[email protected] (DASDBILL2) writes:
> VM was called CP67 in 1967.  It became VM several years later.  CP67
> would only run on a S/360 model 67.  VM would run on any S/370 system
> with paging architecture.

re:
http://www.garlic.com/~lynn/2014d.html#54 Difference between MVS and z / OS 
systems
http://www.garlic.com/~lynn/2014d.html#55 Difference between MVS and z / OS 
systems
http://www.garlic.com/~lynn/2014d.html#56 Difference between MVS and z / OS 
systems

one of the first distributed development efforts using the internal
network technology was joint between the science center and endicott to
modify cp67 to support a 370 virtual machine (cp67 running on real
360/67). then a version of cp67 was modified to run on 370 virtual
memory archived .... running in the 370 virtual machine (on a real
360/67). this was in regular use a year before the first 370
enginneering processing with virtual memory support was operational.

then a couple of engineers came out from san jose and added 3330 & 2305
device support to the (370 version) cp67.

this effort was where the cms multi-level source update process was
developed.

base system was "cp67l" (that actually ran on real 360/67 at the
science center

370 virtual machine support was "cp67h" ... ran in 360/67 virtual
machine ...  the issue at the science center was that students and other
people from univ. systems in the boston area had access to the science
center system ... and there was concern about them being exposed to
unannounced 370 features.

370 version was "cp67i" which ran in a 370 virtual machine (provided
by cp67h).

for quite some time cp67i (or cp67sj with 3330 & 2305 support) was the
primary operating system running internal on 370 hardware ... even after
vm370 became available (even if cp67 370-version wasn't shipped to
customers). some discussion in this recent post
http://www.garlic.com/~lynn/2014d.html#22 Mainframe memories

One of the primary issues was that in the initial morph from cp67 to
vm370 they did a lot of simplification and dropped a lot of features
.... including my fastpath work (that significantly cut pathlengths), my
page replacement algorithm (that was significantly better at choosing
page to replace) and my dynamic adaptive resource management.

old email eventually moving those features from cp67 base to vm370
base
http://www.garlic.com/~lynn/2006v.html#email731212
http://www.garlic.com/~lynn/2006w.html#email750102
http://www.garlic.com/~lynn/2006w.html#email750430

after that there was increasing use of my csc/vm system distribution
internally ... as well as decision to pickup some of the features for
release in product.

as an aside, I also had an argument early in the os/vs2 prototype its
page replacement algorith ... that they were doing a very poor design
... and eventually resulting in agreeing to disagree. the issue was that
they were taking a very myopic optimization at the micro level ... that
resulted in making things worse at the macro level. Well into the MVS
product cycle, somebody realized that they were choosing non-changed,
highuse, shared, linkpack page for replacement before choosing changed,
lowuse, private application data pages.

-- 
virtualization experience starting Jan1968, online at home since Mar1970

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to