> dave5550 wrote:

><snip>

> - The version 3 CD dates from July/2002. Version 4 has been promised
> for a long time. Anyone know if and when it is coming?

Several of us folks have worked on TK4 and then some of us got sick,
really sick. Volker which has been the primary force for the MVS
turnkey has had little to no time in the past year. I recently
underwent a hospital stay for an unexpected problem.

I spearheaded a beta TK4 release back in December. A fixpack
jobstream added corrections to items we overlooked or discovered after
some public testing was done. The Beta was posted on a pay site for a
couple of months and removed after downloads dropped off. Not much
activity has been going on in the TK4 developer group since that time.

> - As I'm sure other people have found, there is a Cygwin problem in
> the CD's setup script, which causes it to fail, even though I told
> it not to install Cygwin or Hercules. When setup tries to use some
> commands, Cygwin complains that there are 2 copies of file
> cygwin1.dll and refuses to do the command.

><snip>

This is a know issue, but don't know if there is a workaround posted
on Volkers' site or not. My individual fix was to rename the CYGWIN
directory to CYGWIN-real, install TK3, then rename CYGWIN-rel back to
CYGWIN.

> - For the future, I think it would be good to "decouple" the MVS 3.8
> install from Cygwin and the Hercules install, at least for the
> Windows case. Most people don't use Cygwin now (it seems to be
> fragile, especially for new users), and the Hercules install process
> is fairly well documented elsewhere. And, if MVS Turnkey really
> needs a setup script, it should not depend on bash or Cygwin. My
> feeling is that a script is not necessary, and I would prefer not to
> run a large "black box" like that unless I know it is bullet-proof.
> Why not just provide the DASD volume files and a sample Hercules
> config file that users can edit, along with instructions on what
> they may need to change in the config file? The config file is well
> documented on the Hercules home page.

There has been plenty of discussion on scripting language and choice
of Hercules executable selection in the TK4 group.

The selection of minimum vs full-blown MVS DASD farm ( source & CBT
files ) is currently a function of the scripts. This has also been
discussed.

> - The Turnkey MVS home page www.bsp-gmbh.com/turnkey/index.html is
> very useful, and I appreciate Volker's efforts. It would help if
> some of the pages had dates on them. I spent considerable time
> looking at the fixes page, for example, before realizing all those
> fixes are already in the version 3 CD.

Don't know that anyone other than Volker has access to modify those
pages. Don't know if Volker has date/time stamps on them offline or not.

><snip>

Phil - keep the suggestions coming - we do listen to them

•  2
  New Members
•  1
  New Files

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

I am trying to use the DELINKI program to delink a load module
(convert a load module to object files) on the Turnkey MVS 3.8
system. DELINKI is CBT090 on www.cbttape.org . Since the PL/1 (F)
compiler on MVS 3.8 is too low a level to compile the PL/I DELINKI
source (many compile errors), I got the DELINKI and DWNSPDSR load
module members from CBT135. I loaded them to MVS from the cbt135.xmi
file via the DASDLOAD utility of Hercules, to a 3350 DASD volume.
But when I try to run DELINKI, I get abend S0C4 (a Program Interrupt
for a Protection Exception) and no other output.

Has anyone got DELINKI to work on MVS 3.8? If so, how?

As I said, I can't build it from the source because the PL/1 (F)
compiler is inadequate (MANY errors at compile time).

A search on Google found a reference to Greg Price's REVIEW utility
(CBT134), which uses DELINKI. It says "set &SYSSPLV to '1' to avoid
S0C4 abend on pre-XA systems due to STAX macro expansion
differences". So, I suspect the problem is that the DELINKI load
module in CBT135 was created on a later version of MVS and does not
work in MVS 3.8 because of a problem in the STAX macro. This is just
a guess. But I have no idea how to fix a problem like that.

Any help or ideas would be much appreciated.

- Dave Edwards

•  1
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

Dave,

IMO you won't get DELINKI to work under MVS 3.8 in a fit.
It requires an upgraded compiler which is not currently available.

However, you can use the DELINK0 program from CBT file 316.
(I have thought about plugging that into REVIEW for MVS 3.8,
but so far thinking is all I've done.)

BTW, the STAX macro is what is used to handle attention interrupts
(usually from the PA1 key) under TSO.

The parameter address word is a victim of byte-stuffing, so to
allow STAX to set up a parameter for an attention exit where the
parameter is above the line a new STAX parameter list format
was invented.

However, the new STAX parameter list format is not understood
properly by pre-XA systems such as 3.8. Hence, I discovered
long ago that PA1 would trigger an S0C4 abend because of a
bad address.

I altered REVIEW to always use the original format, but now I find
that my copy of the PDS 8.5 load module has the same problem.
(Being able to use interrupt under the PDS command is *very*
useful.) I should get around to assembling my own load module
but I haven't.

But I digress...

Cheers,
Greg

•  1
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

Thanks Greg. I'll try delinko. Also, I may try to write my own
delink under MUSIC/SP. Its LOADPDS utility can read load module
members from an IEBCOPY/unload tape and store the lmod records as
variable-length records in a MUSIC/SP file. I have a LKED PLM
describing the records - looks simple enough. My goal is to convert
the Fortran G lmods to object, so I can port Fort G to MUSIC/SP.
Other IBM Fortrans can run under MUSIC/SP but they have license
issues.

- Dave Edwards

--- In turnkey-mvs <at> yahoogroups.com, "Greg Price" <greg.price <at> ...>
wrote:
>
> Dave,
>
> IMO you won't get DELINKI to work under MVS 3.8 in a fit.
> It requires an upgraded compiler which is not currently available.
>
> However, you can use the DELINK0 program from CBT file 316.
> (I have thought about plugging that into REVIEW for MVS 3.8,
> but so far thinking is all I've done.)
>
> BTW, the STAX macro is what is used to handle attention interrupts
> (usually from the PA1 key) under TSO.
>
> The parameter address word is a victim of byte-stuffing, so to
> allow STAX to set up a parameter for an attention exit where the
> parameter is above the line a new STAX parameter list format
> was invented.
>
> However, the new STAX parameter list format is not understood
> properly by pre-XA systems such as 3.8. Hence, I discovered
> long ago that PA1 would trigger an S0C4 abend because of a
> bad address.
>
> I altered REVIEW to always use the original format, but now I find
> that my copy of the PDS 8.5 load module has the same problem.
> (Being able to use interrupt under the PDS command is *very*
> useful.) I should get around to assembling my own load module
> but I haven't.
>
> But I digress...
>
> Cheers,
> Greg
>

•  3
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

More on this... I found a MUSIC/SP utility called DLINK that can
delink an MVS lmod nicely, in combination with the LOADPDS utility.
LOADPDS copies an MVS load module (or modules) from an IEBCOPY/unload
tape to MUSIC/SP files. DLINK was written by Kevin McNamee back in the
early 1980s. I had forgotten about it. It worked fine for the Fortran
G IEY* compiler modules, and I now have the Fort G compiler working on
MUSIC/SP. Next step is the library. I will distribute this stuff in
the next update to the MUSIC/SP Demo system.

- Dave Edwards

--- In turnkey-mvs <at> yahoogroups.com, "dave5550" <dave5550 <at> ...> wrote:
>
> Thanks Greg. I'll try delinko. Also, I may try to write my own
> delink under MUSIC/SP. Its LOADPDS utility can read load module
> members from an IEBCOPY/unload tape and store the lmod records as
> variable-length records in a MUSIC/SP file. I have a LKED PLM
> describing the records - looks simple enough. My goal is to convert
> the Fortran G lmods to object, so I can port Fort G to MUSIC/SP.
> Other IBM Fortrans can run under MUSIC/SP but they have license
> issues.

•  2
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

Hi,

Does anyone still have a copy of this file by Enrico Sorichetti hanging
around. I'd like to see how he uses REXX to extract files from XMI
files - I'm trying to write something that can actually create a PDS in
XMIT format and REXX is both readable and easily traceable, which might
make his code a good start.

The file seems to have been uploaded in Feb 2005, it's no longer there
and Enrico hasn't replied to a private email.

Thanks,

Robert

•  3
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

--- "Robert AH Prins" <robert_ah_prins <at> ...> wrote:
> Does anyone still have a copy of this file by Enrico Sorichetti
> hanging around. I'd like to see how he uses REXX to extract files
> from XMI files - I'm trying to write something that can actually
> create a PDS in XMIT format and REXX is both readable and easily
> traceable, which might make his code a good start.

Hi Rob,

I know I have a copy somewhere but I can't find it right now.
You might also check out the recv390 program in C, it has a
lot of debug aids, particularly in expert mode. You can
trace lots of things, and the source isn't all that bad if
you can read C. Let me know if I can be of some assistance,
I've dug into this stuff in the past.

http://cbttape.org/~jmorrison/mvs38j/recv390.zip

•  3
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

Hi Jim,

--- In turnkey-mvs <at> yahoogroups.com, "Jim Morrison" <jim.n9gtm <at> ...>
wrote:
>
> --- "Robert AH Prins" <robert_ah_prins <at> > wrote:
> > Does anyone still have a copy of this file by Enrico Sorichetti
> > hanging around. I'd like to see how he uses REXX to extract files
> > from XMI files - I'm trying to write something that can actually
> > create a PDS in XMIT format and REXX is both readable and easily
> > traceable, which might make his code a good start.
>
> I know I have a copy somewhere but I can't find it right now.

If at some stage you do find it, drop me a line please!

> You might also check out the recv390 program in C, it has a
> lot of debug aids, particularly in expert mode. You can
> trace lots of things, and the source isn't all that bad if
> you can read C. Let me know if I can be of some assistance,
> I've dug into this stuff in the past.
>
> http://cbttape.org/~jmorrison/mvs38j/recv390.zip

I've got a copy of your RECV390, but have some (understatement...)
trouble reading C - right now I'm trying to translate it into Pascal
so that I can add code to handle PS, VB and possibly even packed
datasets, the latter using code from MVShelp. At a later stage I
would like to create something that can actually create a PDS (a la
Tachyon's Re-Xmit) - creating a FB PS dataset in XMIT format is
actually pretty simple, the exec I use to backup my 'toolbox'
datasets actually creates a .BAT file in xmit format to 'recv390' all
members and give them a proper extension and the correct date.

Robert

•  4
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network

--- In turnkey-mvs <at> yahoogroups.com, "Robert AH Prins" <robert_ah_prins <at> ...> wrote:
>
> Hi,
>
> Does anyone still have a copy of this file by Enrico Sorichetti hanging
> around. I'd like to see how he uses REXX to extract files from XMI
> files - I'm trying to write something that can actually create a PDS in
> XMIT format and REXX is both readable and easily traceable, which might
> make his code a good start.
>
> The file seems to have been uploaded in Feb 2005, it's no longer there
> and Enrico hasn't replied to a private email.
>
> Thanks,
>
> Robert
>

Hi Robert,

sorry for not replying to Your email, but I gone thru a few changes in my internet
connection and i lost a few/many emails.

I have reworked my rexx a little bit and I will reload it shortly

creating a xmit rexx is in my objectives but a few things have to be straightened out
both in the receive an the xmit side.

to make analysis easier I split the problems in two parts each

receive side

strip the xmit wrapper in order to get a sequential "iebcopied" dataset

write a pc IEBCOPY to extract the members ,
this step might be useful also to manage "IEBCOPY" datasets residing on AWS/HET tapes

xmit side the other way around

create a iebcopy like sequential dataset

build the xmit wrapper around it

the only issue to meditate on is how to fake a nonexstent directory structure and dataset
allocation; I have to study on how to build the directory ttr's

regards

enrico

PS

sorry again for losing my emails

•  4
  New Members

• Operating system
• Operating system software
• Window operating system software
• Computer operating system software
• Software operating system network