¿ªÔÆÌåÓý

Writing / reading text files with tapes


 

Before I go down the route of punching files etc. is there a simple tool to read and add ASCII files, at the pc end, to tapes, covering to ebcdic and so on. I have looked but can't find much although I had assumed this was key functionality.

Sorry if this is a stupid question!


 

Have you explored OMA tapes?? Never tried them under VM, but they work under MVS.

ScottC

On Friday, January 31, 2020, 11:36:31 AM PST, adriansutherland67 <adrian@...> wrote:


Before I go down the route of punching files etc. is there a simple tool to read and add ASCII files, at the pc end, to tapes, covering to ebcdic and so on. I have looked but can't find much although I had assumed this was key functionality.

Sorry if this is a stupid question!


 

On Fri, Jan 31, 2020 at 07:44 PM, ScottC wrote:
Have you explored OMA tapes??
Thanks Scott ... Looking at the manual, am I right in thinking these only work in the direction pc to host?


 

¿ªÔÆÌåÓý

Yes. There are some programs to around to take CMS ¡°tape dump¡± files and extract the files

?

?

Dave

?

From: [email protected] <[email protected]> On Behalf Of adriansutherland67
Sent: 31 January 2020 19:58
To: [email protected]
Subject: Re: [h390-vm] Writing / reading text files with tapes

?

On Fri, Jan 31, 2020 at 07:44 PM, ScottC wrote:

Have you explored OMA tapes??

Thanks Scott ... Looking at the manual, am I right in thinking these only work in the direction pc to host?


 

Content-type: text/plain; charset=UTF-8
Content-transfer-encoding: 8BIT


Yes. There are some programs to around to take CMS ¡°tape dump¡± files and
extract the files
I think the problem with tape dump files is either they are specific to a
version of CMS or they are specific to a disk filesystem (ie EDF or non-EDF).
Whichever it is, the result is that it is very easy to end up with a tape dump
format tape which cannot be read by VM/370 CMS.

For text files containing records that are 80 characters or less wide, the
easiest approach is to use the "real" card reader and card punch. For
everything else, I prefer to use VMARC which is like Zip for VM. There is
a version installed on the Sixpack systems (as VMARC370). There is also a
version called VMA which can be compiled and used on the host system, maybe
with a little effort. Both can manipulate the same archive files in card
image format which can be sent through the reader and punch.

Have we already had this conversation recently?
I feel like it's deja-vu all over again :-)

Regards,
Peter Coghlan.


 

On Fri, Jan 31, 2020 at 08:34 PM, Peter Coghlan wrote:
I feel like it's deja-vu all over again :-)
Apologies ... I made a hack to batch up text files for ind$file, I can extend that and use cards, but from the comments here I was wondering if there was an existing solution.

That said, I get that for a binary distribution for VM/370 tape is the way to go. For source however ...


 

¿ªÔÆÌåÓý

VMARC and friends can handle any type of file format or record size and restore to same.

George Shedlock

On 1/31/2020 3:46 PM, adriansutherland67 wrote:

On Fri, Jan 31, 2020 at 08:34 PM, Peter Coghlan wrote:
I feel like it's deja-vu all over again :-)
Apologies ... I made a hack to batch up text files for ind$file, I can extend that and use cards, but from the comments here I was wondering if there was an existing solution.

That said, I get that for a binary distribution for VM/370 tape is the way to go. For source however ...

Virus-free.



 

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Peter
Coghlan
Sent: 31 January 2020 20:32
To: [email protected]
Subject: Re: [h390-vm] Writing / reading text files with tapes

Content-type: text/plain; charset=UTF-8
Content-transfer-encoding: 8BIT


Yes. There are some programs to around to take CMS ?€?tape dump?€?
files and extract the files
I think the problem with tape dump files is either they are specific to a
version of CMS or they are specific to a disk filesystem (ie EDF or non-EDF).
Whichever it is, the result is that it is very easy to end up with a tape dump
format tape which cannot be read by VM/370 CMS.

For text files containing records that are 80 characters or less wide, the
easiest approach is to use the "real" card reader and card punch. For
everything else, I prefer to use VMARC which is like Zip for VM. There is a
version installed on the Sixpack systems (as VMARC370). There is also a
version called VMA which can be compiled and used on the host system,
maybe with a little effort. Both can manipulate the same archive files in card
image format which can be sent through the reader and punch.

Have we already had this conversation recently?
I feel like it's deja-vu all over again :-)

Probably
Dave


Regards,
Peter Coghlan.



 

On 1/31/20 11:36 AM, adriansutherland67 wrote:
Before I go down the route of punching files etc. is there a simple tool to read and add ASCII files, at the pc end, to tapes, covering to ebcdic and so on. I have looked but can't find much although I had assumed this was key functionality.

Sorry if this is a stupid question!
What are you trying actually to do?? Load files onto VM?? Get files off?


 

On Fri, 31 Jan 2020 at 21:00, George Shedlock <gshedlock352@...> wrote:
VMARC and friends can handle any type of file format or record size and restore to same.

Is the format of the archive used by vmarc documented anywhere?? if one were to use the pack function to put a number of CMS files into an archive file and then transfer this to another system, for?example by IND$FILE, are there likely to be utilities to unpack the archive?? Likewise can one pack files into an archive on another system and unpack onto CMS with vmarc?

The previous discussion may have?been the one I started - I was able to get IND$FILE working and also send files to the real punch but I think the current question is more about being able to transfer a number of files at a time,


 

On Fri, Jan 31, 2020 at 09:56 PM, Steven Fosdick wrote:
?? if one were to use the pack function to put a number of CMS files into an archive file and then transfer this to another system, for?example by IND$FILE, are there likely to be utilities to unpack the archive?
This is the hack I produced to do that ...?



Given no alternatives I will clean it up and make it produce an archive with 80 char lines ...?


 


Is the format of the archive used by vmarc documented anywhere?
I don't know. It probably is. I've never found any need to delve into
the format any more than knowing that it is in 80 column card image format
which can readily be sent through a punch / card reader combination or over
an NJE link and each file starts with a header like this (in EBCDIC):

:CFF FNAME FTYPE FM

and that VMARC doesn't mind if there is garbage in the archive file (such as
USERID or :READ for example) before the first valid file header. The actual
file contents could be compressed with LZW compression and I never found any
compelling reason to know how that works.


if one
were to use the pack function to put a number of CMS files into an archive
file and then transfer this to another system, for example by IND$FILE, are
there likely to be utilities to unpack the archive? Likewise can one pack
files into an archive on another system and unpack onto CMS with vmarc?
This is exactly what VMARC was designed for. VMARC is available as S/370
assembler source which can be built on anything from VM/370 CMS to z/VM CMS.
I've done a mod which enables a limited VMARC version to be built on MUSIC.
I wouldn't be surprised if someone else tweaked it to make it buildable on
MVS and other mainframe operating systems.

If you have a Sixpack system up and running, log in to some CMS userid and
type VMARC370 and you should get brief usage details.

VMA is available as C source which can be compiled on any reasonable system
with a functioning C compiler. I have compiled it on my OpenVMS system, I
think most others use it on Windows or Linux. The purpose of VMA is to
read write and list VMARC format archive files on systems other than VM/CMS.


The previous discussion may have been the one I started - I was able to get
IND$FILE working and also send files to the real punch but I think the
current question is more about being able to transfer a number of files at
a time,
I think VMARC is ideal for this purpose but we seem to keep heading off in
the direction of emulated tapes or newly invented archive formats both of
which I think are not so ideal.

Regards,
Peter Coghlan.


 

On Fri, 31 Jan 2020 at 23:46, Peter Coghlan <groups@...> wrote:
VMA is available as C source which can be compiled on any reasonable system
with a functioning C compiler.? I have compiled it on my OpenVMS system, I
think most others use it on Windows or Linux.? The purpose of VMA is to
read write and list VMARC format archive files on systems other than VM/CMS.

Thanks to Ren¨¦ for e-mailing me a link to a video that describes searching for the source with "vma utility homerow" which leads to

I did try searching before to find a portable implementation of the amarc format but even hearing the utility is called vma didn't seem?to be enough to find it.

> The previous discussion may have been the one I started - I was able to get
> IND$FILE working and also send files to the real punch but I think the
> current question is more about being able to transfer a number of files at
> a time,

I think VMARC is ideal for this purpose but we seem to keep heading off in
the direction of emulated tapes or newly invented archive formats both of
which I think are not so ideal.

I think it's quite common to start to solve a problem yourself, get to a point where you're stuck and ask for help on the thing your stuck on.? Helping you with that specific point can lead to a solution but sometimes, if someone persuades you to rewind to the problem you started out with it turns out there is a better solution.?
?


 

And there is the fantastic VMAGUI that runs on various platforms.? I use it very frequently on Windows 10 to look at the contents of my VM stuff that I save on my laptop as VMARC files.
I even allows to add members to a VMARC file, something I didn't dare yet (all my VMARC files are built on VM; I treat them as R/O files on my laptop).
--
Kris Buelens


 

On Sat, Feb 1, 2020 at 09:39 AM, Steven Fosdick wrote:
Thanks to Ren¨¦ for e-mailing me a link to a video that describes searching for the source with "vma utility homerow" which leads to
?
Good - and it certainly seems to meets all needs

For me it is a bit heavy weight - I just want something as a text file transport: ASCII/EBCDIC can be done by Hercules/IND$FILE, I don't care about compression, and don't want to worry about long term storage - so I will plough my lonely furrow; everyone can use what tools they will!?:-)


 

On Sat, Feb 1, 2020 at 10:30 AM, adriansutherland67 wrote:
so I will plough my lonely furrow
To that end - I have renamed it to SPLCON to YATA "Yet Another Text Archive", and will never mention it again! (probably)


 

¿ªÔÆÌåÓý

Thanks for posting that link to the homerow site. ?The precompiled vma and vmagui utilities work excellently on my Win10 system.

?

I contributed code changes many years ago to the assembler implementation of VMARC maintained by Ross Paterson after he took over from John Fischer (the original author) to allow it to run on VM/370 under Herc.? I went searching for Ross Paterson¡¯s geocities site, which is sadly only available at the oocities archive site now.

?

?

The ¡°full download¡± of the VMARC assembler implementation is still available on that archive in base64 format.

?

?

On Win10 systems you can use the builtin certutil command to convert the base64 text to a file usable with the vma or vmagui utilities.

?

C:\VMARC>certutil -decode vmarc.txt vmarc.vma

?

I think the linux base64 command will do the same conversion for you but I didn¡¯t try it myself yet.

?

I haven¡¯t looked, but I would assume the VM/370 six-pack system already includes this code, but it is nice to know the original is still available for reference.

?

Peter

?

From: [email protected] <[email protected]> On Behalf Of Steven Fosdick
Sent: Saturday, February 1, 2020 4:40 AM
To: [email protected]
Subject: Re: [h390-vm] Writing / reading text files with tapes

?

On Fri, 31 Jan 2020 at 23:46, Peter Coghlan <groups@...> wrote:

VMA is available as C source which can be compiled on any reasonable system
with a functioning C compiler.? I have compiled it on my OpenVMS system, I
think most others use it on Windows or Linux.? The purpose of VMA is to
read write and list VMARC format archive files on systems other than VM/CMS.

?

Thanks to Ren¨¦ for e-mailing me a link to a video that describes searching for the source with "vma utility homerow" which leads to

?

I did try searching before to find a portable implementation of the amarc format but even hearing the utility is called vma didn't seem?to be enough to find it.


> The previous discussion may have been the one I started - I was able to get
> IND$FILE working and also send files to the real punch but I think the
> current question is more about being able to transfer a number of files at
> a time,

I think VMARC is ideal for this purpose but we seem to keep heading off in
the direction of emulated tapes or newly invented archive formats both of
which I think are not so ideal.

?

I think it's quite common to start to solve a problem yourself, get to a point where you're stuck and ask for help on the thing your stuck on.? Helping you with that specific point can lead to a solution but sometimes, if someone persuades you to rewind to the problem you started out with it turns out there is a better solution.?

?