As I rewrote my MVS 3.8 installation instructions, I decided to
add this page as a way of handling miscellaneous questions that might
arise. I also tried to remember and include some of the questions that
have been asked and answered in the group discussions. I will be adding to
this page, so it might be a good idea to check back from time to time,
especially if you have a question you need answered.
[December 2006] I have been made
aware of a compatibility issue when text files (scripts and JCL) from the
archives on my site are used on Linux/UNIX systems:
All your shell-Scripts/jcls and *.cnf have the Control-Character "<CTRL>M" at the end of
the line when I make a "tar -xvzf installmvs.tgz" on the ubuntu Linux-Box and then
shell-Scripts are not executable within the Shell.
Even though I create all archives using tar under Cygwin rather
than using a Windows archive program to make them as compatible as possible
between Linux/UNIX and Windows users, this is a problem that can't have a
universal solution. There is a utility that may be run on Linux/UNIX
systems - fromdos - that will translate these files so that they will function
correctly. There are a number of sources for the utility; use the search
term 'fromdos' in your favorite search engine or use this link:
[February 2018] As I was making an
update to the website today, it occurred to me that I have been remiss in not
mentioning somewhere on my site that the exact situation described above has now
become the reverse. I left the Windows world in 2017, most likely for the
final time, so now I am creating/editing all text datasets under Linux.
This means that the lines are terminated without including the
"<CTRL>M", which will make it necessary for those utilizing the
text datasets on a Windows system to use the Tofrodos utility to add the
characters to the end of the lines. An alternative is to edit the datasets
with the freely available Notepad++ utility, which will correctly edit the
datasets. I apologize that I have not mentioned this in a more timely
manner, but as I have not had anyone pose a question to me about the text
datasets, perhaps it has not been as much of a problem as I thought it might be.