• Subject: RE: V5R1 Library List Enhancement (was: PRTCMDUSG RTVJOBA)
  • From: Jim Damato <jdamato@xxxxxxxxxxxxxxxxx>
  • Date: Tue, 10 Apr 2001 14:30:15 -0500

I still like the idea of an AS/400 mode.  Keep all V4 commands intact in
AS/400 mode just like System 38 mode.  Make all OS changes to the "native
iSeries environment."  IBM could initially make vague threats that AS/400
mode is going away some day, eventually leading up to the iSeries V5R4
announcement that V5R5 will be the last AS/400 mode release and that V6 will
not support old AS/400 commands.  That way folks would have to start taking
responsibility for their dinosaur applications instead of clinging to the
mentality that if an application runs today IBM will always support it.



-----Original Message-----
From: barsa@barsaconsulting.com [mailto:barsa@barsaconsulting.com]
Sent: Tuesday, April 10, 2001 10:39 AM
To: MIDRANGE-L@midrange.com
Subject: RE: V5R1 Library List Enhancement (was: PRTCMDUSG RTVJOBA)



I agree that I don't want to start a war, but this is the first significant
change to CL that makes you recompile ever!  Given the fact that IBM has
allowed us to avoid this for over 20 years, they could have done it now.

There are several possible techniques:

   Support a second portion of the user library list.  Possibly with access
   to both parts together in certain commands.
   Add job level support for short vs. long library lists.  This would
   require change to a *JOBD, and likely an added system value.
   Not support the requirement.  IBM made a statement many years ago that
   there would "never be support for over 25 libraries in the user portion
   of the library list".  They could have kept to that promise.
   Teach people how to code properly.

The only untenable solution that I could imagine was to break user's code.

Al



Al Barsa, Jr.
Barsa Consulting Group, LLC

400>390

914-251-1234
914-251-9406 fax

http://www.barsaconsulting.com
http://www.taatool.com





 

                    Jim Damato

                    <jdamato@dollargene        To:
"'MIDRANGE-L@midrange.com'" <MIDRANGE-L@midrange.com>             
                    ral.com>                   cc:

                    Sent by:                   Subject:     RE: V5R1 Library
List Enhancement (was: PRTCMDUSG RTVJOBA)   
                    owner-midrange-l@mi

                    drange.com

 

 

                    04/10/01 10:34 AM

                    Please respond to

                    MIDRANGE-L

 

 





There are a number of choices.

-You stay on V4R5 until you can switch to a supportable product.

-You hire a consultant who has SEU and a CL compiler to produce the hack to
RTVJOBA I've described.

-You stay on V4R5 forever.  As long as your software is unsupportable you
may as well bring your OS to the same state.

-You recognize that you've been living on borrowed time, and that you won't
be supported forever.

I don't want to start another war here, but I really don't believe that
systems vendors are obligated to provide full backwards-compatibility
forever.  This is a relatively minor change compared to what we've seen in
the Oracle or Windows world.  The poor fools who are in the mess you
describe can run V4R5 fully supported for quite a while, during which time
they can plan and budget for patches, upgrades, or replacements.

The moment your software vendor goes out of business you yourself are
responsible for contingency planning.  Who would you blame if the hardware
vendor went out of business?  And falling behind in software support is a
business decision with associated risks.

We moan about IBM's failure to market properly, but we expect them to
produce a competitive, cutting edge machine that can also continue to run
every third-party nightmare ever produced on this platform and the System
38.

In all seriousness, with your presence in the market why don't you lobby to
IBM for an "AS/400 mode"?  It might help the brand if they were able to
draw
that line in the sand between the iSeries and its legacies.

James Damato
Manager - Technical Administration
Dollar General Corporation
<mailto:jdamato@dollargeneral.com>



-----Original Message-----
From: barsa@barsaconsulting.com [mailto:barsa@barsaconsulting.com]
Sent: Tuesday, April 10, 2001 8:26 AM
To: MIDRANGE-L@midrange.com
Subject: RE: V5R1 Library List Enhancement (was: PRTCMDUSG RTVJOBA)



And what do you do for programs for which you have no source and where the
CL is not retrievable?  What do you do if you don't have SEU?  What do you
do if you are just a dumb user, and have an end-user system with an
unsupported package (that's worked since the year of the flood), with:

a.)  a vendor out of business
b.)  support so far behind that you cannot afford to get current

Al

Al Barsa, Jr.
Barsa Consulting Group, LLC

400>390

914-251-1234
914-251-9406 fax

http://www.barsaconsulting.com
http://www.taatool.com







                    Jim Damato

                    <jdamato@dollargene        To:
"'MIDRANGE-L@midrange.com'" <MIDRANGE-L@midrange.com>
                    ral.com>                   cc:

                    Sent by:                   Subject:     RE: V5R1
Library
List Enhancement (was: PRTCMDUSG RTVJOBA)
                    owner-midrange-l@mi

                    drange.com





                    04/09/01 08:15 PM

                    Please respond to

                    MIDRANGE-L









Gee, if we all had externalized our library list retrieval along with our
DB/IO programs we wouldn't have this problem, would we?

In all seriousness, couldn't you just write a new command (call it
OLDRTVJOBA) and CL program to interpret the new library list results of
RTVJOBA?  The new command would run the V5R1 RTVJOBA and return the 275
character library string we know and love.  Scan source or run Pathfinder
command usage to find as many occurrences as you can.  The code
modification
would just require that you replace RTVJOBA with OLDRTVJOBA.  If you missed
any programs you could easily change them as they blew up.  The command
processing program for OLDRTVJOBA could also detect when you've crossed the
25 library threshold.  Eventually someone's going to take advantage of the
new feature and throw you out of compliance, so you might as well report it
on QSYSOPR.

Those poor folks who don't have all their source (and those folks who don't
want to change their code) could make OLDRTVJOBA into the new RTVJOBA and
put it in a system library at a higher level than QSYS, and have the
command
processing program execute QSYS/RTVJOBA.

As a veteran of software package hell I look forward to an increase in the
number of libraries in the user library list.  I always liked letting the
OS
environment do the work for me instead of configuring software environments
via CL for all my packages.  Unfortunately it was impossible to provide
library entries for coexisting merchandising, financial, EDI, spool
management, etc. package library lists.

I agree that V5R1 has ripped the rug out from under us, but I also think
that providing a system value or old and new library list strings in
RTVJOBA
is a bit of a hack.  I never got really liked the legacy fields in the
output files for DSPOBJD and DSPFD either.

We're expecting the system to grow and respond the changing face of
technology, remaining competitive in a complex market, but we also want it
to painlessly run our 1989 legacy applications, gracelessly migrated off
the
System 38.  Maybe IBM should set up an "AS/400 mode" for the iSeries.

James Damato
Manager - Technical Administration
Dollar General Corporation
<mailto:jdamato@dollargeneral.com>


-----Original Message-----
From: barsa@barsaconsulting.com [mailto:barsa@barsaconsulting.com]
Sent: Friday, April 06, 2001 11:07 AM
To: MIDRANGE-L@midrange.com
Subject: V5R1 Library List Enhancement (was: PRTCMDUSG RTVJOBA)



Hi,

Just for the record, the correct spelling of the word "enhancement" is
"f-i-a-s-c-o".

I intend to be very vocal about the V5R1 increase in number of libraries in
the user portion of the library list.  I have refrained from comment in
this forum until I received a clearance from Rochester, which I received on
Tuesday.  I had received clearance to speak about this informally at the
Fall COMMON conference in Baltimore.  Prior to that conference, I also
conducted about 10 to 12 interviews on the topic, and reported those
results to IBM.  IBM paid about as much attention to my findings as the
Morton Thiokol engineers did to the space shuttle o-rings.  As far as I can
determine, they never contacted any of the interviewees (whose names and
identification I provided) on a timely basis when the resolution of this
problem was being determined.

The problem is that IBM increased the number of libraries in the user
portion of the library list from 25 to 250 in the unannounced release of
OS/400.  This will cause any properly coded RTVJOBA command (As well as
some APIs) that specified the USRLIBL to fail if more than 25 libraries are
found on the list  (My definition of properly coded is that the return
value has to be 275 bytes.)

IBM has provided a poorly designed band-aid for V5R1 via PTF.  I was the
person that requested IBM to code the fix, and they coded it improperly.
(They cannot complain that they didn't know how to write the fix,as I gave
them the pseudo code.  Depending on how inadequate the PTF proves, I may
clean it up and publish the pseudo code here, but I'm too busy at the
moment.)  When you exceed 25 libraries on the list, the IBM PTF provides a
different escape message, so you code abends with a different error
message.  This is about as exciting as kissing your sister.  The fix
provides a system wide patch (no, not scoped over the job, which is what is
needed) that will only be supported for a few releases.  (In fairness to
IBM, a system wide patch was about what we could have expected from them at
the time I discovered the problem anything else would have been too
expensive to code, based on how complete that release of OS/400 has
progressed.  This negates the fact that the both the functional addition
and the patch were not well thought out.)

They're fairly mad at me for complaining about this, but what else is new?
The last time I complained about anything as severely as I plan to complain
about this, it was when I said that "V3R1 sucked", and of course, IBM said
that V3R1 was stable and told me I was wrong.

The long term solution is that you must find every RTVJOBA command that
uses the USRLIBL parameter, and replace the returned variable from 275 to
2750 bytes.  Depending on what you do with that data*, this could cause
other parts of that program to fail.

Assuming that you have all of your source, this is not an impossible task.
   You could scan for every RTVJOBA using PDM, and then examine every
   command by hand for USRLIBL.
   There is a new TAA Tool called Scan Command Keyword (SCNCMDKWD).  You
   can specify the command name and the keyword name.

In both cases it is your responsibility to make sure that the program will
still run.  You also (reasonably speaking) need a license to SEU.

   *  Exactly what you do with the returned data will determine the
   complexity of the fix.  As far as I can determine, most people stuff it
   into a few different variables, so the fix is easy.  If you stuff it
   into a data area, this is tougher, because the maximum length of a data
   area is 2000 bytes.  I know of one vendor that puts every library name
   into a different field in a database file - oh god forbid!

If you don't have your source, you have been @#$%ed by IBM.  Why is this
significant?  This is the first time that IBM has done this to you (making
a change to the architecture of the system and requiring you to go back to
source) ever in the system.  There are some notable exceptions:
   In Release 3.0 of CPF, IBM required you to recompile every CL program.
   However in that release they added the new RTVCLSRC command, and of
   course, prior to that, there was no notion of ALWRTVSRC(*NO).
   The first release of the System/38 Migration Aid (5714MG1) had no notion
   that observability could have been removed.  When IBM discovered that
   some vendors were removing the program template, they added a diagnostic
   aid to this product.
   The RMVOBS parameter of CHGPGM was added in V1R2M0 of OS/400, but (IMHO)
   IBM provided adequate warning of the drawbacks.

What IBM should have done (and should still do in a future release of the
system) is add this feature as a system value, allow the system value to
default into an attribute of a job description, and at job initiation time,
the value should be propagated to the job.  The value needs to be
consistently added to both the native AS/400 functionality, and the
System/38 compatibility command set.  (This sounds like a lot of work, but
it's really trivial.  In fact the current PTF is inconsistently applied
over the native commands and the System/38 commands.)  It also needs to be
extended to save/restore.  I have privately submitted my proposed changes
to IBM in detail, and they have yet to respond with any intentions to do
anything other than file them in the circular file.

In my opinion, the change was not well thought out by IBM.  (My upcoming
magazine article on this subject might be less polite in terminology.)

Al

Al Barsa, Jr.
Barsa Consulting Group, LLC

400>390

914-251-1234
914-251-9406 fax

http://www.barsaconsulting.com
http://www.taatool.com








                    MacWheel99@aol.com

                    Sent by:                   To:
MIDRANGE-L@midrange.com (AS400 & family discussion group),
                    owner-midrange-l@mi        BPCS-L@midrange.com (BPCS
Users Discussion Group)
                    drange.com                 cc:

                                               Subject:     PRTCMDUSG
RTVJOBA


                    03/31/01 12:00 PM

                    Please respond to

                    MIDRANGE-L









From
MacWheel99@aol.com (Alister Wm Macintyre) (Al Mac)

Below is cut & paste of information from an AS400network newsletter that I
want to talk about ... I left their advertiser URLs included in hopes they
will not get annoyed with me for forwarding their copyrighted stuff.  Item
#
4 affects BPCS & probably every other ERP & other software package that
anyone is using on the 400.

PRTCMDUSG gets at a list of all programs that use a particular command.
I learned about this command in prior MIDRANGE-L discussion about
cross-referencing where various software objects are used.

We have 800 CL programs in our BPCS 405 CD live environment library list
that
use RTVJOBA.  Some of them are our modifications, but most are vanilla base

BPCS.
RTVJOBA is the first of the 5 commands listed in the newsletter that will
return more information starting OS/400 V5.  This is not the whole picture,

but I have to start somewhere.

RTVJOBA can be used to retrieve a lot of stuff about a job ... library list

information is not its only usage, so in many cases the fact that OS400 V5
is
going from 25 to 250 user libraries in the list won't make a bit of
difference, but when the retrieval is to access the library list
information,
the fact that more data is coming back could have a detrimental effect
depending on how the software is written, and depending on if & when we
utilize the extra libraries.

The situation for BPCS V6 users is different than for BPCS 405 CD because
V6
users do not have access to the source code, rather all code is via SSA's
"case" AS/Set.  SSA had announced that they dropping support for 405 CD
effective end of May 2001.  This IBM V5 is due out beginning of May 2001.
Now there are SEVERAL places that offer good quality tech support for BPCS
405 CD when SSA's ends, so that is not a problem.  My thoughts are

a) Does SSA know about this? (I sent a general inquiry to SSA tech support
to
ask)
b) Can we expect a final REL 03 aggregate collection of BMRs at the end of
405 CD that includes a fix for this nuance?

Assuming that we can not depend on such an expectation, there is a joint
challenge of identifying inventorying what all our retrieve library list
software is doing to figure out the impact & what needs fixing.

I think there is a PDM search & substitute command string that I need to
learn, except I like to look at what exactly is happening in each instance.

However PDM search might help in mapping out how RTVJOBA is used in our 800

programs.

Subj:    Club Tech iSeries Programming Tips - 03.29.01
From:   ClubTechiSeriesPrgrmTips@list.as400network.com
(ClubTechiSeriesPrgrmTips)

*********** Club Tech iSeries Programming Tips Newsletter ***********
An AS400 Network Publication              http://www.as400network.com
Home of NEWS/400 Magazine
Issue 41                                               March 29, 2001

Sponsored by Generic Software, Inc., at (800) 698-5669 or visit
http://www.genericsoftware.com/html/save_output_queue.htm .

<snip>

THIS WEEK:
> APIs by Example: Read/Write an IFS File Line in RPG IV
> APIs by Example: Read an IFS File Line in Cobol
> Data Area Editor Utility
> Poor Man's Cross-Reference
> Maximum Libraries in *LIBL to Change from 25 to 250
< big snip >

* Make your RPG Programs happy! Download RPG-Alive...
  http://www.RPGAlive.com

<snip>

4. MAXIMUM LIBRARIES IN *LIBL TO CHANGE FROM 25 TO 250
The V5 release of OS/400, due out in May, changes the maximum number
of libraries in the user part of a library list (*LIBL) from 25 to
250. This will alleviate some problems that arose from the previous
limitations, but it may cause other problems with your existing code.

The new versions of the RTVJOBA (Retrieve Job Attribute) command and
the QUSRJOBI (Retrieve Job Information), QWCRTVCA (Retrieve Current
Attributes), QUSRSPLA (Retrieve Spool File Attributes), and QWDRJOBD
(Retrieve Job Description) APIs can now return more data than in
previous releases. Be sure that any applications you have that use one
of these interfaces provides enough room for 250 libraries in the
return value.

When you increase the size of a return variable, you can still safely
call V4R5 and earlier releases of these interfaces because there is no
harm in providing more space than needed. Just be sure that your own
application logic correctly handles however many library entries are
returned.

For more information, see:
http://www.ibm.com/eserver/iseries/developer/os400/lib_list.html

Thanks to Paul Conte for the above item

<snip>

http://as400network.com/str/books/uniquebook2.cfm?NextBook=181 .

This newsletter is edited by Chuck Lundgren,
mailto:clundgren@as400network.com .

FOR NEW SUBSCRIPTIONS, you can subscribe by joining the AS400 Network
with a handy Web form at http://www.as400network.com/join/ .

IF YOU WANT TO SPONSOR a Club Tech iSeries Programming Tips
Newsletter, please contact your AS400 Network sales manager. Click
here for details:
http://www.as400network.com/info/mediakit/Sales/Index.htm .
___________________________
Copyright 2001, NEWS/400
http://www.as400network.com
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---






+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---






+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to
MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---






+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator:
david@midrange.com
+---
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---

This thread ...

Follow-Ups:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2020 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].