| T.R | Title | User | Personal Name
 | Date | Lines | 
|---|
| 1943.1 | Try the log file... | MOVIES::GROSSMITH | Mike - VMS Devo/Europe | Tue Dec 19 1989 11:16 | 9 | 
|  | It might be worth looking in DECW$SM.LOG, to see if there is anything of
interest in there.
Of course, this might more difficult than it sounds, if you can't log in at all!
You might have to do a MINIMUM boot, and start the machine without a window
system, then just log in on OPA0:
Good luck,
/Mike
 | 
| 1943.2 | No luck there | ADDIT::OWEN | Call me mad, but then who's sane ? | Tue Dec 19 1989 11:52 | 9 | 
|  |     
    I've looked there, and all it has is a description of my working set and
    nothing else. I've also sifted through various log files with DECW in
    front of them and come up with a big zero there too.
    
    I'm considering doing the upgrade again, but if anyone could save my
    time I would appreciate it.
    
    John.
 | 
| 1943.3 | Any special "customizations"? | AGBEAR::HORNER | A.G.Bear, Old fashion teddy bear | Tue Dec 19 1989 12:32 | 10 | 
|  |     Did you have any "custom" methods for starting applications before the
    upgrade?  We ran into many incompatibilities between the way we did
    things in DECW V1 and the way things now work in DECW V2.  By removing
    all of our workarounds and unsupported utilities and starting with a
    clean slate, all problems disappeared.  None of them were as severe as
    what you are experiencing, but you never know.  I've also noticed that
    DECW V2 is a little more resource hungry than V1.  Make sure that your
    UAF and SYSGEN quotas and parameters are sufficient.
                Dave
 | 
| 1943.4 | Another suggestion... | MOVIES::GROSSMITH | Mike - VMS Devo/Europe | Tue Dec 19 1989 13:24 | 12 | 
|  | Yet another idea. Have you done an AUTOGEN lately? Might be an idea to log into
OPA0:, when DECwindows is waiting for login, and do an AUTOGEN with feedback.
.-1's statement about resources is very valid. We saw problems too with V2, only
in our case we didn't even get the LOGIN window. I think the problem is that
when the UPGRADE does an AUTOGEN, DECwindows is not running at all, and if any
feedback data you have lying around is not to AUTOGEN's liking, (too old
perhaps), then it is ignored. The result is that AUTOGEN gives you fewer
resources than you had before, when you may actually need more.
Just a thought,
/Mike
 | 
| 1943.5 | do you define decw$display by hand ? | WAIT::DESAI | Jatin Desai | Tue Dec 19 1989 16:50 | 5 | 
|  | 
We had some trouble with people who define decw$display instead of 
SET/DISPLAY/NODE command.
Jatin
 | 
| 1943.6 | Nothing seems to work. | ADDIT::OWEN | Call me mad, but then who's sane ? | Wed Dec 20 1989 05:47 | 12 | 
|  |     Hi.
    
    I've Autogen'd it daily (over 24 hrs) and have removed all previous
    customisations of windows etc, I have also check my SYSGEN paramaters
    with those of another machine, though a different machine. If anyone
    has 5.3 and DECWindows running on a WS3200 could they please post a
    copy of thier SYSGEN parameters  and an example UAF account for me to
    check off against my own.
    
    Thanks,
    	John.
    
 | 
| 1943.7 |  | KOBAL::SCAER |  | Wed Dec 20 1989 14:16 | 9 | 
|  | 
	Re: .0
	I had the same problem as you are describing and it was fixed when
	I put the following lines at the top of my login.com:
	$ DEVICE = F$TRNLNM("SYS$COMMAND")
	$ DEVICE = F$EXTRACT(F$LOCATE("$",DEVICE)+1,99,DEVICE)
	$ IF F$EXTRACT(0,2,DEVICE) .EQS. "MB" THEN EXIT
 | 
| 1943.8 | Check for SET TERM/INQ or INQ | DECWIN::FISHER | Burns Fisher 381-1466, ZKO3-4/W23 | Wed Dec 20 1989 15:14 | 7 | 
|  |     I was just about to suggest something like .-1.  If you do anything
    in login.com that assumes that INTERACTIVE implies that you have a
    terminal, symptoms like this will happen.  Thus, $INQUIRE, $SET
    TERM/INQ, etc will all cause trouble.  Look in SYLOGIN.COM too.
    
    Burns
    
 | 
| 1943.9 | I've found the problem | ADDIT::JOHN | Call me mad, but then who's sane ? | Thu Dec 21 1989 04:42 | 12 | 
|  |     
    Problem solved.
    
    It was a call to part of securpack in the sylogin that asked for input
    before the window was created, so causing it to give up.
    
    It is still a bit of a puzzle as to why this used to work on V5.2 but
    not on V5.3.
    
    Anyway, thanks for the help. Merry Christmas.
    
    John
 | 
| 1943.10 |  | DECWIN::FISHER | Burns Fisher 381-1466, ZKO3-4/W23 | Tue Dec 26 1989 11:39 | 4 | 
|  | I believe that the "interactiveness" of the DECwindows process changed between
5.2 and 5.3.  Maybe someone else can give more details, cuz I don't know them.
Burns
 |