* copyright (C) 1984-2019 merrill consultants dallas texas usa

Image object the name includes the full file path name of

the loaded modules, while Image has only the filename.

Oct 24: The new ADSM object had no test records, so its

INPUT statement had not been executed. Compiling

under SAS V8 did not detect an MXG coding error

(there were no comments around the label text in

the ADSM INPUT statement), but SAS V6 did detect

an error at compile, but only accidentally: both

versions of SAS saw each word in the comment as a

variable to be input, but under V6, "variable"

TRANSFERRED was MORE THAN 8 CHARACTERS long.

CHARACTERS! This MXG coding error would have

been eventually detected, when test records were

read, but by using both V6 and V8 in the MXG QA,

yet another, later, change has been avoided!
Change 18.244 Candle's Omegamon for VTAM TCP record documentation was

VMACOMVT incorrect; revised DSECT was received and the code for

Oct 1, 2000 TCPA and TCPB datasets now matches the data records.
Change 18.243 While writing my CMG Workshop Paper on BUILDPDB, I was

MXGSAS stunned to find REGION=4096K on the EXEC PGM= statement

Sep 28, 2000 in the MXGSAS JCL Procedure that I distribute! That

should have been removed long ago. You should specify

either REGION=0 or REGION=64M on your JOB card.
Change 18.242 MXG 18.05-18.08 only. CICSTRAN variable PROGRAMB was not

VMACCIMS kept because it was spelled as ROGRAMB in the KEEP= list

VMACEDGS fortunately PROGRAM is the same as PROGRAMB, so the name

VMACSRMH is still there, and this was not reported, but instead

VMACLDMS was detected by Freddie continued source analysis. Other:

VMACRSDF -VMACIISL - Include of IMACZDAT added to create ZDATE.

VMACIISL -VMACEDGS - MDNDS removed from KEEP= list

Sep 28, 2000 -VMACSRMH - _KSRMNNU changed to _KSRMHNU

Oct 23, 2000 -VMACLDMS - corrected to PCDJJNAM,PCDJPSTN,PCDJCOND,

PCDDRLNG,DIRSVERS,DIRSCTL1 which were missing

the last character in KEEP= list, and LPCABR

changed to LPCAGR.

-VMACRSDF - RSDUNKN1-RSDUNKN4 instead of 2-5.

Oct 23: %%INCLUDE for IMACZDAT.

Thanks to Freddie Arie, TXU Utilities, USA.
Change 18.241 Macro _LNTTN32 was not defined, creating dataset _LNTTN32

VMACNTSM instead of TN3270SV. Insert, before MACRO _WNTTN32:

Sep 28, 2000 MACRO _LNTTN32 &PNTTN32..TN3270SV %

Thanks to Greg Jackson, National Life of Vermont, USA.

Change 18.240 For SAS user SMF record, new format $MGSASPR maps the

FORMATS variable SASPROC (Procedure Name) to the new variable

VMACSASU SASPROD (SAS Product Name that owns that Procedure).

Sep 27, 2000

Thanks to Raff Rushton, Kaiser Permanente, USA.
Change 18.239 Cosmetic. Freddie's post-QA analysis of every line of

VMAC90A MXG source code continues to get better! This iteration

VMAC28 identified variables in the KEEP= list in MXG source

UTILXREF members that were never referenced:

UTILXRF2 -In VMAC90A, variables NEWWKLD and MCPNAME in the KEEP=

Sep 26, 2000 list should be spelled NEWFWKLD and MPCNAME. This does

not cause an error, but those variables did not exist in

the output dataset because of the spelling error; only if

you went looking for those fields would you have noticed

the did not exist.

-And when he ran under SAS Version 6, he got VMAC28 NOTE:

LABEL VALUE FOR VARIABLE NITIFODP HAS BEEN TRUNCATED TO

A LENGTH OF 40, but that note did not appear under my SAS

Version 8 QA run, because V8 permits 256 byte Labels!

I have always limited my LABELS to 40 positions, only so

that you didn't get that TRUNCATED message on your log.

In fact, it is only by searching the SAS Log as part of

my own Post-QA analysis for the string TRUNCATED that I

find and correct these long labels in MXG source.

But now that the message goes away in SAS V8, should I

still keep LABELS to only 40 characters? I say YES:

- For sites still running under SAS V6, eliminating any

unneeded or non-impacting message might save me a tech

support call/email, and will save new users confusion

and wasted time.

- Since PROC PRINT SPLIT='*' is the MXG way to print the

label as column heading, if I were to increase label

length, it could mess up our nicely arranged reports.

Besides, you can always use your own LABEL statement

with your PROC PRINT if you want a wordier heading.

- MXG ADOCxxxx and DOCVER members only have room for 40.

- Forty has been enough to describe the first 112,498

variables I've created in MXG.

So I have modified my UTILXREF/UTILXRF2 programs to

detect any labels longer than 40 bytes as part of the

MXG QA stream, and shorten them before you see them.

Thanks to Freddie Arie, TXU Utilities, USA.
======Changes thru 18.238 were in MXG 18.08 dated Sep 25, 2000======
Change 18.238 The new logic fails when the SPIN.SPINUOW dataset exists,

ASUMUOW because a VIEW cannot be used when the input and output

VMXGUOW datasets are the same. Remove /VIEW=_TMPSPIN from the

Sep 25, 2000 DATA _TMPSPIN; statement.

Sep 26, 2000 Members IMACKEEP and IMACUOW were removed from %INCLUDE

Nov 17, 2000 in member ASUMUOW, and IMACUOW was added to the existing

INCLUDE of IMACKEEP in member VMXGUOW, so that MACKEEP is

only invoked once.

Sep 26: If ASUMUOW were executed in a separate step, and

not in the step that built CICSTRAN, the program fails

with CHAR/NUM conflicts, because macro _LCICTRN did not

exist. Now, new logic validates the existence of that

macro, and defines it with the MXG default if not found.

Nov 17: Under SAS V8, the error message was:

YOU CANNOT OPEN SPIN.SPINUOW.DATA FOR OUTPUT ACCESS

WITH MEMBER-LEVEL CONTROL BECAUSE SPIN.SPINUOW.DATA

IS IN USE BY YOU IN RESOURCE ENVIRONMENT SASDSVX.

Bottom line: Scratch and Reallocate the SPIN Library.

Thanks to Hank Oerlemans, Western Australia Police Service, AUSTRALIA
Change 18.237 The TYPS102A member was expected by the test job TESSIBM3

TYPS102A but that member did not exist.

Sep 22, 2000

Thanks to Dr. Alexander Raeder, Karstat AG, GERMANY.

Thanks to Harmuth Beckmann, Karstat AG, GERMANY
Change 18.236 Support for Landmark TMON for VTAM creates datasets:

EXTMVTAA dddddd dataset

EXTMVTAE TMVTAA TMVTAA

EXTMVTAS TMVTAE TMVTAE

EXTMVTIB TMVTAS TMVTAS

EXTMVTIN TMVTIB TMVTINB

EXTMVTSE TMVTIN TMVTIN

EXTMVTSI TMVTSE TMVTSE

EXTMVTSS TMVTSI TMVTSI

EXTMVTTN TMVTSS TMVTSS

EXTMVTVR TMVTTN TMVTTN

IMACTMVT TMVTVR TMVTVR

TYPETMVT Records AA, IN, SI, TN, and VR have been tested; the

TYPSTMVT TN records were discovered to have some fields that are

VMACTMVT accumulated, the _STMVTTN macro will deaccumulate TMVTTN

VMXGINIT dataset, and _STMVTTN is added to the TYPETMVT member so

Sep 21, 2000 that dataset TMVTTN is always deaccumulated; using the

TYPSTMVT member is recommended, since the TYPSxxxx member

invokes the _Sdddddd sort macros, which is where the MXG

DIF() logic for deaccumulation is invoked.

Thanks to Siobhan Hansen, Merrill Lynch, USA.
Change 18.235 Summarization and trending of STC datasets.

ASUMSTC Input dataset(s) Summarized dataset

TRNDSTC PDB.STCLMU PDB.ASUMLSM

Sep 21, 2000 PDB.STCENTER, PDB.STCEJECT PDB.ASUMENEJ

PDB.STCHSC PDB.ASUMHSC

Thanks to Chuck Hopf, MBNA, USA.

Change 18.234 This analysis uses SYNCSORT SMF records to identify jobs

ANALUAFF that used tape-to-tape sort steps, but that did not use

Sep 21, 2000 UNIT=AFF=SORTIN on the //SORTOUT dd statement, causing

the job to use an extra tape drive.

Thanks to Al Holz, Merck, USA.
Change 18.233 Both VMXGDUR and VMXGSUM now have new argument SYNC59 to

VMXGDUR add one minute to DATETIME for Interval calculations, so

VMXGSUM that sites which write at :59 minutes can have the start

Sep 21, 2000 of the interval be one minute later.

Thanks to Ian Davies, Canadian Utilities Ltd, CANADA.
Change 18.232 NETSPY 'I' record with NSPYENTL=124 caused STOPOVER ABEND

VMACNSPY because MXG expected NSPYENTL=170. The INPUT statement

Sep 20, 2000 is interrupted after variable OPTSEGSZ and the remaining

fields are only INPUT if NSPYENTL GE 170.

Thanks to Mark Bailey, HM Land Registry, ENGLAND.
Change 18.231 Support for Landmarks The Monitor for DBCTL creates four

ADOCTMDC new datasets from the 'CN', 'CT', and 'CX' records.

EXTMDCCN These new datasets are created:

EXTMDCCV dddddd DATASET Description

EXTMDCCT TMDCCN TMDCCN Interval Statistics from CN record.

EXTMDCCX TMDCCV TMDCCV VSAM Intervals from CN record

IHDRTMDC TMDCCT TMDCCT PSB/Thread Detail from CT record.

IMACTMDC TMDCCX TMDCCX Exception record

TYPETMDC See member ADOCTMDC for notes on some data values.

TYPSTMDC These TMON records can be dumped in compressed format,

VMACTMDC and MXG member EXITMON6 is the code and instructions to

VMXGINIT install the TMON INFILE exit so MXG will read compressed

EXITMON6 TMON records directly; comments and examples in EXITMON6

Sep 19, 2000 were revised and updated for all TMON products.

Thanks to ???, ???, EUROPE

Thanks to Daniel Strgarsek, SAS EMEA, GERMANY.

Change 18.230 Cosmetic. The LABEL for variables PGPEXCP and PGPIOTM

VMAC7072 are now *BY THIS*PERFGRP/SRVCLASS instead of the old

Sep 19, 2000 PERF GROUP PERIOD. These variables exist in both TYPE72

and TYPE72GO datasets, so their contents were clarified.

Thanks to Joe Martin, Amdahl, USA.
Change 18.229 Support for Omegamon for VTAM V500 (COMPAT) adds new

EXOMVTCA TCP/IP measurements in four new MXG datasets:

EXOMVTCB dddddd Dataset Description

EXOMVTCC OMVTCP OMVTTCP TCP/IP Address Space

EXOMVTCP OMVTCA OMVTTCPA TCP/IP Application

IMACOMVT OMVTCB OMVTTCPB TCP/IP Buffer Pool

VMACOMVT OMVTCC OMVTTCPC TCP/IP Connection

VMXGINIT Only the first three datasets have been tested with data.

Sep 19, 2000

Thanks to Eric Barnes, Prudential, ENGLAND.

Thanks to Norman Hollander, Candle, USA.

Thanks to Dave Crandall, Farmers Insurance, USA.

Change 18.228 Add logic to MXGCPU report for number ICF's and CP's

ANALRMFR within the Partition data report.

Sep 18, 2000 Error PDB not assigned, when parameters are PDB=SMF,

REPORT=IOQU, and PDBOUT=PDBO was corrected.

Change _STY78 to _STY78CF;

Add _STY78CU; _STY78IO;

Error Physical file does not exist. When parameters are

PDB=SMF,REPORT=WKLD,PDBOUT=PDBO,RPTOPT=PERIOD corrected.

Change all _N72 to _N7072.

Change all VMAC72 to VMAC7072.

Change all _CDE72 to _CDE7072.

Change all _VAR72 to _VAR7072.

Change 18.227 Support for EDA, Enterprise Data Access, SMF user record.

EXEDALOF Two datasets are created from the four subtypes:

EXEDALON dddddd Dataset Subtype Description

FORMATS EDALON EDALOGON 1,4 EDA LOGON OR BEGIN QUERY

IMACEDA EDALOF EDALOGOF 2,5 EDA LOGOFF OR END QUERY

TYPEEDA The EDALOGON dataset seems useless, since it has only the

TYPSEDA SMFTIME of the start, which becomes MXG variable EDASTIME

VMACEDA in the EDALOGOF dataset. Sorted by EDASTIME, EDALOGOF

VMXGINIT has the 2:LOGOFF event record first, which has the total

Sep 18, 2000 CPU Time and EXCP count, and then each of the individual

5:END QUERY event records follow that LOGOFF record. The

CPU time in the 5:END QUERY records is already contained

in the 2:LOGOFF even record.

Thanks to Glen Yee, California Health & Human Services, USA.

Change 18.226 Variable JOB was blank if QMDACORR was blank, for DB2

VMACDB2 attachment QWHCATYP IN (1,2,0Bx) (TSO, CALL ATTACH, and

Sep 15, 2000 UTILITY JOBS). For those QWHCATYP values, MXG now tests

IF QMDACORR GT ' ' THEN JOB=QMDACORR;

so JOB won't be overwritten if QMDACORR is blank.

Thanks to Diane Parker, Bergen Brunswig, USA.

Thanks to Warren E. Waid, JC Penny, USA.
Change 18.225 Change 17.303 was wrong. The DBORG='00'x segment does

VMACCIMS not contain SQL information, and should not have been

Sep 15, 2000 output to CIMSDB2. That segment (with DBDNAME='ALLDBS')

is the I/O activity for DLI buffers during Syncpoint

processing and is now output back into CIMSDBDS (as it

was before Change 17.303), and its I/O counts are summed

into the CIMSTRAN observation for that transaction. It

is called IOWAITS because only IOWAITS code generates

statistics during this processing.

Thanks to Thomas Heitlinger, FIDUCIA IT AG, GERMANY.

Change 18.224 The DCOLLECT _Sdddddd and _Sxxxx SORT macros are updated

VMACDCOL as PROC SORT NODUP and the _Bdddddd BY list will remove

Sep 12, 2000 duplicate observations in the input DCOLLECT data.
Change 18.223 MXG 18.02-18.07 GOAL MODE only. Report 2 printed too

UTILRMFI much because test for RPRTCLAS=Y should have been='Y',

Sep 12, 2000 and reports were revised to use PDB.SMFINTRV/PDB.STEPS

or TYPE30_V/TYPE30_V instead of interval and Job Term,

and the comments were clarified.

Thanks to Bob Falla, The Mutual Group, CANADA.

Change 18.222 Support for AS/400 Collection Services records, which is

EXQAPCON a separate product, different from the AS/400 Performance

EXQAPDIS Monitor that is still available (free) and supported by

EXQAPJOB MXG's TYPEQAPM code. This new product captures only a

EXQAPPOO small subset of the data in the Performance Monitor.

IMACQACS This new TYPEQACS support creates these datasets:

TYPEQACS DDDDDD INFILE/DATASET Description LRECL

TYPSQACS QAPCON QACSCONF CS Configuration data 16

VMACQACS QAPDIS QACSDISK CS Disk Storage data 352

VMXGINIT QAPJOB QACSJOBS CS Job data 544

Sep 14, 2000 QAPPOO QACSPOOL CS Main Storage data 78

Aug 21, 2008 (QACSYS/QACSYSM/EXQCSSYS reference/member removed 2008.)

The AS/400 CS data is completely dependent on the correct

LRECL having been specified when the data was uploaded

to MVS, or in the FILENAME statement on ASCII systems.

Member VMACQACS documents the correct LRECL sizes.

Thanks to Joseph J. Faska, Depository Trust, USA.

Thanks to Colin Bowen, CSC, SOUTH AFRICA.

Thanks to Stuard R. Burnett, Reynolds Metal Company, USA.
Change 18.221 Cosmetic. A single byte value, 10995116167810048 was

FORMATS skipped in the MGBYTES format and would not have been

Sep 12, 2000 formatted; the range for TB was increased to include.

Thanks to Danny Ball, CNF, USA.

Change 18.220 SAS V8.0 (TSM0 and TSM1), but not SAS V8.1. SAS/GRAPH

DOC "missing values" warning message in SAS V6 became a real

Sep 9, 2000 _ERROR_ condition in SAS V8.0, causing either CC=12, or a

USER 999 ABEND (because MXG sets option ERRORABEND so

that any true error causes an ABEND instead of just a

condition code), if any graph has missing values for one

variable! This has been corrected in SAS V8.1; only a

WARNING message is printed on the log, as it should have

been. If you are still at SAS V8.0, you can specify

OPTIONS NOERRORABEND; for the steps that use SAS/GRAPH,

and the USER 999 will be avoided, although the condition

code will still be non-zero.

Thanks to Caron Knox, Willis Group Services, ENGLAND.
Change 18.219 ERROR: CHARACTER VARIABLE QW0140TX HAS TOO LONG A VALUE

VMAC102 THE SASEB ENGINE when you execute BUILDPDB or TYPE102

Sep 8, 2000 under SAS V8, but point your output to a pre-existing

V6-built SAS data library (i.e., your "PDB"). You should

build V8-format data libraries when executing under V8;

you must allocate new physical OS data sets and write to

them: you CANNOT use PROC DELETE to clear a V6 library

and then write to it with V8 to create a V8-built data

library, since the SAS Directory is still V6-built.

See SAS Technical Note 11 in Newsletter THIRTY-SEVEN.

However, if you intentionally must keep your PDB in the

V6-format, executing under SAS V8, you can circumvent

the above error by using %LET SASCHRLN=100; as your

first //SYSIN statement.

Thanks to Caron Knox, Willis Group Services, ENGLAND.
Change 18.218 Report macro failed if PDB=SMF and PDBOUT=PDB was used.

ANAL94 The internal logic was revised so that macro &PDB was not

Sep 7, 2000 reset.

Thanks to Steve Gossage, Cummins Engines, USA.

======Changes thru 18.217 were in MXG 18.07 dated Aug 31, 2000======
Change 18.217 First MXG 18.07 only. First day's tape (dated Aug 30).

BUILDPDB Member BUILDPDB was missing the _S21 invocation, causing

Aug 31, 2000 dataset PDB.TAPES to have not been built.

Thanks to Bruce Widlund, Merrill Consultants, USA.

======Changes thru 18.216 were in MXG 18.07 dated Aug 30, 2000======
Change 18.216 Erroneous "BAD HFS" segment messages were produced from

VMACTCP SMF 118 errors with legitimate data; COL-LENGTH should

Aug 30, 2000 have been LENGTH-COL, and LT LENGTH-1 should have been

LE LENGTH-1.

Thanks to Bruce Widlund, Merrill Consultants, USA.
Change 18.215 Each INFILE statement was enhanced with END=ENDOFTAN

VMACTAND so that the end of input file is flagged, as this can

Aug 30, 2000 be used to simplify duplicate data checking in ITSV.

Thanks to Peter Christie, IT Performance Consulting, USA.

Change 18.214 INPUT STATEMENT EXCEED RECORD LENGTH for SMF ID=65 record

VMAC6156 because the '04'x Catalog Segment in the Catalog Record

Aug 30, 2000 that is appended to the SMF record without documentation,

now has Low and High Key field lengths of 112 bytes when

previously that field was only 64 bytes long. The IBM

development programmer who "owns" these catalog records

that are output in the type 61, 65, and 66 records has

refused to make their documentation available ("its OCO")

even for this Vendor, so I have to stumble into changes

in field lengths in the data in the field. The logic to

read those varying length fields was revised to protect

for whatever length is now found.

Thanks to John Doan, TotalSystem, USA.
Change 18.213 Support for BMC MainView for MQ Series History File V2.1.

EXBBMQCH Two new subtypes, 'E4' and 'E5' are now supported and new

EXBBMQLM data was added to the existing 'E1' Queue Manager record.

IMACBBMQ DDDDDD Dataset Description

VMACBBMQ BBMQCH BBMQCHAN 'E5' CHANNELS RECORD

VMXGINIT BBMQLM BBMQLMGR 'E4' LOG MANAGER RECORD

Aug 29, 2000 These data sets have been created and sorted, but only

Sep 5, 2000 just examined; the interval DURATION does not exist in

these records, and I can't tell if the values are exact

or whether they are accumulated until I look at more

data, but the datasets and variables look good.

Sep 5: CHLLRTRY and CHLSEQHI are now stored in length 8,

because their default hex value of '3B9AC9FF'x=999999999

became '3B9AC900'x internally when stored in length 4.

Thanks to Martyn Jones, ABN AMRO, THE NETHERLANDS.
Change 18.212 Cosmetic. Spelling in comments of asterisk are now all

DOC the same; variants asteric, asterics, and asterisk were

Aug 29, 2000 corrected in many places. And occurrences of occurance

Sep 5, 2000 and occurence were corrected, too!

Thanks to Bruce Green, MIB, USA.
Change 18.211 This "BUILDPDB" utility now knows about the inconsistent

UTILBLDP spelling of the SMF "MACRO _IDxxxx" in 17 MXG members.

Aug 29, 2000 While most of the SMF ID macro names are _IDxxxx, these

products have non-standard macro names, typically _xxxxID

and they are now properly generated if you request these

products be added to the BUILDPDB code built by UTILBLDP:

ARB, DLM, DMON, HBUF, HURN, IPAC, M204, NSPY, PDSM, RMDS

ROSC, RTE, SYNC, TSOM, VVDS, WYLA, WYLB, X37.

Thanks to Ian Davies, ATCO I-Tek, CANADA.
Change 18.210 Documentation. Support for Netview 1.3 SMF records

VMAC37 type 37, 38, and 39 has been in MXG since before 16.16,

VMAC38 but there was no MXG note to that effect. No changes

VMAC39 were made by this change.

Aug 29, 2000

Thanks to Harald Seifert, HUK-COBURG, GERMANY

Change 18.209 Cosmetic. The "LAST RECORD IN GROUP" message was not

VMACSMF written if the last record was not a subtype 3 record,

Aug 29, 2000 as when reading only part of an SMF file.
Change 18.208 Y2K Error. MainView for CICS Statistics SMF type 110

IMACICBB special subtype 0B02x records cause year 2020 in 2000:

VMAC110 -in the IBM Statistic Header COLLTIME/STARTIME variables,

Aug 28, 2000 because the BMC type 110 subtype 2 record for CICS 4.1

did not update the SMFMNMFL Maintenance Level field.

That field was updated for IBM type 110 subtype 2 by

PN69653, which changed the date format from YY to YYYY.

With the MFL still zero, MXG assumed the old YY format

and generated the wrong dates. As it's too late for

BMC to fix their error, MXG now tests the ORIGSUBT to

recognize 4.1 BMC records with the YYYY date format.

Logic in VMAC110 was enhanced to protect BMC.

-in the unique datetime variable in all CICSBBxx datasets,

the BMC documentation showed binary (PIB) for their new

YYYY,MO,DD dates, but now with data records, those fields

turn out to be unsigned-packed (PK) format, so member

IMACICBB was corrected.

Thanks to Hartmut Peter, R+V Allgemeine Versicherung AG, GERMANY.

Change 18.207 Documentation for MXG Support for BMC MainView products.

DOC Member Infile Product/Description

Aug 28, 2000 ASMMNVW n/a INFILE Exit for Compressed BMC data

TYPEBBMQ BBMQVSAM MainView for MQ

TYPECMFV CMFVSAM MainView for OS/390, MMR, CMF-VSAM.

TYPECMF SMF MV for OS/390, CMF User SMF record.

TYPECIMS IMSLOG MainView for IMS

TYPEMVCI CMRDETL MainView for CICS CMRDETL/CMRFILE.

IMACICBB SMF MV for CICS Type 110-2 Stat records.

TYPEDB2 SMF MainView for DB2 (no unique records).

Change 18.206 Three TMON V2 Segments did not have the offset variable

VMACTMO2 in their INPUT statement, causing misalignment. The

Aug 27, 2000 offset variables that were added to their INPUT:

TRDSAOFF,TREXPOFF, and TRXMCOFF, for datasets MONIDSA,

MONIEXP, and MONIXMC.

Thanks to Alex Benny Nielsen, TeleDanmark IT, DENMARK.

Change 18.205A The xxxxCMFV members were named for "CMF-VSAM", because

VMACCMFV member xxxxCMF already existed (for the CMF user SMF

Aug 27, 2000 record), and the records came with the BMC CMF product.

Change 18.205 However, the records that xxxxCMFV processes are actually

created by the MMR component of the "New CMF" product,

but as the records are common to both the "CMF" and the

"MMR" products, they exist even if you have only "MMR"

(for example, if you use RMF instead of the "old CMF"

part of the "New CMF" product).

Five datetime variables in ZNTS format are now decoded