Re: Permission Problems On 3rd Open Of Same PDS

[Kenneth E Tomiak]

On Fri, 31 Aug 2007 21:18:58 -0700, Harry Goldschmitt <[EMAIL PROTECTED]> 
wrote:

>I've got an LE C++ program that fopens 3 different members of the
>same PDS.  The first 2 fopens and fcloses work fine, but the 3rd
>fopen gets a permission error for some users.  I've traced the
>problem a to missing OMVS segments for the failing users.  I've
>resolved this in the past by just defining an OMVS segment in RACF
>for the users that wish to run the program.  Now, however, I'm trying
>to run it at a site with ACF2.  All users have a default OMVS segment
>defined but I'm still getting the problem.
>
>My application gets an "EDC5111I Permission denied.
>(errno2=0x5b450002) DD:TEST3" message.
>

I do not follow your logic. If you had permission to use the dataset 
USERID.TEST for the first two fopens then you have it for the third and your 
real problem should be something else. Is it really USERID.TEST or 
&SYSUID..TEST?

When you looked up errno2, what did it tell you?

http://publibz.boulder.ibm.com/cgi-
bin/bookmgr_OS390/BOOKS/BPXZA870/CCONTENTS

The current (requesting) process
does not have the requested access  
authority to the file, or   
directory.  

Do they have ACF2 access to USERID.TEST?

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: DFSMSdss SYSPRINT attributes

[Kenneth E Tomiak]

Get a license to use z/OS and report your dissatisfaction to IBM. They really 
do not take complaints off IBM-MAIN.

On Wed, 29 Aug 2007 16:50:08 -0600, Paul Gilmartin 
<[EMAIL PROTECTED]> wrote:

>In:
>
>Title: z/OS V1R7.0 DFSMSdss Storage Administration Reference
>Document Number: SC35-0424-06
>
>Ouch.  By experiment, some of this is false or incomplete, and
>much of what is true is harmful.
>
>Further, why reduce LRECL if it exceeds 137?  (Again, this
>is permanent; in the label.)  There's no constraint against
>writing records shorter than LRECL to a RECFM=V[B]A data
>set.  And why, then, set BLKSIZE to 141? This is an adverse
>blocking factor nowadays.  Use what the programmer specified
>if compatible with LRECL; else let OPEN fail and report it.
>Again,  I am far from a fan of utilities overriding the
>attributes of existing data sets.
>
>This is likely WAD, but it feels as if no one reviewed that
>design.
>
>-- gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Jes2 OFFLOAD leaving out some JOBNAME

[Kenneth E Tomiak]

On Tue, 28 Aug 2007 17:36:41 -0500, Alvaro Quintupray B. 
<[EMAIL PROTECTED]> wrote:

>Hi.
>
>  Can I do OFFLOADn,type=transmit  from spool  and   leaving some
>JOBNAME  out   off  Tape ...?
>

I believe you include what you want. So do not include what you do not want.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Calling a AMODE 31 program from AMODE 64

[Chris Mason]

J.R.

1. "The comma"

And, by the way, I never stopped using the comma between index and base 
registers, so I never had to restart when AR-mode came along. Good 
practice never goes out of style.


By which I expect you mean you never stopped using the comma in order to 
indicate that you wanted the register in the address to be the base register 
rather than the index register.


I brought this up as recently as a fortnight ago[1] as an excuse to recall a 
story about the famous comma and Wayne Driscoll was kind enough to add his 
comment that it indeed mattered still.


2. Register Equates

After using register equates for forty years, instructions just don't look 
right without them.


When I started doing some serious Assembler long, long ago, I decided that 
the R0-R15 register equates were not really going far enough in terms of 
assisting documentation of the program. I decided a character string that 
gave some idea what the register was doing was more appropriate - although I 
always retained the leading "R" - innate conservatism I suppose.


I discovered another benefit in using this approach. I no longer had to 
bother solving my problem - some would say "issue" these days[2] - alongside 
worrying about assigning numbers to registers. The routine got written and 
then I bothered about the EQU statements to assign a number to my register 
symbols.


As a purely pragmatic rule, I decided that if I just couldn't shuffle the 
numbers to match the symbols, the routine must be too large and so needed 
"subroutinising". This actually never happened that I can recall.


This technique did yield one very well remembered situation. I found myself 
with an LR instruction as follows:


LR RREC,RBUF

While coding up the register EQU statements, I found that I had finished 
working with RBUF at precisely the point I started working with RREC - which 
is very sensible when a reasonable guess is made as to what sort of routine 
was being written - *and* RBUF and RREC happened to have been assigned the 
same register number.


The "problem" was resolved by placing an asterisk in column 1!

Chris Mason

[1] "JCL passing parms to ASM module?", 17 Aug.

[2] Being British, I don't always insert "smileys" or "grins" where list 
conventions would normally require them.


- Original Message - 
From: "J R" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 9:34 PM
Subject: Re: Calling a AMODE 31 program from AMODE 64

(Selectively quoted above)

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Chris Mason]

John

Speculating on your "other information", although I don't know DHCP in 
detail I expect it provides the 5 data elements I used to have to tick off 
when setting up a PC with OS/2 or an RS/6000 with AIX. At one time OS/2 got 
a GUI for this but it still made sense to be sure each of the 5 items 
previously fed in with one line command or another had been dealt with.


The 5 were/are

1. a name
2. an IP address
3. a subnet mask
4. a default router address
5. a name server address

This was a long time ago and I wonder if I might not have missed a 6th 
element out, namely the name server suffix. On the other hand, I wonder how 
vital really is the name configured in the machine itself - as opposed to 
the name held in the name server.


Anyhow, items 2, 4 and 5 are vital. Item 3 is a nice-to-have but also not 
really vital - as I once discovered when the location where I worked 
implemented a change without having to update every last workstation PC.


So for the case of a printer which never performs a name lookup, I guess 
only the IP address and the default router IP address are needed. As long as 
the IP logic assumes that both IP addresses are in the address range 
associated with the LAN, packet routing is possible.


Actually, after I'd got this response ready, I checked the Communications 
Server IP Configuration Guide for references to DHCP to see if the 
documentation was there for Lizette. It appears to be and when flitting from 
"hit" to "hit" I saw the 6, in fact, elements in a bulleted list - page 701.


Chris Mason

- Original Message - 
From: "McKown, John" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 3:12 PM
Subject: Re: Print Distribution and IP connected Devices


...

However, it is also possible to use DHCP and assign an specific IP
address based on the MAC address of the NIC. I do this on my home
network. Why do this? So that the IP address stays the same, but the
other information is dynamic and so that the IP configuration is
centralized at the DHCP server and not on each workstation. This allows
for a uniform "workstation image".

--
John McKown

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Chris Mason]

Lizette

I understood your "new equipment" to be another z system, hence the 
reference to VIPAs which are purely a Communications Server IP component 
feature. Obviously I'm quite unfamiliar with whatever sort of application 
corresponds to "CA Spool or Dispatch".


I'm sorry for the confusion - and my bewilderment over everybody else's 
contributions until the penny dropped.


Since this is all about actual IP printers, there is no way round needing to 
know how they can be addressed since, in the client-server model, the 
printer represents the server.


Even with all this talk about DHCP - which I know only because it's a 
facility implemented by the little black box on the floor of my office which 
connects me to the wide world through my telephone wires - you can't escape 
the fundamental need to be able to initiate a message exchange of some sort 
with the printer and, because you initiate the exchange, you need an 
identification for the printer.


You've had some suggestions which appear to be

1. know the IP address

1a the IP address can be written into the printer storage

1b the IP address can be held in a table in the DHCP server and the printer 
is provided with the IP address when the printer acts as the DHCP client on 
start-up[a] and is identified by, say, the MAC address


2. know a name - but the software initiating the exchange needs to be able 
to translate the name to an IP address


2a. have a name server entry - or a local file entry - which associates the 
name to an IP address


2b. have a dynamic name server entry where the DHCP server keeps the entry 
up-to-date after the DHCP server has assigned an IP address to the printer 
dynamically when the printer acted as the DHCP client on start-up[a] and 
again is identified by, say, the MAC address[b]


Credit where credit is due, on review I see that this is much the same 
advice as Maarten offered.


I realise you've been told that none of this is possible but, as others have 
mentioned, you need to put these options to your management and get some 
*right* answers.[c]


In addition to the opinion of a number of us that the information you have 
from the CA representative is just plain wrong, from his experience with the 
product, Ulrich has stated that the information really is just that, plain 
wrong.


Chris Mason

[a] I wondered what other information might be required in order to 
configure the printer. Probably of the 5 data elements I identified in 
another post, only the default router address is really needed. The printer 
is never going to perform name lookups.


[b] I believe it has been stated that a DHCP server can assign an IP address 
based on information provided by the printer on start-up. However I am not 
sure whether or not a DHCP server can assign a name to be fed into a name 
server based on information provided by the printer on start-up. On review, 
I see Len sort-of suggests this is possible.


Also on review, I saw that Rex uses this option but the key is actually the 
name - also used, suitably suffixed by the DHCP server probably, by the name 
server - rather than the MAC address.


[c] I have in mind, of course, the sort of conversation which ends up with 
the response "Wrong answer" just before a violent conclusion. 


- Original Message - 
From: "Lizette Koehler" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 1:28 PM
Subject: Re: Print Distribution and IP connected Devices



Since I am not involved (unfortunately) with either the IP or Network part
of this move, I am not as familiar with this problem as I would like.

However, I believe that Hal has identified my issue.  They use DHCP to
assign IP addresses.

So what would be a better way.

I am not familiar with why you would use DHCP vs. some other method.

Is there a benefit or is there a better way.

Thanks all for the help.  It is starting to look murky instead of just 
dark.


Lizette


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Chris Mason]

J.R.

The URL not the best analogy. Within an URL you can use an IP address just 
as easily as a name if you have to. I've even done that in anger and not 
just to test out the idea when I had to contend with a dead name server.


It's having a name in the name server which may be thought of as analogous 
to cataloging a data set - sort-of.


Incidentally, in connection with an earlier post, I guess you realised that, 
by "no name server", Ulrich meant he wasn't *running* a name server not that 
he wasn't *using* a name server - and the name server specifications are 
actually in the resolver or the TCPIP.DATA files, not the PROFILE. You 
probably knew that too but your fingers let you down.


Chris Mason

- Original Message - 
From: "J R" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 3:09 PM
Subject: Re: Print Distribution and IP connected Devices



...
Using a URL is analogous to cataloging a dataset.  It
enables one to get there by name alone, without needing
to know the address, volser, etc.
...


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Chris Mason]

Ulrich

The AIX I knew deserved also to be classified as "UNIX".

One possible name server configuration could have been - or could be now - 
to have the bulk of the corporate names managed by a dedicated, typically 
traditional UNIX, machine, but to have the "mainframe" environment supported 
by a Communications Server (CS) IP name server linked into the higher level 
name server. At the very least the CS IP name server could act as a cache.


Chris Mason

- Original Message - 
From: "Ulrich Krueger" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 12:42 AM
Subject: Re: Print Distribution and IP connected Devices


Ted,
In that particular shop, a pair of UNIX (or was it AIX?) servers were used 
as the corporate DNS servers. All we ran on the mainframe were the usual 
suspects: FTP, SMTP, Telnet. Even though it was (and still is) the most 
important data processing environment of the company, the mainframe was 
never big enough to handle a DNS service task for the entire corporation.


Regards,
Ulrich Krueger

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Chris Mason]

Does anyone remember IBM's RMDS (Report Management and Distribution System), 
5648-048? I can quote the program number only because there's still some 
trace of it in the IBM publications site.


When this first appeared - together with the 3290 - it was supposed to 
herald the paperless office. Indeed some early design suggestions in the 
late '70s were that what became the 3290 display would actually replace the 
desktop and you would do all your reading electronically gazing down at the 
surface of your desk!


Chris Mason

- Original Message - 
From: "Timothy Sipples" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Thursday, August 23, 2007 10:03 AM
Subject: Re: Print Distribution and IP connected Devices


Thinking only slightly out of the box here, if you're going to have new
printers and the team setting them up refuses to give you their IP
addresses for whatever reason(s) -- I'd certainly like to understand more
about why -- would it make sense to e-mail print jobs as PDFs to end users?
That might save paper anyway since a lot of printouts really don't need to
be printed out.  It's also quite likely to provide some user productivity
benefits since, for example, it's very easy to perform keyword searches on
PDFs.  And it could be quite handy indeed for customer service applications
to, say, e-mail a PDF to a customer with a quotation or electronic copy of
an order.

IBM-MAIN archives contain some recent discussions about print-to-PDF and
the several ways to do it along with ideas about security and management
implications.

- - - - -
Timothy Sipples

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Strange behavior of START command(Keyword parms)

[Chris Mason]

Peter

Just what exactly is that "Note" from the description of the START command I 
quoted earlier in the thread supposed to mean? Perhaps some examples would 
help.


Now it has been acknowledged to be "working as coded" rather than "working 
as designed", is it possible than the logic parsing the character string is 
trying to treat the string following the "devicetype" as a "volumeserial" 
which, of course, has a 6-character limit. This assumes that the equal sign 
is regarded as a delimiter. There would also need to be some subsequent 
processing which somehow "lost" the false "volumeserial" and interpreted, 
for example, the DISP=SHR parameter as intended - although this isn't 
actually confirmed by Johnny's example.


Chris Mason

- Original Message - 
From: "Peter Relson" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Monday, August 27, 2007 5:13 PM
Subject: Strange behavior of START command(Keyword parms)



Getting
IEE308I STARTTERM LENGTH ERROR
when you issue
S STC1,3390,JOBNAME=JOHNNY,SUB=JES2

is an error. We suggest you open a PMR.

Peter Relson
z/OS Core Technology Design


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Strange behavior of START command(Keyword parms)

[Chris Mason]

I wonder what "bad advise" that may have been.

Reviewing my post I see only that I implied some suggestions regarding 
testing - and unearthed  a rather confusing comment regarding commas in the 
description of the START command in the manual.


Perhaps this is yet another case of not bothering to read everything in the 
post - typical of this contributor - "par for the course".


I wonder what significance using lower case for a name has. Perhaps there is 
a cultural angle here.


Given that the observed behaviour is declared errant by a representative 
from development, perhaps "the same *ought to be* true for operator 
commands" would be more accurate.


From the context "advise" takes on the role of a noun. Had the adjective 
been modified to become an adverb as in "bad" becoming "badly", "advise" 
could be understood to be a verb - with some other adjustments to the phrase 
of course. I wondered if this wasn't some transatlantic variation of the 
word "advice", as in "defence" becoming "defense" for example, but Googling 
didn't offer any confirmation of this surmise. Indeed, the normal English 
use was confirmed. For example:




American Heritage Dictionary

ad·vise (ad-viz')
v. ad·vised, ad·vis·ing, ad·vis·es

v. tr.
1. To offer advice to; counsel.
2. To recommend; suggest: advised patience.
3. Usage Problem To inform; notify.

v. intr.
1. To take counsel; consult: She advised with her associates.
2. To offer advice.



So it's a mystery.

Chris Mason

- Original Message - 
From: "Shmuel Metz (Seymour J.)" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Tuesday, August 28, 2007 1:51 AM
Subject: Re: Strange behavior of START command(Keyword parms)



...

Thanks. I learned from dealing with assembler macros that only if
some positional parms are omitted between two positional parms commas
are required to occupy the position.


The same is true for operator commands; chris gave you bad advise.
...


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Future date & USS

[Chris Mason]

Arun

The way the date relates to the USS I know is the option to set a date field 
in an USS message. The end-user can be presented with the - supposedly - 
current date by means of the DATE string included in the text of the USS 
message.


The manual says the following about DATE:



Character string: DATE

In message text: Current Date

Format: In the format and with the delimiter specified in the DATEFRM and 
DATEDLM operands on the USSTAB macroinstruction




I expect this means that, an USS message including the character string 
DATE while the system date was set as you described, would indeed have 
shown the date set, that is, your "future" date.


It hadn't occurred to me before that reference to the date - and time, in 
fact - could be another source of ambiguity caused by the (mis)use of the 
TLA USS.


Chris Mason

- Original Message - 
From: "arun kumar" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Tuesday, August 28, 2007 5:26 PM
Subject: Future date & USS


Hi Listers
 On our test LPAR, development and test team wanted us to set a future date 
to test a particualr scenario. We IPL-ed the system with CLOCKxx as below,

OPERATOR PROMPT
ETRMODE  NO
ETRZONE  NO
 and we were able to set to a future date without any problem from TSO.
 IKJ56650I TIME-03:29:10 PM. CPU-00:00:00 SERVICE-64068 SESSION-00:37:01 
JANUARY1,2008
 But, when we checked the date from USS, it seem to be displaying the 
current date/time, not the future.

//:>date
Tue Aug 28 15:29:02  2007
 The test team is more interested in USS having a future date. Request your 
suggestion please on how to set the USS time to future date.

TIA
Arun

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

DLSw Test Program? (Was OSA-Express2 and SNA Pcomm - loss of connectivity)

[Chris Mason]

Also posted in IBMTCP-L.

-

A week or so ago I requested if anyone knew of a, typically Windows-based, 
test program which would allow an 802.2 Test frame to be sent with a 
specified MAC and SAP address in order to test connectivity over, generally, 
DLSw.


Such a program exist(ed/s) in the OS/2 world and was called OS2PING.

I am repeating the appeal - one would hope, more clearly - given that the 
original was buried in a mass of text in a well-advanced thread.


Chris Mason

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: OSA

[Chris Mason]

Sheila

Indeed there is a wealth of information available from an OSA feature when 
using OSA/SF. Much fuss is made of OSA/SF not being needed when using the 
OSA feature purely for IP rather than the much derided - shudder - SNA. 
However, while not being actually *needed*, it is still very handy for the 
purposes of checking what is going on.


The OSA/SF command that I have found most useful in providing what is[1], I 
think, a comprehensive presentation of the information available in the OSA 
feature is


IOACMD QUERY ONE_OSA chpid

My notes say the following:



The data presented by this command is a complete list of all static data 
associated with the chpid comprising


1. the OSA itself
2. the partition identifications to which it is defined
3. each of the ports
4. the status of the channel unit addresses for this LPAR
5. the OSA Address Table (OAT)



You'll note incidentally that the command above is a TSO command and does 
*not* rely on the GUI. When I installed OSA/SF, setting up the GUI options 
was the biggest chore so I'm not sure the possibility to employ the OSA/SF 
GUI should be presented as a benefit!


Furthermore, being a TSO command, it fulfils the requirement John assumed 
that the OSA information could be made available without seeking out some 
means of access to the HMC.


Having said all that, I can't guarantee that the information Philip 
specifically wanted will be present in that QUERY command output. All I can 
say is that it jolly well ought to be!


Chris Mason

[1] Strictly "used to be" in my case since I haven't had "hands-on" in a 
while.


- Original Message - 
From: "Sheila Weissborn" <[EMAIL PROTECTED]>

Newsgroups: bit.listserv.ibm-main
To: 
Sent: Friday, August 24, 2007 9:13 PM
Subject: Re: OSA


It can also be queried using OSA/SF.  There is a REXX procedure IOACMD that
will do this.  OSA/SF also has a GUI interface.
Sheila Weissborn
Ohio Casualty Insurance

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: DFSMSdss SYSPRINT attributes

[Ted MacNEIL]

>report your dissatisfaction to IBM. They really 
do not take complaints off IBM-MAIN.

That has been suggested many times, to no avail.

-
Too busy driving to stop for gas!

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: How to switch UID under OMVS

[Shmuel Metz (Seymour J.)]

In <[EMAIL PROTECTED]>, on
08/28/2007
   at 03:40 PM, Tommy Tsui <[EMAIL PROTECTED]> said:

>Anyone know how to switch the UID under OMVs.

The same as in any Unix.

>I try to use SU but failed.

Not based on the data you posted.

>#id
>uid=0(EDPXPER) gid=2(SYS1)

Looks like UID 0 to me.

>when I invoke using the TSO OMVS with userid TEST001, it always show
>the EDPXPER and cannot be changed to other TEST001.

The su command changes the effective UID, not the userid. If you have
two userids[1] with UID 0, that's 2 too many.

[1] Other than those for daemons that require it.
 
-- 
 Shmuel (Seymour J.) Metz, SysProg and JOAT
 ISO position; see  
We don't care. We don't have to care, we're Congress.
(S877: The Shut up and Eat Your spam act of 2003)

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Print Distribution and IP connected Devices

[Ulrich Krueger]

Chris,
Am I showing my age, if I confess to knowing RMDS?

I once evaluated RMDS as a possible replacement for CA-View. We never
actually implemented RMDS in production, because halfway through the
project, CA offered such a sweet deal on license renewal for all CA
products, the RMDS project was cancelled.
But there was another reason for why we didn't convert to RMDS. As compared
to CA View and other report management software packages that I know of,
RMDS created individual VSAM files for each and every report or SYSOUT it
stored. File and disk space management (migration and backup) was left to
DFHSM. In our system, one of the CA View databases targeted for conversion
processed about 15000 reports arriving daily and being held for 100 days ...
can you imagine the size of the ICF catalogs and the disk space required for
1.5 million VSAM files? Shudder! 
And that was just one area. We had other report archive needs for different
report retention requirements, up to a year.
No offense to IBM and the RMDS developers, they had a pretty good product.
But it just didn't scale well to suit our needs.
As far as I know, that company is still using CA-View for SYSOUT archiving
and another product, RSD EOS, for a more complex report distribution system.

Regards,
Ulrich Krueger

-Original Message-
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf
Of Chris Mason
Sent: Sunday, September 02, 2007 06:55
To: IBM-MAIN@BAMA.UA.EDU
Subject: Re: Print Distribution and IP connected Devices

Does anyone remember IBM's RMDS (Report Management and Distribution System),

5648-048? I can quote the program number only because there's still some 
trace of it in the IBM publications site.

When this first appeared - together with the 3290 - it was supposed to 
herald the paperless office. Indeed some early design suggestions in the 
late '70s were that what became the 3290 display would actually replace the 
desktop and you would do all your reading electronically gazing down at the 
surface of your desk!

Chris Mason

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: DFSMSdss SYSPRINT attributes

[Ed Gould]

On Sep 2, 2007, at 7:42 AM, Kenneth E Tomiak wrote:

Get a license to use z/OS and report your dissatisfaction to IBM.  
They really

do not take complaints off IBM-MAIN.



Kenneth,

He (or the company he works for)does have license. Gil like to  
complain and some of his complaints are valid, IMO. This sounds (to  
me) like a valid SHARE requirement as it goes to a design issue. I  
would think his company belongs to SHARE and he should write up a  
requirement. Although I am not sure SHARE could with stand the number  
of requirements he generates:)


Ed

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Permission Problems On 3rd Open Of Same PDS

[Harry Goldschmitt]

At 7:21 AM -0500 9/2/07, Kenneth E Tomiak wrote:

On Fri, 31 Aug 2007 21:18:58 -0700, Harry Goldschmitt <[EMAIL PROTECTED]>
wrote:


I've got an LE C++ program that fopens 3 different members of the
same PDS.  The first 2 fopens and fcloses work fine, but the 3rd
fopen gets a permission error for some users.  I've traced the
problem a to missing OMVS segments for the failing users.  I've
resolved this in the past by just defining an OMVS segment in RACF
for the users that wish to run the program.  Now, however, I'm trying
to run it at a site with ACF2.  All users have a default OMVS segment
defined but I'm still getting the problem.

My application gets an "EDC5111I Permission denied.

 >(errno2=0x5b450002) DD:TEST3" message.




I do not follow your logic. If you had permission to use the dataset
USERID.TEST for the first two fopens then you have it for the third and your
real problem should be something else. Is it really USERID.TEST or
&SYSUID..TEST?


No, the DD statements match those I initially sent in my first message:

//TEST1 DD DISP=SHR,DSN=USERID.TEST(TEST1)
//TEST2 DD DISP=SHR,DSN=USERID.TEST(TEST2)
//TEST3 DD DISP=SHR,DSN=USERID.TEST(TEST3)


When you looked up errno2, what did it tell you?

http://publibz.boulder.ibm.com/cgi-
bin/bookmgr_OS390/BOOKS/BPXZA870/CCONTENTS

The current (requesting) process   
does not have the requested access 
authority to the file, or  
directory. 


Do they have ACF2 access to USERID.TEST?


They are all being opened and closed in order in the same task.

I did look up the reason code/errno2, and I can't find it.  It could 
be that it's being issued from ACF2.


I don't have the RACF reason code memorized, but when I've run into 
this in RACF, defining an OMVS segment for the users trying to run 
the program fixed the problem.  In fact, creating the OMVS segment 
with ACF2 worked on z/OS 1.7 at this location.  What doesn't work 
anymore is ACF2 on z/OS 1.8.


One of the earlier replies to my post suggested that the member may 
not exist.  I can't check at the ACF2 site till Tuesday, but when I 
try it under RACF, I get an S013 ABEND, just like you'd expect.


Harry

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Permission Problems On 3rd Open Of Same PDS

[JackOfAllTrades]

It is illogical that an OMVS segment is required to open the third file but 
not the first two.

http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/BPXZA870/CCONTENTS
Here is the link to the book's table of contents. From there look at 
Hierarchical File System (HFS) Reason Codes.

Then Module IDs
Scroll down to x'45' - 
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/BPXZA870/6.3.30?DT=20060615093634#HDRGCREA


- Original Message - 
From: "Harry Goldschmitt" <[EMAIL PROTECTED]>

To: "IBM Mainframe Discussion List" 
Cc: "Kenneth E Tomiak" <[EMAIL PROTECTED]>
Sent: Sunday, September 02, 2007 12:14 PM
Subject: Re: Permission Problems On 3rd Open Of Same PDS



At 7:21 AM -0500 9/2/07, Kenneth E Tomiak wrote:

On Fri, 31 Aug 2007 21:18:58 -0700, Harry Goldschmitt <[EMAIL PROTECTED]>
wrote:


I've got an LE C++ program that fopens 3 different members of the
same PDS.  The first 2 fopens and fcloses work fine, but the 3rd
fopen gets a permission error for some users.  I've traced the
problem a to missing OMVS segments for the failing users.  I've
resolved this in the past by just defining an OMVS segment in RACF
for the users that wish to run the program.  Now, however, I'm trying
to run it at a site with ACF2.  All users have a default OMVS segment
defined but I'm still getting the problem.

My application gets an "EDC5111I Permission denied.

 >(errno2=0x5b450002) DD:TEST3" message.




I do not follow your logic. If you had permission to use the dataset
USERID.TEST for the first two fopens then you have it for the third and 
your

real problem should be something else. Is it really USERID.TEST or
&SYSUID..TEST?


No, the DD statements match those I initially sent in my first message:

//TEST1 DD DISP=SHR,DSN=USERID.TEST(TEST1)
//TEST2 DD DISP=SHR,DSN=USERID.TEST(TEST2)
//TEST3 DD DISP=SHR,DSN=USERID.TEST(TEST3)


When you looked up errno2, what did it tell you?

http://publibz.boulder.ibm.com/cgi-
bin/bookmgr_OS390/BOOKS/BPXZA870/CCONTENTS

The current (requesting) process   does not have the requested access 
authority to the file, or  directory.

Do they have ACF2 access to USERID.TEST?


They are all being opened and closed in order in the same task.

I did look up the reason code/errno2, and I can't find it.  It could be 
that it's being issued from ACF2.


I don't have the RACF reason code memorized, but when I've run into this 
in RACF, defining an OMVS segment for the users trying to run the program 
fixed the problem.  In fact, creating the OMVS segment with ACF2 worked on 
z/OS 1.7 at this location.  What doesn't work anymore is ACF2 on z/OS 1.8.


One of the earlier replies to my post suggested that the member may not 
exist.  I can't check at the ACF2 site till Tuesday, but when I try it 
under RACF, I get an S013 ABEND, just like you'd expect.


Harry




--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: ISPF FREED SYS1.PARMLIB

[Skip Robinson]

We've never entertained the idea of concatenating the IPLPARM library along
with other PARMLIBs. The name says it all. It is used only at IPL time at
the earliest possible stage of initialization, so early that it does not
even need to cataloged because LOADxx actually tells the OS where to find
the master catalog. IPLPARM can also contain NUCLSTxx, used only to build
the nucleus.

None of the 'normal' members of IPLPARM are needed once IPL takes off.
Hence concatenating it with, say SYS1.PARMLIB, while not harmful, satisfies
no useful purpose.

.
.
JO.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler
SHARE MVS Program Co-Manager
626-302-7535 Office
323-715-0595 Mobile
[EMAIL PROTECTED]


   
 Steve Horein  
 <[EMAIL PROTECTED] 
 COURT.COM> To 
 Sent by: IBM  IBM-MAIN@BAMA.UA.EDU
 Mainframe  cc 
 Discussion List   
 <[EMAIL PROTECTED] Subject 
 .EDU> Re: ISPF FREED SYS1.PARMLIB 
   
   
 09/01/2007 08:23  
 AM
   
   
 Please respond to 
   IBM Mainframe   
  Discussion List  
 <[EMAIL PROTECTED] 
   .EDU>   
   
   




Thanks to everyone that contributed!

As suggested, SETLOAD was the way to go for me. I once again have a
properly sized SYS1.PARMLIB. However, while reviewing SETLOAD
documentation, questions spring to mind: Do you folks concatenate your
SYSx.IPLPARM (if you use it) to PARMLIB in LOADxx?

I was also curious if IEFPRMLB could be made to include by default the PDS
where LOADxx was found at IPL time, much like it includes by default
SYS1.PARMLIB.

I've recently taken to using LOADxx found in SYSx.IPLPARM on the IODF
volume instead of finding LOADxx in SYS1.PARMLIB. When I first issued my
SETLOAD command, it failed because I didn't specify DSN=SYSx.IPLPARM as I
should have, because at this time, I don't concatenate SYSx.IPLPARM. I
know, how I concatenate PARMLIB is my choice, but I'd hate to be viewed
as 'out there' if this isn't considered standard practice! ;)

I don't know if I'm OCD, but I like having any/all of my LOADxxs found in
one
place. I happen to choose SYSx.PARMLIB for that. I've also considered
creating a SYS1.LLA.PARMLIB (and system specific SYS1.LLA.PARMLIBs) to
contain CSVLLAxx members, excluding CSVLLA00.

I guess there's a line to draw between being organized and easily
maintainable.
Thanks again for being the group that you are!

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: SPAM: Re: Password expiration message at every SUBMIT at z/OS 1.8?

[Skip Robinson]

I'm not so sure about the guarantee of a 'real TSO' environment. I get the
idea that password expiration comes back in response to any new task
entering the system with userid validation. While the TSO SUBMIT command is
a common mechanism, there are others as well. CICS, for example, allows a
user to 'submit' a job. A batch job or a started task can input a job to an
internal reader. Where would PUTLINEs go in such cases?

I bring this up because a company I once worked with had very elaborate
RACF exit routines designed to send valuable but sometimes complex message
back to the 'submitting' user. TSO was a cake walk compared to some other
environments. Just keeping up with changing 372x devices was a challenge
that made me question the ultimate value of the messaging process.
.
.
JO.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler
SHARE MVS Program Co-Manager
626-302-7535 Office
323-715-0595 Mobile
[EMAIL PROTECTED]


   
 Binyamin Dissen   
 <[EMAIL PROTECTED] 
 FTWARE.COM>To 
 Sent by: IBM  IBM-MAIN@BAMA.UA.EDU
 Mainframe  cc 
 Discussion List   
 <[EMAIL PROTECTED] Subject 
 .EDU> Re: SPAM:  Re: Password expiration  
   message at every SUBMIT at z/OS 
   1.8?
 09/01/2007 07:47  
 PM
   
   
 Please respond to 
   IBM Mainframe   
  Discussion List  
 <[EMAIL PROTECTED] 
   .EDU>   
   
   




On Fri, 31 Aug 2007 16:54:27 -0500 Rick Fochtman <[EMAIL PROTECTED]> wrote:

:>
:>Wouldn't PUTLINE be a better choice than TPUT?
:>---
:>That would depend on how early in the process. Remember that some
:>control blocks may not be fully populated yet.

If I recall correctly, PUTLINE could be used at TSO pre-prompt time, which
is
before RACF validation.

On the other hand, as this code is guaranteed to be running on real TSO and
will be running where nothing is trapping the stack, there ain't much
difference.

--
Binyamin Dissen <[EMAIL PROTECTED]>

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Stand-Alone Dump Process

[Skip Robinson]

Once again my perspective. Store Status either will or will not 'save the
exact same data' depending on how you define 'same'. The process is more
(mostly? entirely?) a function of hardware. What you get is effectively the
register, PSW, etc. (too lazy to crack the book) data reflecting the state
of the control program that was last running in the LPAR at moment your
favorite operator Mashed Load. For an ordinary SAD, you almost certainly
want that data to represent the z/OS or VM or Linux (or TPF?) operating
system that failed beyond expectation of repair within a reasonable outage
window. So you almost always want exactly one and only Store Status for SAD
to capture.

In real life, many shops take advantage of HMC wizardry to define SAD load
profile(s). These profiles must be different from regular OS load profiles
because the IPL device for SAD is never a sysres volume and may differ from
one LPAR to another. You have the option in an SAD load profile--still, I
think?--to set the Store Status option. The temptation is to set it because
this is by definition a SAD profile. I used to do that until I was advised
otherwise: if the first shot at SAD goes awry for some transitory and
immediately correctable reason, you DO NOT want a second Store Status on
retry. Only the first Store Status, the one from the OS, is of any use
UNLESS you trying to debug SAD itself. BTW I have never once done that in
my whole career.
.
.
JO.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler
SHARE MVS Program Co-Manager
626-302-7535 Office
323-715-0595 Mobile
[EMAIL PROTECTED]


   
 Binyamin Dissen   
 <[EMAIL PROTECTED] 
 FTWARE.COM>To 
 Sent by: IBM  IBM-MAIN@BAMA.UA.EDU
 Mainframe  cc 
 Discussion List   
 <[EMAIL PROTECTED] Subject 
 .EDU> Re: Stand-Alone Dump Process
   
   
 09/01/2007 07:44  
 PM
   
   
 Please respond to 
   IBM Mainframe   
  Discussion List  
 <[EMAIL PROTECTED] 
   .EDU>   
   
   






Can issuing a STORE STATUS directly hurt?
Or does the IPL store status save the exact same data?

--
Binyamin Dissen <[EMAIL PROTECTED]>

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: DLSw Test Program? (Was OSA-Express2 and SNA Pcomm - loss of connectivity)

[Timothy Sipples]

OS2PING is very handy indeed.  (There's also OS/2's NETPING, which does
much the same for NetBIOS connections.)  You could boot into OS/2 and run
this command.  OS2PING (and NETPING) are not installed automatically, but
you can find them contained in the MPTSAPLT.ZIP file.  This file is located
somewhere within the MPTS (Multiprotocol Transport Services) feature of
OS/2.  Simply unzip what you need.  I would imagine today's modern
eComStation still has this, and conceivably you could make a bootable CD to
run it.

I've found a not-quite-as-handy Linux substitute, described here:

http://www.gcom.com/linux/lis/cmds.html#ldltest

See the second ldltest example for the relevant syntax.  Likewise, you may
be able to use this from a bootable Linux CD and thus make it a
near-universal troubleshooting tool from a particular workstation.

I haven't found a real equivalent for Windows, but perhaps someone else has
an idea.

- - - - -
Timothy Sipples
IBM Consulting Enterprise Software Architect
Specializing in Software Architectures Related to System z
Based in Tokyo, Serving IBM Japan and IBM Asia-Pacific
E-Mail: [EMAIL PROTECTED]
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

ISPF Display eXcluded lines (was: zOS Unix DOS2UNIX)

[Paul Gilmartin]

On Wed, 29 Aug 2007 21:17:37 -0500, McKown, John wrote:
>
>I use both "vi" (well actually vim on Linux) and OEDIT. Of course, I
>mean in the appropriate environment (OEDIT in TSO OMVS) and vi on a
>telnet login. Both have strengths and weaknesses. There are things that
>I can do so easily in OEDIT but have never figured out how to do in "vi"
>(like C 'A' 'B' ALL 5 to change all "A"s in column 5 to "B"s. Or use
>start and end labels). But I absolutely adore the use of regular
>expressions in "vi".
>
Me, too, on regex.

While we're on editor techniques, can I do with ISPF something
I've found useful in XEDIT:  momentarily and reversibly display
excluded lines.  For example, after doing the analogue of
EXCLUDE ALL and a few FINDs, I can do

DISPLAY 0
  ... which displays only the remaining excluded lines

DISPLAY 0 1
  ... which displays all lines, whether excluded or not

DISPLAY 1
  return to displaying only the nonexcluded lines.

(Please, no argument that this violates the essential semantic
of eXclude -- it's useful regardless.)

And one thing I value in vi is the ease with which I can imbed
sample output from a command in the document I'm editing:

:r ! some-command

... or pass the content of the edit buffer to a command, even to
shell itself:

:w ! sh

... this means I can do an instant test of a script without saving
to a file.  I wish I could do the same with ISPF and Rexx.

(I know, "You can do that with an edit macro."  And, in one case,
I have.  But with vi it's intrinsic and the technique is portable
to any system where vi runs (and vi runs on more of the systems
I use than does ISPF), and I can do it without schlepping along a
toolkit with me.)

-- gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Re: Stand-Alone Dump Process

[Barbara Nitz]

>So you almost always want exactly one and only Store Status for SAD
>to capture.

>In real life, many shops take advantage of HMC wizardry to define SAD load
>profile(s). These profiles must be different from regular OS load profiles
>because the IPL device for SAD is never a sysres volume and may differ
>from one LPAR to another. You have the option in an SAD load profile--
>still, I think?--to set the Store Status option. 

Yes, the option is still there. 

>The temptation is to set it because this is by definition a SAD profile.
>I used to do that until I was advised otherwise: if the first shot at SAD 
>>goes awry for some transitory and immediately correctable reason, you DO >NOT 
>want a second Store Status on retry. Only the first Store Status, the >one 
>from the OS, is of any use UNLESS you trying to debug SAD itself. BTW >I have 
>never once done that in my whole career.

Well, this thread comes in handy. I have just generated (and tested) sadump for 
1.8. And during testing, I clicked on the option for store status (despite 
having read NOT to do it, as I was really unsure). The sadump I then took was 
definitely readable and was NOT a dump of an sadump. It was showing everything 
I expected it to show, including all system address spaces that were up at the 
time.

I had been regenerating sadump from tape to DASD (dsntype LARGE), and my first 
attempt went wrong due to operator error. I re-IPL'd sadump (without store 
status, I think), and *that* was definitely a dump of a dump (sadump even tells 
me so).

So I guess now I am really confused with the whole issue of store status or not.

Regards, Barbara Nitz
-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html