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

TYPETMMQ 27.145 Support for TMON for MQ record 'QA' (APPLICATION).

TYPETMNT 27.009 MSGID IEF233D supported, no DSNAME in SYSLOG doc'd.

TYPETMNT 27.336 SYSLOG message text size increased to 32384 bytes.

TYPETMO2 27.042 Support for ASG TMON for CICS V3.2.

TYPETMO2 27.091 ASG TMON/CICS variables WTSCWTTM,WTSCWTCN reversed.

TYPETMO2 27.215 TMON/CICS Version 3.1 INPUT EXCEEDED on 'TI' record.

TYPETMS5 27.111 New TMSLIB variable, support for multiple TMS cats.

TYPETMS5 27.168 MXG 27.05-27.06. _KTMSTMS dropped, impacts ITRM.

TYPETMS5 27.190 TMS.TMS DEVTYPE blank for 3590 and 3592 devices.

TYPETMVT 27.347 TMON/VTAM "SI" record Interval variables now INPUT.

TYPETPMX 27.093 TYPETPMX variable JESNR now 7-digits, was 5 digits.

TYPEULOP 27.029 Support for BMC's Ultra Op Product's User SMF record.

TYPEULTM 27.069 Serena's Ultimizer user MV moved subtype location.

TYPEVMXA 27.008 z/VM 5.2 RECORD ERROR SYTSYP/STORSP/STOSXP/PRCPRP.

TYPEVMXA 27.156 Support for z/VM 6.1.0 in MXG 27.01+; no new data.

TYPEVMXA 27.156 z/VM 6.1 support is in MXG 27.01 or later.

TYPEVMXA 27.264 z/VM MONWRITE example processes only USER domain.

TYPEXAM 27.030 Only the last MDISK was kept in XAMDEV.

TYPEXAM 27.070 Variable DESCR truncated in XMHSTMEM, changed.

TYPEXAM 27.151 XAM TCP record INPUT STATEMENT EXCEEDED error.

TYPEXAM 27.216 Variable SYTLPNAM restored kept in XAMSYT dataset.

TYPEXAM 27.281 Updates/corrections/ for XAM Version 3.7.

UDB2GTF 27.015 Revised Support for processing DB2 GTF records.

VGETDDS 27.083 Concatenate PDB libs, dynamically allocate them.

VGETDDS 27.248 Logic revised when DDNAMES= syntax is used.

VGETDDS 27.310 New DATEJUL= created DSNAMES with Julian YYYDDD.

VGETDDS 27.330 New DATEJUL= correctly generates julian dsnames.

VGETDDS 27.359 WAIT=N option protects for DSNAME already in use.

VGETOBS 27.237 Enhancement if no DDNAME argument.

VGETSYSI 27.049 New %VGETSYSI gets (z/OS only) SYSTEM, SU_SEC values.

VMACDB2 27.131 DB2STATS DIF() or no-DIF() corrections.

VMACEDGR 27.128 Syntax error after Change 27.046 (GT. GT.).

VMACORAL 27.306 Support for restructured ORACLE SMF records.

VMACSMF 27.058 SMF SUBTYPE GT 255 for BMC CICS subtype 2818/47874.

VMACSMF 27.341 WARNING: SUBTYPE GT 255 message now not defaulted.

VMXGALOC 27.355 NOWAIT added, create/allocates are now conditional.

VMXGCNFG 27.356 The standard SAS JCL Proc can be used for MXG.

VMXGDUR 27.214 New FLORCEIL=FLOOR/CEIL argument for begin/end calc.

VMXGDUR 27.214 SYNC59=YES now default with FLOOR, always safe.

VMXGDUR 27.308 SYNC59-NO/VMXGDUR/VMXGSUM (final?) enhancements.

VMXGOPTR 27.051 %VMXGOPTR changed, CURRENT vs ORIGINAL value is used.

VMXGOPTR 27.092 SAS V8-ONLY: OVERFLOW HAS OCCURRED after SUMSTATB

VMXGOPTR 27.092 SAS V9.2-ONLY: NO MATCHING %IF FOR %THEN.

VMXGOPTR 27.124 Macro %TRIM() function removed from VMXGOPTR.

VMXGSET 27.343 VMXGSET permits multiple datasets with APPEND=YES.

VMXGSUM 27.071 VMXGSUM-using programs support DROPed variables.

VMXGSUM 27.234 Revision to eliminate OUTCODE= argument sometimes.

VMXGTAPE 27.114 "Tape-aware" programs now support LIBNAME allocation.

WEEKxxxx 27.005 Weekly logic enhanced to support nonexistent dataset.

WPS 27.239 WPS 2.4 GA has been tested, requires MXG 27.09.
See member CHANGESS for all changes ever made to MXG Software.

Inverse chronological list of all Changes:

NEXTCHANGE: Version 27.
====== Changes thru 27.361 were in MXG 27.27 dated Jan 20, 2010========
Change 27.361 MXG 27.06-27.11. Variable BUFNO in TYPE1415 was always

VMAC1415 zero; the label in the comment for SMF14ABD, added by

Jan 19, 2009 Change 27.148, had a / instead of */ at the end of that

line, which swallowed (without error) the IF BUFNO test.

Note that even when fixed, BUFNO=0 occurs frequently for

non QSAM files; for example, TYPE1415 records for SAS

data libraries always have BUFNO=0 (because the access

method for SAS Data Libraries is EXCP Access Method).

Thanks to Tom Parquette, AXA Technology Services, USA.

Thanks to ???, ???, CANADA.

Change 27.360 Fixed in MXG 27.10, but in MXG 27.09, PDB.ASUM70PR could

VMXG70PR have PCTCPUBY much greater than 100%, for systems with

Jan 19, 2009 IFL engines, if the IFL also has the highest LPARNUM.

Similar to Change 27.123, fixed by Change 27.294/325,

but invalid values in PCTCPUBY were not mentioned in

the text of those changes.

This is only change text; no code was changed.

Thanks to Tee Brown, Blue Shield Blue Cross of South Carolina, USA.

Change 27.359 The new WAIT=N argument in %VGETDDS causes allocations to

VGETDDS be WAITed for N minutes if the DSNAME is already in use.

Jan 19, 2009 SAS tests every 15 seconds and if the DSNAME is freed in

those N minutes, the allocation proceeds as normal.

Thanks to George Pandzik, USAA, USA.
Change 27.358 Support for OMEGAMON ONDV SMF 112 SUB-SUBTYPE '0100'X,

EX112USD optional USREVNT1 or User Function clock/count section.

EX112UST dddddd Dataset Description

IMAC112 112USD T112USRD USREVNT1 Detail

VMAC112 112UST T112USRT USREVNT1 Totals

VMXGINIT In Omegamon/CICS modules KC2GLB or KC2GLBOL, in the

Jan 19, 2010 RKANPARU or RKANPARM library, you can define up to 10

Jan 29, 2010 "User Functions" that can populate the 10 clock/counter

pairs in this new segment. You define the Function Names

you plan to use, and CLOCK START MISC and CLOCK STOP MISC

will accumulate the MISC duration and count of starts in

the first pair. Names MISC, Sybase, Tablebase are used

in this first example, so those names are used to label

the first three sets of detail/total count/clocks, and

only the first three sets are kept by default. If you

create more User Function data, you can use the _K112UST

and K112USD macros to keep more than three, and you can

use the EX112UST exit to change the existing three or to

add new labels for the other counters.

Thanks to Henry Steinhauer, Northwestern Mutual, USA.

Change 27.357 -Some (new) TRANSLATE() functions had '00' or '80' where

VMAC80A '00'x or '80'x should have been specified.

Jan 17, 2010 -TYPE8066 dataset is enhanced with variables from RACFTYPE

6, 318, 319, and 320.

Thanks to Matthew T. Chappell, Queensland Dept. Transport, AUSTRALIA.
Change 27.356 -The standard SAS JCL procedure can now be used for MXG on

VMXGCNFG z/OS. You do not need a separate MXGSASVn JCL procedure;

MXGNAMES instead, use this JCL example (in member JCLMXG), after

JCLMXG you EDIT the DSNAMES of your MXG Source, MXG "USERID" and

CONFIMXG MXG Formats datasets into your MXGNAMES member in your

Jan 17, 2010 MXG "USERID" tailoring library:

// EXEC SAS,CONFIG='MXG.SOURCLIB(CONFIMXG)'

//MXGNAMES DD DSN=MXG.USERID.SOURCLIB(MXGNAMES),DISP=SHR

or you can provide the names in the jobstream, with:
// EXEC SAS,CONFIG='MXG.SOURCLIB(CONFIMXG)'

//MXGNAMES DD *

%LET MXGUSER1=HLQ.MXG.USERID;

%LET MXGSOURC=HLQ.MXG.SOURCLIB;

%LET MXGFORMT=HLQ.MXG.FORMATS;

-In addition, the VMXGCNFG macro that was designed by Rich

allocates the //SOURCLIB with OPEN_ED-1047 encoding; by

doing so, the setting for NLSCOMPATMODE is moot, and by

doing this, all NLS sites running with a locale that is

non-ENGLISH_UNITEDSTATES will never need to worry about

NLSCOMPATMODE, so MXG never has to worry about those SAS

language encoding issues again.

There can NOT be a LIBRARY DD in JCL with CONFIMXG, but

you can have a USER FORMAT library. The CONFIMXG member

%INCLUDEs MXGNAMES and then %INCLUDEs VMXGCNFG from the

&MXGSOURC path and then runs the %VMXGCNFG %macro.

The MXGNAMES member defines MXGFORMT and MXGFORMU and

VMXGCNFG LIBNAME-allocates MXGFORMT to LIBRARY LIBREF and

LIBNAME-allocates MXGFORMU to the USRFORMT LIBREF/DDNAME.

If both MXGFORMT and MXGFORMU are specified in the

MXGNAMES then the SAS system option FMTSEARCH is set:

OPTIONS FMTSEARCH=(USRFORMT LIBRARY)

so the user's format library is searched first.

Member JCLINSTL has the example JCL for ALOCUSID and FORMATS.

Both of those examples use MXGSAS94 JCL Procedure to create FORMATS,

but sites with National Language Support, should consider CONFIMXG

which protects for a future SAS version in which the NLSCOMPATMODE

option will be removed. CONFIMXG uses your standard site's SAS JCL

procedure and options:

// EXEC SAS94,CONFIG='MXG.SOURCLIB(CONFIMXG)'

//MXGNAMES DD DSN=MXG.USERID.SOURCLIB(MXGNAMES),DISP=SHR

But to create or update your FORMATS library with CONFIMXG:

a. The DSNAME of that format library must be named in MXGNAMES

b. There must be no //LIBRARY DD in the JCL of this job step.

c. You must use this syntax in the SYSIN for a new FORMATS:

// EXEC SAS94,CONFIG='MXG.SOURCLIB(CONFIMXG)'

//MXGNAMES DD DSN=MXG.USERID.SOURCLIB(MXGNAMES),DISP=SHR

//SYSIN DD *

LIBNAME LIBRARY CLEAR;

LIBNAME LIBRARY 'MXGV3603.FORMATS'

DISP=(NEW,CATLG)

SPACE=(CYL(10,3))

UNIT=SYSDA;

or use DISP=OLD to replace your existing formats library.

Thanks to Rich Anderson, SAS Institute Technical Support, USA.

Thanks to Chuck Hopf, Independent Consultant, USA.

Change 27.355 For execution under Windows, the unconditional create of

VMXGALOC existing directories or the delete of non-existent ones

Jan 17, 2010 caused popups that delay jobs until the popup is cleared.

An OPTIONS NOXWAIT was relocated around the TREND dataset

allocations, which will eliminate the popup messages, but

the creations/deletions are also now done in conditional

tests, now that we know these Windows commands exist:

This command tests if a folder exists before trying to

delete its files

If exist "m:\asdf\fdsa" del /q "m:\asdf\fdsa\*.*"

and this command makes sure a folder doesn't exist

before trying to create it:

If not exist "m:\asdf\asdf" md "m:\asdf\asdf"

Thanks to Jim Quigley, ConEd, USA.

Change 27.354 An MXGNOTE now prints the value of _IMSVERS on the log of

IMACIMSA the TYPEIMSA and TYPEIMSB steps of the JCLIMSL6 job. If

JCLIMSL6 IMS V9 records are read with _IMSVERS set to 10, an 08x

TYPEIMSA record is dumped with INVALID YYYY error; in V9, the YYYY

TYPEIMSB is located byte 81, but with _IMSVERS of 10, MXG tries to

Jan 15, 2010 to read the YYYY starting in byte 101. Before MXG 27.01

and Change 27.033, _IMSVERS was set either in IMACIMSA or

in your IMACKEEP, but now its value is set with statement

%LET MACKEEP= MACRO _IMSVERS 10.0 % ;

in the //SYSIN test in JCLIMSL6 (twice), so that you do

NOT have to EDIT IMACIMSA/IMACKEEP to define _IMSVERS.

This change just adds that diagnostic MXGNOTE so you can

see the actual value, if you should also see a hex dump!

Comments were revised to document this change.

Thanks to Douglas G. Wells, First National Bank of Omaha, USA.
Change 27.353 While I expected RMFINTRV would be used to create a small

VMXGRMFI number of "WORKnn" workloads, like 20 or so, for ease in

Jan 15, 2010 consolidation of scores of Service/Reporting Class into

logical workloads, RMFINTRV can now be created with up to

999 sets of "WORKnnn" variables, and also supports up to

9999 Service and Reporting Classes per Workload.

Thanks to Wayne Bell, UniGroup, Inc, USA.
Change 27.352 Variable SMF42JOQ was incorrectly "spelled" with a zero

VMAC42 instead of an "oh".

Jan 12, 2010

Thanks to Ambat Ravi Nair, CitiGroup, SINGAPORE.

Change 27.351 -READDB2 didn't invoke the EXdddddd member when selection

READDB2 generated an _Edddddd macro with only OUTPUT _Wdddddd,

Jan 12, 2010 causing any new variables you created in your tailored

EXdddddd to be not created. READDB2 now always %INCLUDEs

the EXdddddd member in its generated MACRO _Edddddd text.

EXCEPT: The EXPDBACB exit is NEVER called when READDB2

was asked to create DB2ACCTB, because the MXG EXDB2ACB

does NOT output DB2ACCTB (because of potential size), and

if READDB2 calls your tailored EXDB2ACB that did output,

then we would output DB2ACCTB twice.

-Logic for selection with DB2xxxx=AAAA/BBBB/CCCC/DDDDD

was corrected and simplified so it always works as doc'd.

-PROC COPY when PDBOUT= is specified only copies DB2 data

(and not any other datasets that happened to be in WORK).

Thanks to Raff Rushton, IBM Global Services, USA.
Change 27.350 This 1994 example analysis of "bands" of Usage is updated

ANALUSAG to use TYPE72GO instead of TYPE72 and an example JCL was

Jan 12, 2010 added in the comments.

Thanks to R. Wells, American General Finance, USA.

Change 27.349 Support for RMM APAR OA28930 which relocates the fields

VMACEDGR RDBLKCNT and RDTOTAL and expands them to 20 (EBDCIC) NUM

Jan 12, 2010 charaters in the DEXT and XEXT records.

-Support for RMM APAR OA24025 which adds fields RDBESKEY

(DEXT) and XDBESKEY (XEXT) with the CA Tape Encryption

Key value.

Thanks to John Grasing, MetLife, USA.
Change 27.348 The Optional SYSLOG Message Capture in MXGTMNT Tape Mount

ASMTAPEE Monitor program expected a maximum of 255 lines in a

Jan 12, 2010 multi-line Console Message (see Change 27.336 text), but

HASP636 message caused an SVC DUMP "MXG Monitor Extension

Subtask Abend" with its 1541 lines! This update, ML-46,

avoids the SVC Dump by skipping messages with over 255

lines, but is only a circumvention; once we can create a

similar large message on our test system, so we can see

the control block structure, we will create an ML-47

update that will support any multi-line console message.

Thanks to Beau Chavis, Bank of America, USA.

Thanks to Skip Abadie, Bank of America, USA.

Change 27.347 TMON/VTAM "SI" record Interval variables were not all

VMACTMVT input; I apparently had an incorrect DSECT or misread it.

Jan 8, 2010

Thanks to Paul Volpi, UHC, USA.

Change 27.346 Analysis of Hourly CPU Times in MXG TYPE70, TYPE72GO, and

ANAL307X SMFINTRV datasets (or in ITRM XTY70, XTY72GO, & XSMFINT)

Jan 14, 2009 to compare times captured in LPAR Dispatch and Effective,

Feb 22, 2010 captured in Service Classes, and captured in SMF Interval

Address Space records. Two reports, one BY SYSTEM and

one with detail BY SYSTEM SRVCLASS are produced.

The identification of the "Hour" of RMF and SMF Interval

observations is not straightforward; you must start with

the "projected interval end time" SYNCTIME/SMF70GIE and

then subtract from it to get the STARTHOUR, and the value

to subtract is different if you have SYNC(59) specified

in SMFPRMxx instead of the recommended SYNC(0) option.

Note: MXG has ALWAYS recommended SYNC(0) so that RMF

and SMF interval records are written at 00/15/30/45 to

create clean, comparable intervals. But if you still

have MICS, then you are unfortunately stuck with using

SYNC(59) to write at 14/29/44/59 minutes, because MICS

still uses SMFTIME to define its intervals, a (poor)

design that requires records be written early.

Feb 22, 2010: The output datasets created by ANAL307X

now include TYPE30_6 (deaccumulated) interval data and

SMFINTRV, so the subtypes 2/3/6 data is included in

this comparison of 30, 70, and 72 interval CPU times.

Thanks to Dick Cook, North Carolina Dept of Info Technology, USA.

Thanks to Francisco Ojeda, SAS Institute, USA.

Thanks to Joe Piechota, SAS Institute, USA.

Change 27.345 This JCL example uses UTILBLDP to process CICS and DB2

JCLCIDB2 SMF records to create PDB.ASUMUOW and PDB.CICS for

Jan 7, 2010 complete analysis. The output datasets are COMPRESSED

but COMPRESS=NO is used for the intermediate datasets,

to save CPU time (at the cost of doubling WORK space).
Change 27.344 Variables LPCTBY and PCTLPBY were missing in dataset

VMXG70PR PDB.ASUMCELP for the LPARNAME='PHYSICAL'.

Jan 7, 2010

Thanks to Karl Lasecki, Chemical Abstracts, USA.

Change 27.343 %VMXGSET is enhanced to permit multiple datasets to be

VMXGSET read from each data library, and new option APPEND=YES

Jan 6, 2010 removes the semi-colon from the constructed SET statement

so that you can add other dataset(s). This syntax:

%VGETDDS(DDNAMES=PDB);

DATA COMBINE.JOBS;

%VMXGSET(DATASET=JOBS,APPEND=YES) DAY.SPUNJOBS;

will read all of the JOBS datasets in the PDBx DDNAMEs

and the DAY.SPUNJOBS dataset.

Thanks to George Pandzik, USAA, USA.

Change 27.342 XAMSYS records from XAM Release 3.4 had SEGLEN=168, but

VMACXAM MXG code for Release 3.7 expected SEGLEN=220, causing

Jan 5, 2010 an MXG ERROR message for each record. Now, the XAMSYS

record's length is tested and only the old variables are

input for SEGLEN=168.

Thanks to Rodger Foreman, Acxiom, USA.

Thanks to Jerry Urbaniak, Acxiom, USA.
Change 27.341 Change 27.257 added a warning and correction if SUBTYPE

VMACSMF in an SMF record exceeded 255 (because the SMF record

Jan 5, 2010 creator incorrectly stored their subtype in the left byte

instead of the correct location in the right byte).

The correction is ONLY needed when just the SMF header

logic is used (only in MACRO _SMF, or UTILGETM), so

that the real SUBTYPE value is available for reporting

or selection.

Now, the WARNING is NOT printed by default; you can print

the warning with %LET MXGDEBUG=VMACSMF;

Thanks to Jerry Urbaniak, Acxiom, USA.
Change 27.340 SRM Thales Security PTF SL24010 INCOMPATIBLY changed the

VMACSRMH Summary Record, which caused zero observations to be

Jan 5, 2010 created in datasets SRMHSMAP and SRMHSMDE. New variables

S04DBUSY S04DOVER S04DINTV S04DNCNT S04DUFLG

are now added to SRMHSMDE and the PTF is now supported.

Thanks to Kim Nguyen, National Australia Bank, AUSTRALIA.

Thanks to Shu chun Lai, National Australia Bank, AUSTRALIA

Thanks to Anne Chung, National Australia Bank, AUSTRALIA.

Change 27.339 Change 27.046 (MXG 27.02) changed four error variables

VMACEDGR RVTRERR RVTWERR RVPRERR RVPWERR

Jan 4, 2010 from character to numeric when IBM changed their lengths

from four to five bytes, but that was not documented.

If datasets built with MXG 27.01 or earlier are combined

with datasets built with MXG 27.02 or later, then this

ERROR: VARIABLE RVPERR DEFINED AS BOTH CHAR AND NUMERIC

will result. The only solution is to convert those old

character variables to numeric variables prior to merge

with the new datasets. This can be done with

DATA PDB.EDGRVEXT;

SET PDB.EDGRVEXT

(RENAME=(RVTRERR=XRVTRERR

RVTWERR=XRVTWERR

RVPRERR=XRVPRERR

RVPWERR=XRVPWERR));

RVTRERR=INPUT(XRVTRERR,5.);

RVTWERR=INPUT(XRVTWERR,5.);

RVPRERR=INPUT(XRVPRERR,5.);

RVPWERR=INPUT(XRVPWERR,5.);

DROP XRVTRERR XRVTWERR XRVPRERR XRVPWERR;

LENGTH DEFAULT=4;

DATA PDB.EDGRXEXT;

SET PDB.EDGRXEXT

(RENAME=(RVTRERR=XRVTRERR

RVTWERR=XRVTWERR

RVPRERR=XRVPRERR

RVPWERR=XRVPWERR));

RVTRERR=INPUT(XRVTRERR,5.);

RVTWERR=INPUT(XRVTWERR,5.);

RVPRERR=INPUT(XRVPRERR,5.);

RVPWERR=INPUT(XRVPWERR,5.);

DROP XRVTRERR XRVTWERR XRVPRERR XRVPWERR;

LENGTH DEFAULT=4;

Thanks to Steve Sombke, American Century, USA.
Change 27.338 -The new DATEJUL= argument to generate dataset\directory

VGETDDS names that contain the Julian Date didn't stop at 2009365

Jan 3, 2010 but tried to create 2009366, 2009367, etc. The arguments

are now validated and if both start and end are Julian,

then the names are built with Julian dates. Otherwise,

if the arguments are numeric, then the names contain just

numeric sequences. When Julian dates are detected, then

SAS Date functions are used, so the names will correctly

roll forward or backward and both year-end AND leap-year

dates are correctly generated as the names.

-The length of the Julian Date argument can be either five

(09360) or seven (2009360) digits, but the directory or

z/OS dataset name must have the same number of digits.

So you can use

%VGETDDS(DATEJUL=h:\mxg\d,start=09360,end=10010);

for directory names h:\mxg\d09365 thru h:\mxg\d10010

or you can use

%VGETDDS(DATEJUL=h:\mxg\d,start=2009360,end=2010010);

for directory names h:\mxg\d2009365 thru h:\mxg\d2010010

Thanks to George Pandzik, USAA, USA.

====== Changes thru 27.337 were in MXG 27.11 dated Dec 31, 2009========
Change 27.337 The sort order for TYPE70PR was wrong in 27.10 MONTHBLD;

MONTHBLD a second MACRO _BYLIST and _MNTHBLD should have been

Dec 30, 2009 removed. The correct sort order is reduced to now be

SYSPLEX SYSTEM SYSNAME STARTIME

to avoid NOT SORTED conditions.

Thanks to Winnie Pang, Hawaii Medical Services Association, USA.

Change 27.336 Variable SYSLTEXT was arbitrarily INPUT as length $1024,

VMACTMNT sufficient for all tape-mount-related SYSLOG text,

Dec 24, 2009 but other SYSLOG records that can be written by MXGTMNT

can be 32384 bytes long; that is the maximum number of

lines in a multi-line WTO SYSLOG message (255) times the

maximum length of each message (126) plus one byte we add

between each message (254). MXGTMNT processes the

segmented records, which is the way the system presents

multi-line WTOs via the console interface. MXGTMNT

concatenates the messages into one block in one subtype 9

as opposed to creating multiple subtype 9s. So MXGTMNT

is covered unless IBM increases the number of lines or

message length which is not very likely but easily

addressed if they do. SYSLTEXT is now 32384.

Thanks to Beau Chavis, Bank of America, USA.

Thanks to Skip Abadie, Bank of America, USA.

Change 27.335 A stray */ at the end of ANALFIOE prevented any other

ANALFIOE subsequent %INCLUDEs to be bypassed. Unmatched comment

Dec 22, 2009 pairs often cause strange/unpredictable results.

Thanks to Brian Harvey, HCL America, USA.

Change 27.334 MXG 27.10 only. Macro variable &EPDBINC was accidentally

EXPDBINC left in the EXPDBINC member, but it is already correctly

Dec 21, 2009 located in BUILPDx, causing a second unwanted invocation

if %LET EPDBINC= was used to tailor BUILDPDB. However,

it should be noted that there is an inconsistency between

the syntax of the EPDBINC macro variable and the EXPDBINC

member, to include VMACxxxx's for BUILDPDB tailoring:

%LET EPDBINC= MEMBER1 MEMBER2 ;

EXPDBINC code: %INCLUDE SOURCLIB(MEMBER1 MEMBER2);

The value of %LET EPDBINC= is one or more MEMBER names to