Creating a century window in QUIZ/QTP for PROMPT variables.
Deskin, Bob
Bob.Deskin@Cognos.COM
Mon, 6 Apr 1998 11:50:35 -0400
I tested the following on UNIX (7.33.D), MPE/iX (7.29.C and 7.29.C8),
OpenVMS (7.10.G), and MPE/V (5.36C).
DEFINE X DATE CENTURY INCLUDED = 19981231
DEFINE Y DATE CENTURY INCLUDED = ADDCENTURY(X)
REP ALL
GO
In all cases, the attempt to add a century to a date that includes a
non-zero century (ADDCENTURY(X) ) gave a data conversion error. As I said
earlier, this is generic functionality and should work the same on all
platforms. It does, and it works correctly and as designed. We don't want to
change this because it would affect backward compatibility.
The system option applies to SYSDATE and any temporary or defined item that
allows the century included or excluded to default. PDL requires a
specification on the element of size 6 or 8. The system option also affects
whether some of the dates functions return a century included or excluded
result and whether that result (if century excluded) must be within the
default century. I've written an article about this that will appear in this
month's SupportLink.
I am not a proponent of making partial changes to applications. I understand
the difficulty in making wholesale changes to large applications, however
that is why the year 2000 issue is such a problem.
Bob Deskin
Senior Product Advisor bob.deskin@cognos.com
Cognos Inc. (613) 738-1338 ext 4205 FAX: (613) 228-3149
3755 Riverside Drive P.O. Box 9707 Stn. T, Ottawa ON K1G 4K9 CANADA
> ----------
> From: Blair Combs[SMTP:combblai@isu.edu]
> Sent: April 4, 1998 4:20 AM
> To: Deskin, Bob
> Cc: Powerhouse Discussion Listserv
> Subject: Re: Creating a century window in QUIZ/QTP for PROMPT
> variables.
>
> <<File: vcard.vcf>>
> Bob Deskin wrote:
>
> > Where did you here that the ADDCENTURY function would be changed? As
> > far as I know, there are no plans to change it in any version.
>
> We spoke with techincal support about our difficulties with
> the Year 2000 conversion process and they informed us of
> this feature. Apparently, based on what we have been told,
> this feature works on all platforms but MPE/iX. This does
> not give me a warm and fuzzy feeling concerning the level of
> support from Cognos for the MPE/iX platform.
>
> > On the subject of changing data and dictionary options, when you're
> > making changes, if you're going to century included dates, then you'll
> > want to change the dictionary option immediately. Otherwise your date
> > functions won't work correctly with multiple centuries.
> >
> > Bob Deskin
>
> My understanding according to the PDL manual is that any
> date of size 8 includes the century automatically and any 6
> digit date has the century excluded. Therefore the system
> option really only applies to the SYSDATE variable. Is this
> not true?
>
> The problem with changing the system option at the first is
> that we have numerous programs utilizing SYSDATE expecting 6
> digits. As we convert dates to 8 digits we use the
> ADDCENTURY function to be able to place the programs back
> into production right away. Converting dates 1 at a time
> allows us to more easily manage a very large date conversion
> process while at the same time continuing ongoing
> maintenance of systems and their programs.
>
> The ADDCENTURY function makes this possible. Now with the
> news concerning the reluctance of Cognos to correct this
> problem in 7.29 (they have logged the problem but tell us
> not to expect a correction) we will probably be forced to go
> back into all programs utilizing ADDCENTURY to remove this
> function as one of the final steps before changing the
> system option to include century, costing us hundreds of
> additional hours of work.
>
> Are we going to have problems using this method of
> conversion? You mention that date functions will not work
> across centuries without the system option specifying
> century included. Is this true even when a date function is
> performed against an 8 digit date that includes the century?
> Please let me know. I appreciate your insight.
>
> TIA,
>
> Blair Combs
> Idaho State Unversity
> combblai@isu.edu
> 208.236.2399
>
> *************************
>
> > ----------
> > From: Blair Combs[SMTP:combblai@isu.edu]
> > Sent: April 2, 1998 11:50 AM
> > To: Deskin, Bob
> > Cc: powerh-l@lists.swau.edu
> > Subject: Re: Creating a century window in QUIZ/QTP for PROMPT
> > variables.
> >
> > There is one item I thought should be mentioned here for the
> > benefit of the MPE/iX community at least. There are still
> > date related problems in version 7.29.C8 that Cognos seems
> > to be taking lightly that will create serious problems for
> > us and probably others.
> >
> > As you go through your systems converting dates you will
> > normally have a period of time when you will have some 8
> > digit dates as well as 6 digit dates utilized within the
> > same dictionary. I believe this to be very common. In this
> > situation, if your SYSTEM OPTIONS specify CENTURY EXCLUDED,
> > the SYSDATE variable will be incompatible with 8 digit dates
> > included in your dictionary. Consequently, the use of the
> > ADDCENTURY is required to convert the SYSDATE when working
> > with an 8 digit date. ADDCENTURY will be required whenever
> > you are working with two incompatible dates, one 8 digits
> > and the other 6 digits.
> >
> > We were all promised that the ADDCENTURY function would be
> > enhanced in version 7.29.C8 for MPE/iX to recognize a date
> > already includes the century and not attempt to add the
> > century again in this case. We have been in the date
> > conversion process now for two years with the expectation of
> > this feature being available with this release. All our
> > reprogramming has been along these lines. Now we are being
> > told, "Sorry, we are not enhancing 7.29 any longer and you
> > will have to wait for version 8.19 which should have this
> > feature working correctly."
> >
> > With the full release of 8.19 still months away Cognos is
> > putting their MPE/iX customers (or are we the only ones) in
> > a very difficult position.
> > Are we going to be forced into beta testing 8.19 with our
> > production systems because Cognos did not keep their promise
> > and are unwilling to correct the situation?
> >
> > I would really appreciate a response from Cognos on these
> > concerns posted here.
> >
> > TIA,
> >
> > Blair Combs
> > Idaho State University
> >
> > *********************
> >
> > Deskin, Bob wrote:
> >
> > << snip >>
> >
> > > The complaint versions (with the problems fixed) are:
> > >
> > > PowerHouse series 7
> > > MPE/iX - 7.29.C8 and up
> > > OpenVMS - 7.10.F2 and up
> > > UNIX - 7.33.D3 and up, except for DG Intel which is 7.33.E2 and up
> > >
> > > PowerHouse series 8
> > > MPE/iX - 8.09.C and up
> > > OpenVMS - 8.00.C1 and up
> > > UNIX - 8.03.C1 and up
> > > NT - 8.01.C1 and up
> > >
> > > Any appropriately supported customer can get the compliant versions
> > > free of charge. Contact Customer Support or Cognos Direct.
> > >
> > > Bob Deskin
> > > ----------
> >
>
= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
Subscribe: "subscribe powerh-l" in message body to majordomo@lists.swau.edu
Unsubscribe: "unsubscribe powerh-l" in message to majordomo@lists.swau.edu
powerh-l@lists.swau.edu is gatewayed one-way to bit.listserv.powerh-l
This list is closed, thus to post to the list, you must be a subscriber.