[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
cvs edit and cvs co?
From: |
Colby Allred |
Subject: |
cvs edit and cvs co? |
Date: |
Mon, 4 Jun 2001 14:57:00 -0700 (PDT) |
We have run into some interesting behaviour with our repository that you
might want to look into. I think the most brief way to explain it is that
running a checkout removes editor information from the fileattr file!
Following is a list of steps that show the problem.
pluto@garfield:~/cvs_edit_test>uname -a
Linux garfield.us.aha.com 2.2.16-3 #1 Mon Jun 19 19:11:44 EDT 2000 i686 unknown
pluto@garfield:~/cvs_edit_test>cvs -v
Concurrent Versions System (CVS) 1.10.8 (client/server)
Copyright (c) 1989-1998 Brian Berliner, david d `zoo' zuhn,
Jeff Polk, and other authors
CVS may be copied only under the terms of the GNU General Public License,
a copy of which can be found with the CVS distribution kit.
Specify the --help option for further information about CVS
NOTE: the following behaviour also happens with cvs v1.11
pluto@garfield:~/cvs_edit_test>ls -al
total 4
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:12 ./
drwxr-xr-x 33 callred vlsi 2048 Jun 4 14:12 ../
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:12 CVS/
pluto@garfield:~/cvs_edit_test>echo fileA > fileA
pluto@garfield:~/cvs_edit_test>echo fileB > fileB
pluto@garfield:~/cvs_edit_test>cvs add fileA fileB
cvs add: scheduling file `fileA' for addition
cvs add: scheduling file `fileB' for addition
cvs add: use 'cvs commit' to add these files permanently
You have mail in /var/spool/mail/callred
pluto@garfield:~/cvs_edit_test>cvs ci
cvs commit: Examining .
RCS file: /nusers/callred/cvstest/cvs_edit_test/fileA,v
done
Checking in fileA;
/nusers/callred/cvstest/cvs_edit_test/fileA,v <-- fileA
initial revision: 1.1
done
RCS file: /nusers/callred/cvstest/cvs_edit_test/fileB,v
done
Checking in fileB;
/nusers/callred/cvstest/cvs_edit_test/fileB,v <-- fileB
initial revision: 1.1
done
pluto@garfield:~/cvs_edit_test>ls -al
total 6
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:12 ./
drwxr-xr-x 33 callred vlsi 2048 Jun 4 14:12 ../
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:12 CVS/
-rw-rw-r-- 1 callred vlsi 6 Jun 4 14:12 fileA
-rw-rw-r-- 1 callred vlsi 6 Jun 4 14:12 fileB
pluto@garfield:~/cvs_edit_test>cvs watch on file*
pluto@garfield:~/cvs_edit_test>rm file*
pluto@garfield:~/cvs_edit_test>cvs update
cvs update: Updating .
cvs update: warning: fileA was lost
U fileA
cvs update: warning: fileB was lost
U fileB
pluto@garfield:~/cvs_edit_test>ls -al
total 6
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:12 ./
drwxr-xr-x 33 callred vlsi 2048 Jun 4 14:12 ../
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:13 CVS/
-r--r--r-- 1 callred vlsi 6 Jun 4 14:13 fileA
-r--r--r-- 1 callred vlsi 6 Jun 4 14:13 fileB
Watching is on, and I cvs edit the two files.
pluto@garfield:~/cvs_edit_test>cvs watchers
pluto@garfield:~/cvs_edit_test>cvs edit fileA fileB
pluto@garfield:~/cvs_edit_test>cvs editors
fileA callred Mon Jun 4 21:13:37 2001 GMT garfield.us.aha.com
/nusers/callred/cvs_edit_test
fileB callred Mon Jun 4 21:13:37 2001 GMT garfield.us.aha.com
/nusers/callred/cvs_edit_test
Now, I check out another sandbox
pluto@garfield:~/cvs_edit_test>mkdir ../TEMPCVS
pluto@garfield:~/cvs_edit_test>cd ../TEMPCVS/
pluto@garfield:~/TEMPCVS>cvs -d /nusers/callred/cvstest co cvs_edit_test
cvs checkout: Updating cvs_edit_test
U cvs_edit_test/fileA
U cvs_edit_test/fileB
pluto@garfield:~/TEMPCVS>cd cvs_edit_test/
pluto@garfield:~/TEMPCVS/cvs_edit_test>ls -al
total 5
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:14 ./
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:14 ../
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:14 CVS/
-r--r--r-- 1 callred vlsi 6 Jun 4 14:12 fileA
-r--r--r-- 1 callred vlsi 6 Jun 4 14:12 fileB
Ok, so far so good...
pluto@garfield:~/TEMPCVS/cvs_edit_test>cvs editors
The information is gone from the fileattr file!
pluto@garfield:~/TEMPCVS/cvs_edit_test>cd ../../cvs_edit_test/
pluto@garfield:~/cvs_edit_test>ls -al
total 6
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:13 ./
drwxr-xr-x 34 callred vlsi 2048 Jun 4 14:12 ../
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:13 CVS/
-rw-r--r-- 1 callred vlsi 6 Jun 4 14:13 fileA
-rw-r--r-- 1 callred vlsi 6 Jun 4 14:13 fileB
pluto@garfield:~/cvs_edit_test>cvs editors
The files still have the write permissions set, as if the cvs edit is
still in effect, but cvs editors returns nothing.
pluto@garfield:~/cvs_edit_test>ll CVS
total 7
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:13 ./
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:13 ../
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:13 Base/
-rw-rw-r-- 1 callred vlsi 25 Jun 4 14:13 Baserev
-rw-rw-r-- 1 callred vlsi 78 Jun 4 14:13 Entries
-rw-rw-r-- 1 callred vlsi 14 Jun 4 14:12 Repository
-rw-rw-r-- 1 callred vlsi 24 Jun 4 14:12 Root
pluto@garfield:~/cvs_edit_test>ll CVS/Base
total 4
drwxrwxr-x 2 callred vlsi 512 Jun 4 14:13 ./
drwxrwxr-x 3 callred vlsi 512 Jun 4 14:13 ../
-rw-r--r-- 1 callred vlsi 6 Jun 4 14:13 fileA
-rw-r--r-- 1 callred vlsi 6 Jun 4 14:13 fileB
pluto@garfield:~/cvs_edit_test>cat CVS/Baserev
BfileA/1.1/
BfileB/1.1/
pluto@garfield:~/cvs_edit_test>
The admin directory still contains Base/ and Baserev (which is a result of
the cvs edit command, right?)
NOTE: I also tried this with me doing the initial checkout and cvs edit,
and another user doing a checkout after that, and the information was
still removed from the fileattr (i.e. cvs editors returned null)
Am I wrong in thinking that a checkout/update should leave previous editor
information in the fileattr file?
--
Colby Allred callred@aha.com
Advanced Hardware Architectures
http://www.aha.com/
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- cvs edit and cvs co?,
Colby Allred <=