation where one wants this?
From: Mark Johnson
Is there an OCONV method that does nothing, ie IF
X=OCONV(X,CONV) THEN TRUE.
I've had a situation where I could use this before and one
crept up today.
How does one find MORE THAN ONE situation where one
Fine by me, but get your own thread; this one's taken.
(Or at least add another sentence to the subject line.)
From: Clifton Oliver
...Control numbers, on the other hand, are a completely different
issue
Anyone want to discuss why most MultiValue systems don't know
the concept of
From: shuchi bhandari
i had another query reagrding the output of LIST.READU command.
On running the command, we got multiple lines of records that
were exactly similar i.e. all the parameters inncluding the
inode number; process id and Item id were the same.
Is this a case of multiple
Barry,
Could you post your test program?
thx,
cds
-Original Message-
From: Barry Brevik
for UniVerse, TRANS() ( aka XLATE() ) will cache a record
behind the
scenes so that a 2nd I-descriptor that TRANS()s to the same record
Please excuse me for butting in. That is very
Martin,
Thanks for that:
The TRANS cache only operates in the query processor, not
Basic programs.
Would you be willing to tackle any of these other questions? :
- Is T-correlative smart enough to do the same caching as TRANS()?
- The caching is only 1 record 1 file deep,
just one
Always good to get answers from both Glenn Martin. I thank you both.
But combining the 2 answers :
- Is T-correlative smart enough to do the same caching
as TRANS()?
{glenn} A T-correlative calls TRANS under the covers.
The t-correlative is parsed and then
If more people care, they whould weigh in.
Clif moved it into the U2UG Enhancement committee's queue.
I'm on that committee so I'll take it up there, but the more voices the
better.
-Original Message-
From: Bill Haskett
Charles:
No, you guys aren't the only ones. This is a
I don't think that will work, because the WRITEV is still inside the
transaction.
As a matter of fact, I think it will error out if there is no explicit
readu preceding it. That might be isolation-level-dependent, but I
don't think so.
LOCK 50
READV SeqNo FROM FILENAME.FV, 'CTRL', 1 ELSE
Glenn seems to have trouble posting.
I got this answer off-list and I haven't seen it show up yet on-list, so
let me forward it to the rest of you:
From: Glenn Herbert [mailto:[EMAIL PROTECTED]
Sent: Thursday, June 16, 2005 12:12 PM
To: Stevenson, Charles
Subject
One might think there was no useful info in the attached post, but then
one notices that the From address tells us that Glenn Herbert now has
an IBM address.
If one assumes Glenn still has some association with DataStage, one may
deduce that there may have been some truth to the rumour that
I should already know the answer to this but...
for UniVerse, TRANS() ( aka XLATE() ) will cache a record behind the
scenes so that a 2nd I-descriptor that TRANS()s to the same record
(maybe different field) will not have to do the disk access again.
Is T-correlative smart enough to do the
From: Brutzman, Bill
[snip]
crt @(31,19) : Qty.Rcvd M
[snip]
I am writing to inquire about the meaning of M. Suggestions would be
appreciated.
Looks like M is a variable, so it would depend on what it contained.
I'm going to guess that M = R#9
ct CDS.BP BILL
From: Brutzman, Bill
[snip]
crt @(31,19) : Qty.Rcvd M
[snip]
I am writing to inquire about the meaning of M. Suggestions would be
appreciated.
Looks like M is a variable, so it would depend on what it contained.
I'm going to guess that M = R#9
ct CDS.BP BILL
I'm reading through all these responses and probably won't respond to
each individually.
But I do thank you all for your input.
There are several suggestions like the one below that involve
incrementing the control record BEFORE the transaction starts. Either
grab a block, or grab a master
This is a pretty good idea, Ken.
Getting current date time is probably cheaper than reading and
incrementing a control record. If time is in milliseconds, uniqueness
is even more likely so likelihood of locking a new record with SEED_VAL=
000 is good.
date()*1 + time() might be faster than
From: Anthony Dzikiewicz
So, what if you change the READVU to a READVL ?
I don't understand what READL buys me. (Not arguing, just looking to
understand.)
I thought you need a READU in order to do the write within the
transaction.
By the way, I've only been working with ISOMODE 1, if that
Missing sequential numbers isn't important to me, but it's a good point
for general discussion.
cds
From: Bruce Nichol
This is all well and good if the commit goes ahead, but if
rollback is the action, don't you lose a supposedly
sequential root key into the vapours?
Never to be seen
Ah-HA! Some real world transaction experience!
Just what I wanted, Les.
From: Les Hewkin
Start trasnactions
Readu control record then
record.ID = control record
write record
control record += 1
write control record
end
end transaction
This seems to work
From: Bill Haskett
I wonder if the same issue (transactions don't release record
locks until they're committed) applies to the LOCK/UNLOCK statement?
No, those are independent of transactions.
So could they be leveraged for the question posed here?
. . . I don't see it.
cds
---
u2-users
Would this be a worthwhile enhancement request through U2UG?
Are Ken and I the only ones whining about this:
From: [EMAIL PROTECTED]
Ideally of course, the ID would be allocated by the DBMS
transparently when the record is inserted into the file, but
neither U2 product supports that concept
From: Brian Leach
...Windows service app) in Delphi... Mutex wrapper... thread safe.
... Socket API or SOAP API...
smart aleck.
Seriously, if you are going to use transactions I would guess
the only way to avoid maintaining the key for the duration of
the transaction (to go back to the
Charlie,
I don't see how that addresses the TRANSACTION problem.
If the call to your NEXT.AVAIL.ID routine happens after TRANSACTION
START, the id control item will not be updated nor will its lock be
released until TRANSACTION COMMIT (or rollback) is executed, even though
the NEXT.AVAIL.ID
Bill,
The pooh-syndrome has struck me again. I think I've missed your point.
How does this apply to the TRANSACTION scenario?
Oh, wait. You dropped that from the subject line. So this is a new more
general thread?
Maybe we need a contest for the best
For files with sequential numbers as record IDs, it is common practice
to use a control item (in dictionary or in a special control table) to
get the next sequential id counter, then write back the incremented
counter ASAP so it is available for the next process that needs to do
the same. Some
I feel like Winnie-the-Pooh here, Ray. I'm not following you.
Are you saying reserve a block of N IDs before starting the
transaction, by incrementing by N the traditional next-seq-id control
record thereby freeing the ctrl item for use by other processes?
What's the point of the 2 sequential
One thing I would add to this is to always check to see
if the key exists on the file that you are writing to
before stepping on an existing record.
I'm not concerned about fleshing out my little READVU SEQ.ID . . .
WRITEV SEQ.ID+1 . . .
Yes, you have to check for existing records and worry
How about replacing:
Some variation on this theme (probably involving a utility
subroutine):
READVU SEQ.ID FROM ctrl.fvar, ctrl.id, n ...
WRITEV SEQ.ID+1 TO ctrl.fvar, ctrl.id, n
WRITE NEW.REC TO fvar, SEQ.ID
with some variation of:
OPENSEQ path TO ctrl.fseq ELSE ...
READSEQ
Adrian,
Could you provide an example program of it not working? Which version
of UV? I am unaware of the problem you describe.
There is a bug with REVREMOVE, (reverse-remove through a dynamic array).
It was partially fixed at 9.6 or 10.0 (I forget when), but not quite.
The list archives should
From: John Jenkins
Try selecting with NO.INDEX selected
I did, but by that time it was acting normal every which way, so I
didn't learn anything.
- and try it on each part file individually.
did, but see above.
It could be the GTAR but I have seen oddities with one of:
1. A Index based
1st, we do not use UV/NET. This happened on a stand-alone UV system
HPUX 11i, UV 10.0.16.
The error (Exhibit A) occurred SELECTing or LISTing a distributed file
(Exhibit C), or either of its 2 parts, with the selection criteria based
on an indexed field. This is from TCL on a
Bill Diane,
I'd hold suspect just about anything this teacher said.
That is - to be polite - a unique way of thinking about EQUATEs.
There are not 2 slots in memory to update simultaneously.
I believe EQUATE on Microdata, UV, and ALL MV-Basics are substantially
equivalent in that EQUATE is an
Yes , very helpful, but those should be EQUATEs, not assignments.
Equates of quoted texts, CHAR(27), and even concatenations of these will
all be resolved to strings by the compiler. Nothing gets executed at
runtime.
If the file as written were $INCLUDEd in a utility routine that gets
called
I use the method my dad taught me when I was a kid, cuz I couldn't
remember Thiry days hath September
Hold 2 fists side-by-side, forefingers of right left hand touching.
Each knuckle and each valley between knuckles represents a month.
Start at the left pinky knuckle and count off months.
If
From: Les Hewkin
Can we have a quick show of hands of those who have had
there U2 system hacked?
I was accused of it once by a system administrator because I pointed out
the security hole that I was then accused of wriggling through. This
was the same system administrator who once ended up as
If you execute a READU hit the LOCKED clause, STATUS() will tell you
the USERNO of whoever holds the lock you want.
But what is the best way to tell what _kind_ of lock is blocking you?
E.g., is it a record lock or a filelock?
The only way I can think of is to dig through the results of:
Thank-you, Perry!
That's exactly what I wanted.
I've used it before, just forgot it, couldn't find it.
What a wonderful list,
cds
-Original Message-
From: Perry Taylor
I think you can use the RECORDLOCKED() function.
-Original Message-
From: Stevenson, Charles
If you execute
Mike,
You've probably already figured it out or been told, but if you use a
numbered select list in your innermost executed select or selectindex
inside the i-desc subroutine, I'll bet you'll be ok.
I favour a standard that says select list 10 is a very temporary list
that a low level function
...Allow re-dimensioning of dimensioned arrays at runtime...
As others have pointed out, YOU CAN
$OPTIONS -STATIC.DIM
Minus sign means not as in not static dim as in you can redimension
this on the fly at runtime.
It is necessary in Pick- Reality-flavors.
PI- and Ideal-flavors have that
201 - 300 of 417 matches
Mail list logo
