T.R | Title | User | Personal Name | Date | Lines |
---|
2190.1 | Until DaveT answers | AIMTEC::WICKS_A | MUP(pets) coming to ALL-IN-1 soon? | Tue Feb 02 1993 04:50 | 12 |
| Caroline,
P11D sounds a bit too much like some sort of Inland Revenueish account for
a tax exile such as myself to try so I used I18N (obvious I know but)
and it worked on my system (v3.0 unpatched). I guess we might need
some extract from the log or confirmation that you don't have any
old v2.4 system management customisations lying around for me.
Regards,
Andrew.D.Wicks
|
2190.2 | more on the RENAME problem | SUBURB::A1_BRUCE | | Wed Feb 03 1993 13:06 | 25 |
|
On further investigation is seems that this account rename wont work
whatever username you try and rename it to, the reason for failure is,
An invalid VMS account name was supplied
I've run with trace (SM_RENAME.SCP) and its branching to this error at
the following check,
.LABEL check_old_vms_name
GET #old_vms_username = PROFIL.VMSUSR[#a1_username]
.IF UAI$.USERNAME[#old_vms_username] EQS '' THEN .GOTO not_there
nb. #a1_username equals the offending source username
The UAI$USERNAME[#old_vms_username] is returning nothing ?
If I do a <GET OA$DISPLAY = UAI$.USERNAME["A1_BRUCE"] I get A1_BRUCE
returned, If I do it for the account that is failing I get blank
returned. Interestingly I get blank returned for the MANAGER account
also. Any ideas anyone ?
Bruce
|
2190.3 | REMOVE AND ADD THE OFFENDING USERNAME IN AUTHORIZE | SUBURB::A1_BRUCE | | Thu Feb 04 1993 10:44 | 12 |
|
I removed, then recreated, the authorize entry for this account using
the same uic and directory spec etc, and now the UAI$.USERNAME call works.
I have just submitted the RNA from ALL-IN-1 and it worked first time.
Something strange is happening with the UAI$ code I think.
Regards,
Bruce
|
2190.6 | Update, for those who are interested. | SCOTTC::MARSHALL | Spitfire Drivers Do It Topless | Tue Feb 09 1993 10:36 | 22 |
| Well,
I can reproduce it easily enough with ALL-IN-1. But when I run a simple program
to mimic what ALL-IN-1 is doing, that works fine. I've arranged with Caroline
that when I have time I'll stick a debug image on PEKING to see exactly what's
going on.
My current guess is that one of the items in the massive item list ALL-IN-1
uses (and which I didn't try and mimic in the test program) is making VMS look
at some duff data in the UAF record, and hence causing an error. AUTHORIZE
still works though, as it only looks at a subset of the UAF record, and is
presumably ignoring the offending field(s).
The other possibility (very likely considering the number of errors I've found
in GETUAI documentation to date!) is that some of the new fields in VMS V5.4
(upon which UAI$ is based) are incorrectly documented. This might cause buffer
overflows, etc, if VMS tries to return more data than the books say it will,
and hence more than ALL-IN-1 is expecting.
More info as soon as we have it, and remember - you heard it first here :-)
Scott
|