Conference smurf::buildhelp

1909.0. "fwd: ODE 3.0 problems with bsubmit" by AOSG::FILTER (Automatic Posting Software - mail to flume::puck) Fri Oct 27 1995 15:31

Date Of Receipt: 	27-OCT-1995 14:34:00.09
From: 	SMURF::FLUME::jmf "Joshua M. Friedman OSF/UNIX SDE  27-Oct-1995 1431"
To: 	decwet::ode, tpb@DEC:.zko.flume
CC: 	odehelp@DEC:.zko.flume
Subj: 	fwd: ODE 3.0 problems with bsubmit

DECwest ODE team, can you please look at the first problem Tom had with
the visual/novisual/ok menu.

Tom, as for the -l link switch, sorry, this is my fault.  When ODE v3.0
was released this switch wasn't yet completely implemented.  It's since
been completed, but I haven't had an opportunity at a "quiet time" to
"sup" the bug fixes since this fix, since this will cause interruption 
to the user community (it will expire the authorization of all active
workon sessions since the workon image is being updated among others).

I'll try to get this done within a couple of days.

Thanks for the reminder!		-josh

P.S. In the future please mail to 'odehelp' instead of me.  thanks.

------- Forwarded Message

To: jmf
Subject: ODE 3.0 problems with bsubmit
Date: Fri, 27 Oct 95 11:18:24 -0400
From: "Dr. Tom Blinn, 603-881-0646" <tpb>
X-Mts: smtp

This morning, I did a bsubmit to Pt BL9 and it didn't go smoothly.

First, I tried this command:

bsubmit -all -l -auto_out -defect ptos-3465-tpb

Note that I did not explicitly say either "-visual" or "-novisual".

It accepted the command, got the list of files that needed merging, and then
put me into my editor (vi) in what seemed a "-novisual" mode to resolve the
merge conflicts, but when I edited the file and exited, I never managed to
get to an "ok" option -- it kept dropping me back into the editor.  I got
out of this by aborting the merge, cancelling the work so far (there was
only one line in conflict, the "pragma ident"), and restarting the merge.
I should NOT have had to restart -- if it was going to default to "novisual"
and drop me into vi (which was fine as far as I was concerned), it should
have picked up the merged file, figured out I'd removed the conflicts, and
let me proceed.

Then I tried this command:

bsubmit -all -l -auto_out -visual -defect ptos-3465-tpb 

This worked much better, running the visual merge tool (which didn't do as
good a job, as the only REAL conflicts in the merges were the pragma ident
lines but I had to figure that out for myself, the tool doesn't show me the
edit history or offer a convenient way to get the diffs from the common
ancestor to guide the merge activity).  When I completed the two merges to
resolve the "pragma ident" conflicts and get all the changes, I got these
messages:

Choose next merge action on ./kernel/arch/alpha/hal/eb64p.c
Abort, ok, edit, novisual, visual, [r]co, [r]diff, help  [ok]  
[ ./kernel/arch/alpha/hal/eb164.c checked in onto branch 1.1.6 ]
[ ./kernel/arch/alpha/hal/eb64p.c checked in onto branch 1.1.9 ]
[ ./kernel/arch/alpha/hal/eb66.c checked in onto branch 1.1.8 ]
[ ./kernel/arch/alpha/hal/eb164.c Rev 1.1.6.8 checked out ]
[ ./kernel/arch/alpha/hal/eb64p.c Rev 1.1.9.14 checked out ]
[ ./kernel/arch/alpha/hal/eb66.c Rev 1.1.8.16 checked out ]
[ ./kernel/arch/alpha/hal/eb164.c outdated revisions in set
Thomas_Blinn_ptbox ]
[ ./kernel/arch/alpha/hal/eb64p.c outdated revisions in set
Thomas_Blinn_ptbox ]
[ ./kernel/arch/alpha/hal/eb66.c outdated revisions in set
Thomas_Blinn_ptbox ]

Submission completed successfully.
>> FATAL ERROR in mklinks:
   No such option: -files.
USAGE:
mklinks [-copy [-over] -norecurse] [-dironly] [sb_opts]
        <directory> | [abs_opts]
        -copy: copy the files instead of linking them
          -over: overwrite file on copy even if it exists
        -norecurse: do files in the directory, not sub-directories
        directory: the directory to link in from the backing tree.
        abs_opts:
          -link_from <dir>: non-sandbox directory to link from
          -link_to <dir>: non-sandbox directory to link to
          -cmp: compare and link identical files
          -query: with -cmp flag, query for each file to be relinked
        sb_opts:
          -sb <sandbox>, -rc <user rc> -sb_rc <sb rc>
mklinks -usage | -rev

(the >> FATAL ERROR in mklinks happened three times, once for each file).

It didn't replace my outdated files with symbolic links:
csh[17]> pushd ./kernel/arch/alpha/hal
/sbdisk1/tpb/sandboxes/ptbox/src/kernel/arch/alpha/hal
/sbdisk1/tpb/sandboxes/ptbox/src 
csh[18]> ls -l eb*6*.c
lrwxr-xr-x   1 tpb      staff         74 Sep  8 15:35 eb164_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb164_mult.c
lrwxr-xr-x   1 tpb      staff         74 Sep  8 15:35 eb64p_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb64p_mult.c
lrwxr-xr-x   1 tpb      staff         73 Sep  8 15:35 eb66_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb66_mult.c
csh[19]> mklinks -norecurse .

Linking files:
  From: /sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/.
  To:   /sbdisk1/tpb/sandboxes/ptbox/src/kernel/arch/alpha/hal/.

csh[20]> ls -l eb*6*.c
lrwxr-xr-x   1 tpb      staff         69 Oct 27 11:17 eb164.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb164.c
lrwxr-xr-x   1 tpb      staff         74 Sep  8 15:35 eb164_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb164_mult.c
lrwxr-xr-x   1 tpb      staff         69 Oct 27 11:17 eb64p.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb64p.c
lrwxr-xr-x   1 tpb      staff         74 Sep  8 15:35 eb64p_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb64p_mult.c
lrwxr-xr-x   1 tpb      staff         68 Oct 27 11:17 eb66.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb66.c
lrwxr-xr-x   1 tpb      staff         73 Sep  8 15:35 eb66_mult.c -> 
/sbdisk1/tpb/sandboxes/ptbox/link/src/kernel/arch/alpha/hal/./eb66_mult.c
csh[21]> 

If the -l option isn't going to work, the reference page shouldn't list it,
and the tool shouldn't refuse to work in "-novisual" mode the way it did.

Tom



------- End of Forwarded Message

Date Of Receipt: 	27-OCT-1995 16:33:51.68
From: 	SMURF::ALPHA::tpb "Thomas Prescott Blinn USG  27-Oct-1995 1631"
To: 	"Joshua M. Friedman, OSF/UNIX SDE 381-1548" <jmf@DEC:.zko.alpha>
CC: 	decwet::ode, odehelp@DEC:.zko.alpha
Subj: 	Re: fwd: ODE 3.0 problems with bsubmit

Thanks for the prompt and courteous support.  I'll use odehelp in future.

Tom