Nine days ago:  Leonard Berkowitz after Vikram  (I'm hopelessly behind on
email & etc. after going on leave):

> Vikram wrote:
>
> > The latest version of DBUNLOAD/ DBLOAD being released with
> > Lrgefile Dataset solution has a 'DISC' option using which
> > unload/load can be done to/from disc>>
>
> I find this ironic. The larger the data base, the less likely it is that
> the dynamic duo of DBUNLOAD/DBLOAD can offer a solution.
> Aside from taking up a large chunk of disk space for a large data
> base, the the time required will probably be very great. .....

While Leonard makes some valid points, I won't complain that HP
provided this enhancement.....   hmmm....:  Although I will note in
passing that if *I* could have decided how this enhancement would
have been delivered, I would have had HP turn DICTDBL and
DICTDBU into NMPRGs and bundle those tiny two standalone DICT
utilities instead (I believe DICTDBL and DICTDBU are written in
Transact;  and HP has the Transact/iX NM compiler;  so.....):

ACCOUNT=  SYS         GROUP=  PUB

FILENAME  CODE  --------LOGICAL RECORD
                  SIZE  TYP        EOF

DBLOAD    NMPRG   128W  FB         254
DBUNLOAD  NMPRG   128W  FB         277

DICTDBL   PROG    128W  FB         195
DICTDBU   PROG    128W  FB          84

DICTDBL and DICTDBU have several neat features over and above
DBLOAD and DBUNLOAD.  For example, DICTDBU asks:
=========================================

UNLOAD AUTOMATIC MASTER SETS(N/Y)?> N

UNLOAD DETAIL SETS BY CHAIN(Y/N)?> N

UNLOAD EDIT(N/Y)?> Y

DATA-ELEMENT     M: UNLOAD(Y/N/C/X)?> N
========================================

Last above allows you to selectively unload datasets;  or stop;  or
continue with all datasets at any time...

Let me hasten to add that even if HP did bundle DICTDBL and
DICTDBU, I realize DBLOAD and DBUNLOAD would have to "stick
around".

Ken Sletten

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *