Hi Donna, not trying to beat a dead horse but from the output you sent
below, it looks like the logon from the STARTSESS command failed due to
the wrong password and that the default logon that occurs after SYSSTART
is done then succeeded.

Also, note in the first output where you don't supply the password in
STARTSESS it was simply LOGON FOR: and that in the second output where
you do hardcode the password, it says (PROGRAMMATIC) LOGON FOR: (which
is what the console msg looks like for a STARTSESS).

And, yes, you are right that moving the STARTSESS to COOLSTART wouldn't
make a difference. I just wanted to be sure there wasn't a STARTSESS cmd
at the end of STARTUP and another at the end of COOLSTART (and you've
confirmed that there isn't).

Donna Garverick wrote:
<snip>
> =====================================================
>               COOLSTART startup block processing
> =====================================================
>
> STARTSESS 20;CONSOLE,OPERATOR.SYS;NOWAIT;HIPRI
> 13:47/1/INVALID PASSWORD FOR "CONSOLE,OPERATOR.SYS," ....
> (see 'console' is clearly present and it's being sent to the system via
> startsess)

Yes but the password was bad so the logon didn't occur.

> <<snippage>>
> 13:47/#S1/49/LOGON FOR: "OPERATOR.SYS,SYSOPER" ON LDEV #20.

'CONSOLE' is missing from the message. So is '(PROGRAMMATIC)'. Looks
suspicious.

>
> 3 - ok fine!  change the sysstart file again to include the password on the
> startsess line.  blah blah blah.  here's the text from the console....
>
> ==============================================================
>               COOLSTART startup block processing
> ==============================================================
>
> STARTSESS 20;CONSOLE,OPERATOR/.SYS/;NOWAIT;HIPRI
> <<snippage>>
> 13:56/#S1/49/(PROGRAMMATIC) LOGON FOR: "CONSOLE,OPERATOR.SYS,SYSOPER"
>
> friggin' fraggin' -- to me this is clearly a bug.  the absence of the
> password makes the startsess command fail.

<snip>

While I never discount the possibility of a bug, I'm not quite convinced
in this case. I have to install a few patches on a box today anyway so
I'll test SYSSTART vs STARTSESS vs hardcoded passwords at the same time.
I'll let you know if my milage varied from yours :-)