Re: MVS 3.8J Turnkey 5 Update-3 announced.
On Sat, Apr 27, 2024 at 05:15 AM, Rob Prins wrote:
Greg stated: “It's taken 21 years, but this time for sure!”
"That trick never works!"? — Rocket J. Squirrel
|
Re: MVS 3.8J Turnkey 5 Update-3 announced.
toggle quoted message
Show quoted text
On April 27, 2024 7:15:47 AM CDT, Rob Prins <prin0096@...> wrote:
Hi Folks,
Today on the 57th birthday of our Dutch king I am pleased to announce the third update of
MVS 3.8J Turnkey 5. This update contains many new components and a huge amount of
changes.
The summary of updates are:
Larry Belmontes has developed a Tutorial for ISPF. This tutorial, LBTUTOR, can be accessed via ISPF option T.
The installed REVIEW panels have been converted to ISPF 2.2 format.
ISPF option 3.6 now provides a Hardcopy function (CPRT36).
The TSO command TSOEXEC is now available. TSOEXEC was introduced with the TSO/E Program Product not available for TK5 users.
To provide the same functionality for TK5 users Rob Prins has developed the TSOEXEC command. The purpose of
TSOEXEC is to execute authorized TSO command processors in a non-authorized environment.
The only restriction is that the TSO user should have access to resource SVC244 in the FACILITY class. If no security product is installed, then access is always granted.
Usermod RP00003 has been applied to SVC 24 so that the device information returned by DEVTYPE now provides the correct number
of cylinders for the various models of 3380 and 3390 DASD devices. See the document: “TK5 Usermod Directory”.
Hercules has been upgraded from 4.6 to 4.7. Only the Windows 64 and Linux 64 binaries are upgraded.
The bug in LOGON (,)APPLID=xxxxxxxx in ETHLOGVM has been resolved with a rework of usermod ZBP0002.
SYS1.TK5.INSTALL.CNTL has been moved from TK5RES to TK5001 and renamed to TK5.INSTALL.CNTL to better
reflect its purpose in providing the installation JCL used to install the packages on the TK5001 and TK5002 volumes.
ISPF option 3.7.V package DVTOC updated to V1R0M03 to fix an ABEND0C4 when running authorized.
Ed Liss has upgraded his MAP3270 package from 3.1.0 to 3.2.1.
Bob Polmanter has upgraded NJE38 from 242 to 250. The most significant update is the MEMBERS keyword
in the TRANSMIT command. With RECEIVE you can replace members in an existing partitioned dataset.
Package BREXX/370 has been updated from V2R5M2 to V2R5M3L01.
Rob Prins has updated RPF from V1R9M6 to V1R9M7. See RPF option 7.5 for a summary of the changes.
Greg Price has updated Review 50.4 to 51.2.
SPY full screen display of 3270 system console screens has been upgraded to support all terminal sizes.
TK5 Introduction and User Manual page 5
PDS command package upgraded from 8.6.20.0 to 8.6.22.0.
An I/O device system generation has been performed to include the following :-
o 16 CTCs at unit addresses 700-708 and 710-717
o 16 CTCs at unit addresses 800-808 and 810-817
o 32 non shared 3380 DASD at unit addresses 1E0-1EF and 2E0-2EF
o 32 non shared 3390 DASD at unit addresses 1F0-1FF and 2F0-2FF
The 3390 DASD WORKnn volumes now have a use attribute of PUBLIC instead of
STORAGE for standardized allocation of permanent data sets.
ZP60009, a usermod essential for the successful operation of full screen TSO/VTAM
applications, and with a long history of rework, has again been reworked by Greg Price.
Some issues were found with 3270 data stream processing for large 3270 screens. Greg
stated: “It's taken 21 years, but this time for sure!”
From a suggestion by Volker Bandke, the initial CLIST ISPLOGON is updated and the (B)REXX exec $CONCAT is added.
OS/360 Sort/Merge for MVS 3.8 updated to Version 1.03 to increase the seven digit record count fields in information messages to eleven digits.
HTTPD server is upgraded from 1.0 to 2.0.0
Please download MVS-TK5 Update-3 or the complete MVS-TK5 system with Update-3 included from:
To install the system, please read page 22 of the .
Please read also the
Have a lot of fun with this "new" version of MVS 3.8J
Cheers,
Rob
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
|
MVS 3.8J Turnkey 5 Update-3 announced.
Hi Folks,
Today on the 57th birthday of our Dutch king I am pleased to announce the third update of
MVS 3.8J Turnkey 5. This update contains many new components and a huge amount of
changes.
The summary of updates are:
Larry Belmontes has developed a Tutorial for ISPF. This tutorial, LBTUTOR, can be accessed via ISPF option T.
The installed REVIEW panels have been converted to ISPF 2.2 format.
ISPF option 3.6 now provides a Hardcopy function (CPRT36).
The TSO command TSOEXEC is now available. TSOEXEC was introduced with the TSO/E Program Product not available for TK5 users.
To provide the same functionality for TK5 users Rob Prins has developed the TSOEXEC command. The purpose of
TSOEXEC is to execute authorized TSO command processors in a non-authorized environment.
The only restriction is that the TSO user should have access to resource SVC244 in the FACILITY class. If no security product is installed, then access is always granted.
Usermod RP00003 has been applied to SVC 24 so that the device information returned by DEVTYPE now provides the correct number
of cylinders for the various models of 3380 and 3390 DASD devices. See the document: “TK5 Usermod Directory”.
Hercules has been upgraded from 4.6 to 4.7. Only the Windows 64 and Linux 64 binaries are upgraded.
The bug in LOGON (,)APPLID=xxxxxxxx in ETHLOGVM has been resolved with a rework of usermod ZBP0002.
SYS1.TK5.INSTALL.CNTL has been moved from TK5RES to TK5001 and renamed to TK5.INSTALL.CNTL to better
reflect its purpose in providing the installation JCL used to install the packages on the TK5001 and TK5002 volumes.
ISPF option 3.7.V package DVTOC updated to V1R0M03 to fix an ABEND0C4 when running authorized.
Ed Liss has upgraded his MAP3270 package from 3.1.0 to 3.2.1.
Bob Polmanter has upgraded NJE38 from 242 to 250. The most significant update is the MEMBERS keyword
in the TRANSMIT command. With RECEIVE you can replace members in an existing partitioned dataset.
Package BREXX/370 has been updated from V2R5M2 to V2R5M3L01.
Rob Prins has updated RPF from V1R9M6 to V1R9M7. See RPF option 7.5 for a summary of the changes.
Greg Price has updated Review 50.4 to 51.2.
SPY full screen display of 3270 system console screens has been upgraded to support all terminal sizes.
TK5 Introduction and User Manual page 5
PDS command package upgraded from 8.6.20.0 to 8.6.22.0.
An I/O device system generation has been performed to include the following :-
o 16 CTCs at unit addresses 700-708 and 710-717
o 16 CTCs at unit addresses 800-808 and 810-817
o 32 non shared 3380 DASD at unit addresses 1E0-1EF and 2E0-2EF
o 32 non shared 3390 DASD at unit addresses 1F0-1FF and 2F0-2FF
The 3390 DASD WORKnn volumes now have a use attribute of PUBLIC instead of
STORAGE for standardized allocation of permanent data sets.
ZP60009, a usermod essential for the successful operation of full screen TSO/VTAM
applications, and with a long history of rework, has again been reworked by Greg Price.
Some issues were found with 3270 data stream processing for large 3270 screens. Greg
stated: “It's taken 21 years, but this time for sure!”
From a suggestion by Volker Bandke, the initial CLIST ISPLOGON is updated and the (B)REXX exec $CONCAT is added.
OS/360 Sort/Merge for MVS 3.8 updated to Version 1.03 to increase the seven digit record count fields in information messages to eleven digits.
HTTPD server is upgraded from 1.0 to 2.0.0
Please download MVS-TK5 Update-3 or the complete MVS-TK5 system with Update-3 included from:
To install the system, please read page 22 of the .
Please read also the
Have a lot of fun with this "new" version of MVS 3.8J
Cheers,
Rob
|
Hi Rob,
thanks, it was my second try ;o) REVOUT is fine, even I miss some SDSF features.
Greets 骋á产辞谤
Hi Gabor,
Please press F1 for documentation about REVOUT or REVIEW.
Cheers,
Rob
|
Hi Gabor,
Please press F1 for documentation about REVOUT or REVIEW.
Cheers,
Rob
|
MSGCLASS H solved the problem, thank you all!
Nevertheless, I still miss the documentation for the field-developed components as REVOUT in thes case.
|
I use 5655-G53 IBM Enterprise COBOL for z/OS and OS/390 3.1.1
搁别苍é
Le mercredi 17 avril 2024 à 22:07:20 UTC+2, Rene BRANDT via groups.io <rbr146@...> a écrit :
"less memory means more i/o. more memory means less i/o."
With internal sort i give only 27 bytes so in memory it puts a lot more than the external where it use 80 bytes per records
I think the insert into the sort does each time a lot of work like if it initializes again some things
Is it a way to see what it does ?
Many thanks,
搁别苍é
Le mercredi 17 avril 2024 à 21:55:45 UTC+2, Ed Liss <egliss4024@...> a écrit :
I see that this was run on a Z/OS system using something other than MVT COBOL (COBOL 2?).? BUT I would refer you to a COBOL manual for the compiler you are using as well as a DFSORT manual to find out about the defaults for the memory, sort techniques and COBOL special registers related to internal sorts.
Bottom line for sorting comes down to "less memory means more i/o.? more memory means less i/o.".??
|
"less memory means more i/o. more memory means less i/o."
With internal sort i give only 27 bytes so in memory it puts a lot more than the external where it use 80 bytes per records
I think the insert into the sort does each time a lot of work like if it initializes again some things
Is it a way to see what it does ?
Many thanks,
搁别苍é
Le mercredi 17 avril 2024 à 21:55:45 UTC+2, Ed Liss <egliss4024@...> a écrit :
I see that this was run on a Z/OS system using something other than MVT COBOL (COBOL 2?).? BUT I would refer you to a COBOL manual for the compiler you are using as well as a DFSORT manual to find out about the defaults for the memory, sort techniques and COBOL special registers related to internal sorts.
Bottom line for sorting comes down to "less memory means more i/o.? more memory means less i/o.".??
|
I see that this was run on a Z/OS system using something other than MVT COBOL (COBOL 2?).? BUT I would refer you to a COBOL manual for the compiler you are using as well as a DFSORT manual to find out about the defaults for the memory, sort techniques and COBOL special registers related to internal sorts.
Bottom line for sorting comes down to "less memory means more i/o.? more memory means less i/o.".??
|
On 15/04/2024 7:33 pm, Gabor Markon via groups.io wrote:
Hello all,
I've installed recently the TK5, and cannot refind my compile output in REVOUT.
Under TK4 I've found it already out, where a HERC01x output is landed, but I've lost the setting, how can I set it in my jobcard.
For example, my jobname is HERC01P. If I call 'STATUS HERC01P' in the REVOKE' menu, it does not show anything.
Any hints please?
If you issue ST HERC01P as a TSO command (such as from the READY prompt, for example) is your job found? If not, it has already printed and been purged from the spool. Set MSGCLASS to something like X or H on the JOB statement to prevent the job from being printed. Cheers, Greg
|
here is the file with all inside
搁别苍é
|
With internal sort we can have only the used part of the record in the output section, so it should be more records in memory, but with that the external sort is quicker.
So i think that each time the cobol insert a record the sort program redoes a lot of things (for me not needed)...
I will put the programs, jcls and data.
搁别苍é
Le mardi 16 avril 2024 à 15:35:59 UTC+2, Ed Liss <egliss4024@...> a écrit :
Both the internal and external sort use the same sort software.? Factors such a region size, number of sort work files, estimated record counts, etc determine sort speed.? In addition, COBOL internal sort uses defaults optimized for small regions (64k in the case of MVT COBOL).? The sort utility standalone will use all available memory in the region. Internal sorts are constrained because they must share the region with the COBOL program.
|
Both the internal and external sort use the same sort software.? Factors such a region size, number of sort work files, estimated record counts, etc determine sort speed.? In addition, COBOL internal sort uses defaults optimized for small regions (64k in the case of MVT COBOL).? The sort utility standalone will use all available memory in the region. Internal sorts are constrained because they must share the region with the COBOL program.
|
In an external sort it reads and writes the records. >Then a batch reads again teh the records and writes them again.
With a cobol and the internal sort the cobol reads the records gives to sort (memory to memory) and sort gives back the records (memory to memory) and the cobol writes the records.
So with internal sort we have half of i/os so we should be quicker than external sort. So Why external sort is quicker ? Is the internal sort so badly written or do i forgot something ?
搁别苍é
|
Hello all,
I've installed recently the TK5, and cannot refind my compile output in REVOUT.
Under TK4 I've found it already out, where a HERC01x output is landed, but I've lost the setting, how can I set it in my jobcard.
For example, my jobname is HERC01P. If I call 'STATUS HERC01P' in the REVOKE' menu, it does not show anything.
Any hints please?
|
Re: MVS-TK5 startup problem on Linux
Noted, thanks for the revision.
Vince
toggle quoted message
Show quoted text
On 15/04/2024 01:15, Gavin Scott via groups.io wrote: On Sun, Apr 14, 2024 at 12:16 PM, Vince Coen wrote:
Please note every one there is TK5 update 3 available from the usual place : Per Rob's recent message where he said he's waiting for additional software updates and expects to release Update 3 sometime in May, this is almost certainly not what Update 3 will be in its final form, and you're probably just discovering the current work-in-progress version that's being shared between the people developing/testing TK5 along with Rob.
I would not be downloading anything labeled Update 3 until it has been officially announced as being ready and is linked directly from the main TK5 page.
|
Re: MVS-TK5 startup problem on Linux
On Sun, Apr 14, 2024 at 12:16 PM, Vince Coen wrote: Please note every one there is TK5 update 3 available from the usual place : Per Rob's recent message where he said he's waiting for additional software updates and expects to release Update 3 sometime in May, this is almost certainly not what Update 3 will be in its final form, and you're probably just discovering the current work-in-progress version that's being shared between the people developing/testing TK5 along with Rob. I would not be downloading anything labeled Update 3 until it has been officially announced as being ready and is linked directly from the main TK5 page.
|
Re: MVS-TK5 startup problem on Linux
On 04.14.2024 06:26, kb2ghz via groups.io wrote: have tried to install and run MVS-TK5 update 2 on Ubuntu, Debian, and Fedora.
Hercules has crashed! (Illegal instruction)
mvs_ipl: line 57:? 6095 Illegal instruction? ? ?(core dumped) $force_arch hercules $DAEMON -f conf/tk5.cnf > log/3033.log Illegal instruction is probably due to the bundled Hercules binary included in TK5 being compiled for a different processor than what you have, and is trying to use an instruction that your processor doesn't support. You'll probably be better off compiling your own copy of Hercules. hercules-helper can help you out there: <> After you have your own Hercules, I'm not sure how you convince the tk5 start script to use it instead of the bundled one, though. Other folks have done so, though, so might be able to tell you how. -Matthew
|
Re: MVS-TK5 startup problem on Linux
Please note every one there is TK5 update 3 available from the usual place :
then near the end is Download Update 3 which is linked to :
Vince
toggle quoted message
Show quoted text
On 14/04/2024 14:26, kb2ghz via groups.io wrote: ?have tried to install and run MVS-TK5 update 2 on Ubuntu, Debian, and Fedora. Each time i get the error messages below. My ulimit on all of the systems is unlimited, and i followed the instillation instructions in the manual. Any thought or suggestions would be appreciated. Mike McCann KB2GHZ
|
MVS-TK5 startup problem on Linux
?have tried to install and run MVS-TK5 update 2 on Ubuntu, Debian, and Fedora. Each time i get the error messages below. My ulimit on all of the systems is unlimited, and i followed the instillation instructions in the manual. Any thought or suggestions would be appreciated. Mike McCann KB2GHZ
test@kb2ghz:/usr1/MVS-TK5/mvs-tk5$ ./mvs
?
?
?
? ? +++ OOPS! +++
?
Hercules has crashed! (Illegal instruction)
?
Crash dump NOT created due to ulimit setting...
?
mvs_ipl: line 57:? 6095 Illegal instruction? ? ?(core dumped) $force_arch hercules $DAEMON -f conf/tk5.cnf > log/3033.log
|
Rob Prins