On Fri, 2 May 2008 15:13:43 -0700, Schwarz, Barry A wrote:
>Yes it's great if you are allowed to connect your system to the
>internet. For the rest (few?) of us who cannot, we still need
>maintenance.
>
Get a laptop running Solaris, OS X, Linux, or Windows.
Step A:
- Connect the laptop to the Internet (you were able to post to this
list, weren't you?)
- Download the package contents to the laptop.
See:
Linkname: Requirements for a workstation, if used as an intermediate node
URL:
http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/EPD1UG14/3.3
Step B:
- Disconnect the laptop from the Internet, carry it into the computer
room, and connect it to the LAN there.
- Start an FTP server on the laptop.
- Use SMP/E RECEIVE FROMNETWORK to receive the package from the laptop.
I've performed Step B for all the systems mentioned except Linux, for
which I nonetheless have high confidence.
>And while we are on the subject, why does an internet download have to
>go to a Unix file? Is there some reason SMP/E couldn't handle a normal
>dataset? If I wanted a Unix system, I would have bought one.
>
Because UNIX is simpler, easier to use, and allowed implementation of
the function with only base z/OS software.
I was not privy to the design considerations, but I imagine the
designers wanted an archive scheme that was available with base z/OS;
and could be used with a non-z/OS-peculiar server. Compression is
also desirable. What are the options?:
o TSO TRANSMIT? Widely used, but requires TSO TMP. I suppose a
requirement to run SMP/E under TSO or vice-versa was unattractive.
And the format is not compressed, but quite the opposite with a
high overhead.
o AMATERSE? Not in base z/OS at the time the facility was developed.
o ADRDSSU? Don't know. Is it Internet-friendly? And I lately
discovered that there's a rule that to unpack an ADRDSSU archive
the user must have read access in the profile governing the
original data set name, even when unpacking to a different name;
a clear impediment to portability. I understand the rationale
for this rule, but it could have been relaxed somewhat and
still achieved its objective.
I suppose the first scheme they found that met the criteria was
the pax.Z format common in UNIX.
But why didn't they abandon relative files and put everything
inline with GIMDTS, compressed with GIMCPTS?
-- 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
