Re: RTVMSG - 'Text not available for message'

Repost to correct API name. Also it seems, not only does the MSG() get populated from the CPF2457, but so do the other parameters specified, according to the doc link provided in the quoted message. That is, "If the message file is found, but does not contain a description for the message identifier, the message attributes *and* text of message CPF2457 in QCPFMSG are used in place of the missing message description."

Regards, Chuck

CRPence wrote:

Not documented in [all] the /correct/ place(s) perhaps, but it is
[even if not well] documented.

http://publib.boulder.ibm.com/infocenter/iseries/v7r1m0/topic/rbam6/smsg.htm

<ed: did not confirm API name before posting; corrected below>

The RTVMSG CL and the Retrieve Message (QMHRTVM) API are both
affected by the documented result of message CPF2457 in QCPFMSG
being returned for the condition of a message identifier not
found in the specified message file. And although not explicitly
stated, presumably the MSGF(*LIBL/QCPFMSG) is utilized to locate
the CPF2457. I can only guess that the effect for MSGID(CPF2457)
not found in MSGF(QCPFMSG) might be blanks :-) since *that* is
[rightly] not documented ;-)

The status message can be monitored, e.g. MONMSG, and it is
listed as such on the RTVMSG help text.

Bruce Vining wrote:

You are correct -- it's not documented and that is (somewhat) what is returned for the MSG parameter. Which really, really
surprises me.

Is your &TXT variable defined as *Char 30? The reason I ask is
that the actual text returned is 'Text not available for
message XXXXXXX file YYYYYYYY.' which would not compare as *EQ
if &TXT is defined as longer than 30 characters.

The CPF2419 should really be signaled as an *ESCAPE, rather
than *STATUS, which makes me wonder if "someone" in the deep,
dark past was too scared to add an escape to this command
(years after it became available).

Would anyone with a non-English NLV installed care to try this?
I suspect (hope) it's picking up the text from a message file,
in which case this code would not work properly outside of an English environment. If, on the other hand, it still comes back
as English then "someone" has hardcoded English text in their code. A very nasty thing to do.

In any case, this should be reported to IBM.

On Thu, May 6, 2010 at 9:24 AM, Robert Rogerson wrote:

A base package program program (CLLE) uses the RTVMSG command
to decide which message file to use.

RTVMSG MSGID(&MSGID) MSGF(TFMMSG) MSG(&TXT)

IF COND(&TXT *EQ 'Text not available for message') THEN(DO)
SNDPGMMSG MSGID(&MSGID) MSGF(WFMMSG) +
MSGDTA(&MSGDTA) TOPGMQ(*SAME +
&SFPGMQ) KEYVAR(&MSGNO)
ENDDO
ELSE CMD(DO)
SNDPGMMSG MSGID(&MSGID) MSGF(TFMMSG) +
MSGDTA(&MSGDTA) TOPGMQ(*SAME +
&SFPGMQ) KEYVAR(&MSGNO)
ENDDO

It works fine but I'm curious where &TXT is being populated
with 'Text not available for message' after the RTVMSG, if
not found in message file TFMMSG.

I checked and RTVMSG doesn't mention this text is returned if
the message is not found. It does mention CPF2419 is returned
so I checked the WRKRPYLE (Work with System Reply List
Entries) but this message was not found.

Is it just undocumented or did I miss something where RTVMSG
does populate the MSG parameter if the message is not found?