2nd level install duplicate volsers 2495

ref:

for a little drift on the subject of changing from a real memory constrained environment and the uptake of rdbms:

now the original relational-sql implementation was done on vm370

and then there was tech transfer from sjr to endicott for sql-ds.

for a little more topic drift ... when we were doing ha-cmp

I've commented before that two of the people in the following meeting

later showed up in a small client-server startup responsible for something called commerce server

however, one of the other people in that same meeting had commented that he had done most of the initial work for the tech transfer of sql-ds from endicott back to stl for db2 (even tho bldg. 28 and bldg. 90 were only ten miles apart).

another topic drift is the overhead of doing the channel program translation to produce the shadow channel program. i've posted before about doing a lot of early performance work on enhancing (in large part mft & mft) performance on cp67 ... misc. past posts

after the above mentioned presentation at the '68 boston share meeting, i had also done some amount of path length special case optimization for cms filesystem disk i-o. however, i got strongly admonished by adair that I had violated virtual machine architecture and it needed to be morphed (using diagnose instruction) so as to preserve achitecture purity. misc. past posts referencing converting the cms filesystem disk i-o optimization to diagnose instruction:

somewhat in response, one of the things that I did in the early 70s at the science center

for cp67 cms was implement page mapped interface for cms filesystem support ... which eliminated the whole paradigm of having to emulate real i-o (and all the buttociated overhead gorp, at least for disks) in a virtual memory environment

this was later ported to vm370 cms and deployed succesfully on internal systems.

--