Re: QCCSID = 65535 - What release/Notice to users was this discussed?

[CRPence <CRPbottle@xxxxxxxxx>]

  While Jeff and others will relay there are\were no ill-effects from 
changing QCCSID, beware, that should not be inferred to be true for all. 
 Those who have made the change may not even [still] know of some 
negative effects, or may have corrected any errors either without having 
been informed of the need to correct or without proper attribution to 
the origin [the QCCSID change] for the need to correct.  Understand the 
system value change for probable side effects, and try to resolve the 
issues before implementing the change.  If all user profiles currently 
use CCSID(*SYSVAL), changing the system value from 65535 will have the 
maximum impact, if any; users with a _default_ coded character set 
identifier that differs from a new\changed value due to their LANGID 
[language identifier] setting, those users likely will have the highest 
potential for seeing negative effects in that scenario.

  IIRC, the CCSID was introduced in v2r1m1 for the database; source and 
data physical files.  Every database file created before then should 
have been assigned an /assumed/ CCSID according to the installed 
language, with the opportunity to CHGPF CCSID() to override that 
assignment\assumption.  Note: The CHGPF CCSID() invocation has negative 
side-effects for any other utilization, than to correct the implicitly 
assigned CCSID from the transition from pre-v2r1m1.

  If every user profile is established with an appropriate individual 
CCSID [for their language and keyboard], then the SysVal QCCSID is of 
lesser impact and relevance.  In fact, leaving all system user profiles 
CCSID(*SYSVAL), leaving QCCSID *HEX, and changing individual user 
profiles to have a non-hex CCSID would have less impact than changing 
QCCSID and leaving all user profiles with a default of redirecting to 
the *SYSVAL.

Regards, Chuck

On 12/6/10 7:31 AM, Jeff Crosby wrote:
> I changed it in the middle of the day some years back with no ill
> effects. Jerry Adams has stated the same thing.
>
> On Mon, Dec 6, 2010 at 9:42 AM, Vern Hamberg wrote:
> > I think this has been known and discussed since the system value
> > was put on the system - IOW, maybe even the System/38.
> >
> > Although it has seemed to be common knowledge, I understand your
> > need. I don't know without searching ibm.com for QCCSID and 66535,
> > which is what I'd have to do.
> >
> > There was something at V5R3, that the native JDBC driver would not
> > work if your job QCCSID was 66535. Check out the Memo to Users at
> >
> > http://publib.boulder.ibm.com/infocenter/iseries/v5r3/topic/rzahg/rzaq9.pdf
> >
> > And search for 66535
> >
> > On 12/6/2010 7:52 AM, Michael Ryan wrote:
> > > Working with a system manager. I want the QCCSID value changed to
> > > 37 and I feel he could use some reassurance. I'm comfortable with
> > > the change, but being able to point to the Notice to Users would
> > > be a help. Thanks!