[Search for users]
[Overall Top Noters]
[List of all Conferences]
[Download this site]
Title: | RMS asks, 'R U Journaled?' |
|
Moderator: | STAR::TSPEER UVEL |
|
Created: | Tue Mar 11 1986 |
Last Modified: | Wed Jun 04 1997 |
Last Successful Update: | Fri Jun 06 1997 |
Number of topics: | 3031 |
Total number of notes: | 12302 |
2000.0. "YEAR 2000: RMS fully Y2K compliant" by STAR::EWOODS () Sat May 17 1997 15:53
I received a couple of inquiries last week whether RMS had a problem
with Year 2000. It seems appropriate to nip any rumors to the contrary
by adding a notes entry on the subject in the RMS notes conference.
Since there appears to be a proliferation of notes on the subject of
Year 2000, I ask that this note be confined to questions, comments and
information on the topic relevant to RMS file and record I/O.
I want to state unequivocably in this first entry that RMS is fully
Year 2000 compliant. This conclusion is based on a line-by-line inspection
of all the RMS and FAL code (and RMS utilities) by the RMS engineers.
There are two changes related to Year 2000 that RMS is planning to include
in the next version of OpenVMS after V7.1. Both changes are being added
as "enhancements" that provide options for RMS users which may facilitate
efforts to make their own code or data files Year 2000 compliant:
1. Segmented collated key support
Some users have ASCII date keys in their RMS indexed data
files that are based on a 2-digit year (for example,
YYMMDD or a Julian date of YYddd). By adding segmented
support to the existing collated key type, the 2-digit year
could be reordered as 80, 81, ..., 98, 99, 00, 01, ... 79,
without adding a century prefix to the existing date field.
2. FAL 128-bit UTC binary date/time support
OpenVMS RMS and FAL exchange date and time information about
files (e.g., file creation or revision dates and times) using
a 2-digit year in accordance with the Data Access Protocol (DAP,
Version 7.2, Section 5.15). Currently, OpenVMS RMS and FAL
implement this by pivoting the YY year field around the year 1970.
That is, YY's from 70 to 99 map to 1970 through 1999 and YY's from
00 to 69 map to 2000 through 2069. Since the year is pivoted,
OpenVMS will not see any problems with DECNET FAL access to file
dates until the year 2070. In other words, it is year 2000
compliant.
However, non-VMS clients of OpenVMS FAL (DOS FAL for example)
have the potential of breaking at the year 2000 if they haven't
built a similiar pivoting algorithm into their code. For this
reason, as an enhancement for non-VMS clients, support will be
added to OpenVMS RMS and FAL for passing dates using the 128-bit
Coordinated Universal Time (UTC). Support for UTC was added to
Version 7.2.3 of the DAP specification. A client of OpenVMS FAL
must enable the system capabilities (SYSCAP) bit 65 in order to
indicate any dates exchanged should be expressed in the UTC format.
Implementing this capability will not introduce any
incompatibilities with existing clients of OpenVMS RMS and FAL
since the old method will still remain intact.
Again I emphasize that OpenVMS RMS is fully Year 2000 compliant without
either of these changes.
RMS is also willing to backport both of these enhancements starting with
OpenVMS V6.1 through an existing distribution media: the RMS remedial
TIMA kits.
The first enhancement is currently being tested by a customer. To quote
him, he thinks this change will save his company "quite a wad of money."
Not only will it save them from having to add a century prefix to many data
files, but more importantly from having to modify code in a number of
applications. He is now testing it with their applications to verify
whether he is correct.
I should note, however, that the usefulness of the first enhancement is
language dependent for applications that use language RTL I/O
rather than direct calls to the RMS services. The basis for file and
record I/O that this customer will be using is the FORTRAN RTL. This
change (segmented collated key type) is transparent for existing files to
FORTRAN RTL. It will not work with COBOL. In fact, COBOL does not support
a collated key type -- segmented or not -- in an RMS indexed file.
Elinor Woods
PL, OpenVMS RMS Engineering
May 17, 1997
T.R | Title | User | Personal Name | Date | Lines |
---|
2000.1 | YEAR 2000 - Note # 2000 - not a coincidence | STAR::EWOODS | | Sat May 17 1997 16:22 | 8 |
| Within a few minutes of my posting .0 on a Saturday, Hein responded.
He felt this topic might be popular enough to merit a special place.
Leave it to Hein to come up with this very easy number to remember.
With a few nips and snips (nice to own a notes conference), it was made
possible.
Thanks, Hein!
|
2000.2 | OK to repost? | GIDDAY::GILLINGS | a crucible of informative mistakes | Mon May 19 1997 00:17 | 5 |
| Elinor,
It might be a good idea to place this in one or both of EVMS::Y2K and
siog::year2000. (I don't know why there are 2 conferences).
John Gillings, Sydney CSC
|
2000.3 | EVMS::Y2K | XDELTA::HOFFMAN | Steve, OpenVMS Engineering | Mon May 19 1997 18:55 | 8 |
|
: It might be a good idea to place this in one or both of EVMS::Y2K and
: siog::year2000. (I don't know why there are 2 conferences).
EVMS::Y2K is the OpenVMS Engineering Y2K conference and is the
right place for any OpenVMS discussions around the Year 2000.
(The other conference is rather more generic in nature.)
|
2000.4 | Nice addition... | KAOFS::B_CROOK | Brian @KAO | Wed May 21 1997 11:32 | 8 |
|
Elinor, thanks for that update, that is the kind of thing that will
save customers $$$'s.
I'm curious as to how the "Segmented collated key support" will be
added to RMS while remaining transparent, can you share an overview of
how the change will look to application developers who use RMS?
brian
|
2000.5 | | EVMS::EWOODS | | Mon May 26 1997 17:35 | 7 |
| brian --
I will write up the details and post them but it may not be for a week or
two. I have a few hot things on the fire and am pressed for time at the
moment. But I won't forget.
-- Elinor
|