| Date Of Receipt: 29-MAR-1994 19:16:22.36
From: FLAMBE::"[email protected]" "29-Mar-1994 1615"
To: "Joshua Friedman, OSF/UNIX SDE 381-1548" <[email protected]>
CC: [email protected], [email protected], [email protected]
Subj: Re: Workaround for as0main.p
Josh,
Precisely what I was going to suggest. I'm curious though, is this the only ".p"
file in your pool? I'm surprised we haven't seen this problem before.
The bad news is that the modified bci (soaking here, not yet available for sup)
behaves just as badly :-(
Mike
|
| Date Of Receipt: 30-MAR-1994 11:06:10.72
From: FLAMBE::"[email protected]" "30-Mar-1994 0804"
To: [email protected]
CC: [email protected], [email protected], [email protected]
Subj: Re: Workaround for as0main.p
I think the problem is in bci. Don, could you confirm how you provided the
checkin comments to bci (via the -m flag, via text after the HISTORY line in the
file, or both)?
If my hunch is correct, you should see that the checked-in file still has your
text after the HISTORY line, and this is what bsubmit is choking on. It should
be possible to remove that line (you'll need to make it writeable, edit it, then
make it readonly), and submit it.
If you've already submitted it by changing the leader to BIN, I'm still
interested in seeing whether that text is indeed there after the HISTORY line,
and if so setting the leader back to " ", checking it out, removing the line in
question, checking it in & submitting it.
Thanks ... Mike
|
| Date Of Receipt: 30-MAR-1994 19:25:46.92
From: FLAMBE::"[email protected]" "30-Mar-1994 1624"
To: "Joshua M. Friedman, OSF/UNIX SDE 381-1548" <[email protected]>
CC: [email protected], [email protected], [email protected]
Subj: Re: Workaround for as0main.p
Josh,
I haven't figured out exactly what yet, but it seems there was something wrong
with this particular file, before you changed the comment leader, and before Don
did his checkin & submit. Everything works fine with a "fresh" .p file.
Regarding the comment still being after "HISTORY" in the file, but without a
comment leader :
I think the fact that the comment leader is missing from this line
after HISTORY is irrelevant and that Don probably left it off when
inserting it in the file. With a copy of the same rcs file
(copied before its leader was changed to BIN), I can reproduce this
(with or without a leading blank, the line following HISTORY gets
ignored and left in the file after the checkin).
For files with non-blank comment leaders, a missing leader generates
the correct error message and lets you edit the file.
For a .p file (with a comment leader of " "), a missing leader is
harmless, and the checkin & submit proceed normally (except for this
one file, because that line is being ignored).
Regarding "BIN" prefixing the revision entry for 4.2.9.5 :
My guess is that when 'bcs -c' was run, it was given an argument of
"BIN" (ie with quotes), rather than just BIN (ie without quotes). That
would have the result of setting the comment leader literally to the
string BIN. Can you or Don confirm this? If that's the case, do :
bcs -c BIN as0main.p
You may have to edit the rcs file to change the "BIN" to " " in the
revision entry for 4.2.9.5.
If you would prefer the file to have a leader of " " right now, you could set it
to that, and just avoid putting text after the "HISTORY" line until I can figure
out what's wrong with the file. I've verified (with my copy that still has a
leader of " ") that the -m flag works.
Mike
|