|
Old CVS FAQ from Gigascale
Below is a copy of the old CVS faq from GSRC
CVS Information about Concurrent Version Control (CVS)
for CHESS webpage authors and developers.
- How do I get a CVS account
-
CVS accounts are available to anyone who has
a login on CHESS, TRUST or Embedded, and is working
on a project that is doing software development.
To request a CVS account, use the
CVS Account Request Form.
You should get email with your
account details within a day or so. (If not,
by all means send email to webmaster
asking about the status of your request!)
Note that CVS accounts cannot be created in response
to a simple email request. The form above
generates passwords and a script that is used
to create your account, saves a lot of time
and generally makes the whole process more
reliable. It helps us a lot if you use it :-)
- What do I do to get started using CVS on the CHESS website?
-
For information about getting started with CVS, see
How do I edit pages in a group with the "CVS Authoring" option?
- What CVS resources are available on the Net?
-
- CVS Clients for Windows and Mac
-
Under Windows, we recommend TortoiseCVS.
Under MacOS, we recommend smartcvs.
- What CVS Resources are available locally?
-
- Common CVS Commands
-
Below is some text from William Wu about CVS commands
-
cvs import [module|file]
- add files/modules into the repository directory.
cvs commit [module|file]
- Check in your modified files/module to the repository,
creates a new file that merge your file and the one in
repository. Note it is a MERGE not a replace, so it could
cause problem. A good rule is to after you modified the
file, do a cvs update (see below) first, so you have the
newest version, and try to recompile and run, so that you
can find any conflict that arise.
cvs update [module|file]
- updates your files to the files in the repository,
if you have made some change to your files, it will
merge the two. This is always a good thing to do before
you do commit, since someone else can modify the file
and check in after you last update, thus it is possible
the changes he made is in conflict with you changes.
By doing a update, you can find out about the changes he made
and spot any conflict. (conflict is indicated by a letter c
in front of file name, in the messages from CVS)
You can run cvs -n update to preview what
would happen if you ran cvs update. The -n
option does not change any files, it only issues reports.
Note that usually you will want to use
cvs update -P -d, which Prunes empty
directories and creates new directories. You can
set this as the default by creating a
~/.cvsrc file that contains the line
update -P -d
If you run cvs update without the -d flag,
then you might not get any new directories created
cvs checkout [module|file]
- Use this when you check out the files/module for the first time.
When someone had import a new file in, you can use this command
to check it out, or I think do a update on the module will also
add the new file to you directory.
cvs status [module|file]
- Give a status report on the files comparing with the newest
version in the repository. The files status can be Up-to-Date,
Modified ... Uses this command before commit, then you can
check if you need to do a update or not.
cvs diff [module|file]
- does a diff with your file and its newest version in repository.
cvs history [module|file]
- gives a report on the history of modication done on this file,
users, date etc.
These are some commonly used ones, you can always use
cvs(1)
man page for help.
- Common CVS Tasks
-
Checking out just one directory
The old symbolic links method we used with SCCS probably won't work
as well. Your best bet is to just check out the entire tree
To check out just one directory outside of your tree, you can use
the directory name with cvs checkout
cvs checkout ptII/ptolemy/plot
will create the directory ptII/ptolemy/plot
in the current directory.
.cvsrc file
Brian Evans suggested that if your ~/.cvsrc file contains
the following:
cvs -z3
update -d -P
Then you will get the following features automatically:
-
cvs -z3
- compression when doing a
cvs checkout. This
can be especially useful when you are dialed in from home
-
update -d -P
- When you do
cvs update, create and prune directories
as necessary.
The alternative is to use the command line arguments each time you
run the appropriate commands.
cvs -d :ext:gigasource.eecs.berkeley.edu:/home/cvs co ptII
cd ptII
cvs update -d -P
cvs update -D 7/14/2000 -d -P
setenv PTII `pwd`
make
If you are under bash, then instead of setenv PTII `pwd`, do
PTII=`pwd`
export PTII
Notes:
After doing the cvs co with the -d :ext:gigasource.eecs.berkeley.edu:/home/cvs
flag, you need not specify the -d flag again. That value is cached
in the CVS subdirectory in each directory
In 'cvs update -d -P', the -d flag means something different than
the -d :ext:gigasource.eecs.berkeley.edu:/home/cvs flag
Here, -d means create directories
-P means prune empty directories.
- How do I use CVS with Eclipse?
-
Using CVS with Eclipse can be tricky.
The
Eclipse FAQ has some info.
Eclipse 2.0 Features page might be of use is of use.
The Ptolemy II Eclipse Instructions has some hints.
- Administrative CVS Commands
-
Below are some common scenarios for fixing up the CVS database.
See the FAQ for more examples.
Ignoring files
/users/cvs/Repository/CVSROOT/cvsignore contains
the patterns for files that are ignored when cvs update is done.
If cvsignore includes *.class,
then when you do cvs update, you won't see messages like
? DEActor.class
? DEDirector.class
Adding a directory
- Create the directory
mkdir foo
- Update the Repository and create
foo/CVS
cvs add foo
- Populate
foo with your files
- Add the files:
cd foo
cvs add file1 file2
cvs commit
Deleting a file
- (optional)
Make sure that you do not have any uncommitted changes to the file:
cvs update foo.c
- Tell CVS you want to remove the file:
cvs delete -f foo.c
The -f option forces cvs to delete the file
from the current directory.
- Commit the changes, move the file to the CVS attic directory
cvs commit foo.c
This is covered in Chapter 9 of the CVS documentation.
If a file gets added without -kb, then you can fix it with
-
cvs admin -kb foo.class
-
cvs update -A foo.class
- Then copy in a good copy of the file from outside CVS.
-
cvs commit -m "Make it binary" foo.class
cvs remove -f foo.c will remove the file for you.
The faq says:
4. How do I rename a file?
CVS does not offer a way to rename a file in a way that CVS can track
later. See Section 4B for more information.
Here is the best (to some, the only acceptable) way to get the effect
of renaming, while preserving the change log:
Copy the RCS (",v") file directly in the Repository.
cp $CVSROOT//,v $CVSROOT//,v
By duplicating the file, you will preserve the change history and the
ability to retrieve earlier revisions of the old file via the "-r
" or "-D " options to "checkout" and "update".
Remove the old file using CVS.
cd / rm
cvs remove
cvs commit
This will move the to the Attic associated with .
Retrieve and remove all the Tags from it.
By stripping off all the old Tags, "checkout -r" and "update -r" won't
retrieve revisions Tagged before the renaming.
cd /
cvs update
cvs log # Save the list of Tags
cvs tag -d
cvs tag -d
. . .
This technique can be used to rename files within one directory or
across different directories. You can apply this idea to directories
too, as long as you apply the above to each file and don't delete the
old directory.
Of course, you have to change your build system (e.g. Makefile) in
your to know about the name change.
Warning: Stripping the old tags from the copied file will allow "-r
" to do the right thing, but you will still have problems with
"-D " because there is no place to store the "deletion time".
See 5B.3 for more details.
Last modified: _6/13/1997_
- EDITOR Variable
-
When CVS commits a change, it needs a log message.
One quick workaround is to use the
-m "log message"
option to commit:
cvs commit -m "Fixed off by one error" foo.java
If you do not use the -m option, then
CVS will need a log message.
When CVS needs a log message, it reads the EDITOR
environment variable to determine which editor to startup.
If you set EDITOR to emacs, then
each time CVS needs a log message it will start up an emacs.
If you always have an emacs process running.
Consider using emacsclient instead.
- Edit your
~/.emacs file and add
(server-start)
to it
- Edit your
~/.cshrc file and add
setenv EDITOR emacsclient
Now, when you CVS needs a log message, it will bring up a window
in your emacs process. When you are done editing,
type C-x #
For more information about emacsclient, start
up emacs and type M-x info and search for
the emacs online documentation.
There is an easier solution:
If you use emacs as your editor, then you can use
C-x C-q to commit your change.
After filling in the message, type C-c C-c to exit the log message editor window.
For C-x C-q to
work with CVS from a laptop you must have your laptop
set up so that it does not require a password when
committing a cvs change.
See How do I use CVS without typing my password each time? for details.
- How do I use CVS without typing my password each time?
-
To use CVS without typing your password, you will need
an individual account on gigasource.
If your CVS account was created after
September 2000, then it is fairly likely that you
already have an individual account.
If you would like to check, log in to your cvs
account on gigasource and run pwd.
For example, the ptolemy users does not have
an individual account, so email should be sent:
cxh@maury 1% ssh gigasource -l ptolemy
ptolemy@gigasource's password:
Last login: Mon Mar 26 22:01:19 2001 from julian.eecs.ber
$ pwd
/home/cvsusers
$ exit
Connection to gigasource closed.
cxh@maury 2%
The steps below are slightly modified from the
CVS SSH instructions
SSH (and therefore CVS) can use RSA and Rhosts style authentication to
make it so that you can login without typing your password. Using Rhosts authentication alone is insecure,
and most CVS servers (this one included) disallow it.
RSA can be used with or without Rhosts authentication.
Using RSA alone, any user with the appropriate RSA
key and passphrase can access the repository.
Using RSA and Rhosts restricts repository access to only computers listed in Rhosts that also have the appropriate RSA key.
Note that for RSA with Rhosts authentication
to work, each host that
you are logging in from needs to be listed in two files.
If you are connecting from multiple hosts via dynamic
DSL, then each time you connect, you are likely to
have a different address, which makes managing
the file difficult. There are several possible
solutions, one is to try to use wildcards
in ~/.ssh/known_hosts and ~/.shosts,
the other is to run a script that updates these
files automatically. Both solutions are complex
and have security issues, so we do not cover them here.
It is much simpler to use RSA authentication alone
in this case.
RSA authentication
Once your account has been set up, do the following:
- Create
~/.ssh/id_rsa.pub on the
local machine:
If ~/.ssh/id_rsa.pub does not exist,
on your local machine,
then create it by running ssh-keygen -t rsa
When prompted for a passphrase, hit return.
If you type in a passphrase here, you will be prompted
for that passphrase each time.
Running ssh-keygen will generate the files
~/.ssh/id_rsa and
~/.ssh/id_rsa.pub, which
are your private and public encryption keys respectively.
- Create
~/.ssh/authorized_keys2 on
gigasource:
Set the permission of ~/.ssh/id_rsa.pub
to 0644 and then
copy ~/.ssh/id_rsa.pub over
to ~/.ssh/authorized_keys2 with the
scp -p flag to preserve permissions.
chmod 0644 ~/.ssh/id_rsa.pub
scp -p ~/.ssh/id_rsa.pub yourgigasourcelogin@gigasource:~/.ssh/authorized_keys2
where yourgigasourcelogin is the your CVS login
on gigasource.
- From the local machine, test ssh with:
ssh yourgigasourcelogin@gigasource cvs
to check the set up.
Below is a sample run
cxh@DOPLAP03 ~
$ ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/cygdrive/c/cxh/.ssh/id_rsa):
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /cygdrive/c/cxh/.ssh/id_rsa.
Your public key has been saved in /cygdrive/c/cxh/.ssh/id_rsa.pub.
The key fingerprint is:
03:2a:8a:3b:96:93:6b:74:86:c8:ea:30:e2:c9:11:68 cxh@DOPLAP03
cxh@DOPLAP03 ~
$ chmod 0644 ~/.ssh/id_rsa.pub
cxh@DOPLAP03 ~
$ scp ~/.ssh/id_rsa.pub cxh@gigasource:~/.ssh/authorized_keys2
The authenticity of host 'gigasource (128.32.171.225)' can't be established.
RSA key fingerprint is 74:57:84:9b:ca:b8:44:1d:fa:f0:e3:27:29:ac:19:c6.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'gigasource,128.32.171.225' (RSA) to the list of know
n hosts.
cxh@gigasource's password:
id_rsa.pub 100% |*****************************| 222 00:00
cxh@DOPLAP03 ~
$ ssh cxh@gigasource cvs
Usage: cvs [cvs-options] command [command-options-and-arguments]
where cvs-options are -q, -n, etc.
(specify --help-options for a list of options)
where command is add, admin, etc.
(specify --help-commands for a list of commands
or --help-synonyms for a list of command synonyms)
where command-options-and-arguments depend on the specific command
(specify -H followed by a command name for command-specific help)
Specify --help to receive this message
The Concurrent Versions System (CVS) is a tool for version control.
For CVS updates and additional information, see
the CVS home page at http://www.cvshome.org/ or
Pascal Molli's CVS site at http://www.loria.fr/~molli/cvs-index.html
cxh@DOPLAP03 ~
$
The above steps should not prompt you for your password each time you run a cvs command. There are two ways to enable yourself to not type a passphrase each time.
The first is to use only RSA authentication as described above, and give an empty passphrase.
Note that this means that if someone steals your laptop and breaks into your account, they will be able to use the ssh command to get on to your Unix account.
RSA authentication works with ssh1 and ss2.
The second way is to set up Rhosts RSA authentication,
which currently only works with ssh1.
In this case, ssh will authenticate your computer instead of you. It is somewhat safer for your computer to log in without a password
because the CVS server can determine where your computer
is located. Your computer can only login without
a password if it also has the correct name and IP.
This method is shown below.
Rhosts RSA authentication
These instructions apply to ssh1 clients, like
the version of ssh that is currently shipped with cvsssh.
For ssh2, see above.
- Setup
~/.shosts on gigasource:
Create a temporary file on your local machine that
contains the name of the host you will be logging in
from.
echo "myhost.eecs.berkeley.edu" > /tmp/shosts
- Change the permissions of the file on you local
machine so that it is only readable by you. Note
that you can't run
chmod on the remote
machine under rksh, so you need to fix the permissions
before you copy the file over.
chmod 0600 /tmp/shosts
- Copy the file over:
scp /tmp/shosts gigasource:~/.shosts
Below is a sample session
ptolemy@myhost 4% echo "myhosts.eecs.berkeley.edu" > /tmp/shosts
ptolemy@myhost 5% chmod 0600 /tmp/shosts
ptolemy@myhost 6% scp /tmp/shosts gigasource:~/.shosts
Enter passphrase for RSA key 'ptolemy@myhost.eecs.berkeley.edu': your passphrase
^Mshosts | 0 KB | 0.0 kB/s | ETA: 00:00:00 | 100%
ptolemy@myhost 7%
- On your Windows machine, check to see if the file
c:\ssh\etc\ssh_host_key.pub exists:
- If it does not exist, generate host keys on your Unix machine
and copy them over. Note that the ssh-keygen command
that is shipped with the Windows CVS SSH package will
not work, you should run
ssh-keygen
on a Unix host. Note that the -N below indicates
that your computer will not have a passphrase.
cd /tmp
ssh-keygen -b 1024 -f ssh_host_key -N ''
On Windows, you can copy the files with scp. If you use scp here,
note that scp does not understand the Windows c: naming convention,
it think c: is a machine named 'c'. To copy the file under bash
on Windows:
cd c:/ssh/etc
scp yourunixmachine:/tmp/ssh_host_key.pub .
scp yourunixmachine:/tmp/ssh_host_key .
On the Unix machine, remove /tmp/ssh_host_key
- If c:/etc/ssh/ssh_host_key.pub exists on your Windows machine,
copy it to a temporary file on the Unix machine:
cd c:/ssh/etc
scp ssh_host_key.pub yourunixmachine:/tmp
- Set up
~/.ssh/known_hosts on gigasource:
On gigasource, ~/.ssh/known_hosts lists hosts
that are allowed to connect without a password. Since
we can't edit files on gigasource from the restricted shell,
we create the file on the local machine and copy it
over to gigascale.
Note that if you are connecting from multiple machines,
you will need to add a line to this file for each
machine you are connecting from. The easiest way
to do this is to use scp to transfer the file
to a local machine, edit the file locally and then
scp it back to gigasource.
- Grab the contents of the copy of ssh_host_key.pub that
you either generated or copied over, and create a temporary
file
- Add the fully qualified domain name of the windows machine
to the beginning of the line.
You should end up with something like
maury.eecs.berkeley.edu 1024 17 27348124368712489214987214872164987243
Where the last number is several lines long.
- Copy the file:
scp /tmp/known_hosts gigasource:~/.ssh/known_hosts
- From the local machine, test ssh with:
ssh -v gigasource cvs
you should not have to type in your password.
Troubleshooting CVS SSH under NT
- Run
ssh -v yourunixmachine cvs
and check the output, which should look like:
ptolemy@carson 8% ssh -v gigasource cvs
SSH Version 1.2.26 [sparc-sun-solaris2.5.1], protocol version 1.5.
Standard version. Does not use RSAREF.
carson.eecs.berkeley.edu: Reading configuration data /etc/ssh_config
carson.eecs.berkeley.edu: ssh_connect: getuid 4035 geteuid 0 anon 0
carson.eecs.berkeley.edu: Connecting to gigasource [128.32.171.225] port 22.
carson.eecs.berkeley.edu: Allocated local port 1021.
carson.eecs.berkeley.edu: Connection established.
carson.eecs.berkeley.edu: Remote protocol version 1.5, remote software version \
1.2.26
carson.eecs.berkeley.edu: Waiting for server public key.
carson.eecs.berkeley.edu: Received server public key (768 bits) and host key (1\
024 bits).
carson.eecs.berkeley.edu: Host 'gigasource' is known and matches the host key.
carson.eecs.berkeley.edu: Initializing random; seed file /users/ptolemy/.ssh/ra\
ndom_seed
carson.eecs.berkeley.edu: Encryption type: idea
carson.eecs.berkeley.edu: Sent encrypted session key.
carson.eecs.berkeley.edu: Installing crc compensation attack detector.
carson.eecs.berkeley.edu: Received encrypted confirmation.
carson.eecs.berkeley.edu: Trying rhosts or /etc/hosts.equiv with RSA host authe\
ntication.
carson.eecs.berkeley.edu: Remote: Accepted by .shosts.
carson.eecs.berkeley.edu: Remote: Your host key cannot be verified: unknown or \
invalid host key.
carson.eecs.berkeley.edu: Remote: The host name used to check the key was 'cars\
on.eecs.berkeley.edu'.
carson.eecs.berkeley.edu: Remote: Try logging back from the server machine with\
the canonical host name using ssh, and then try again.
carson.eecs.berkeley.edu: Server refused our rhosts authentication or host key.
carson.eecs.berkeley.edu: No agent.
carson.eecs.berkeley.edu: Trying RSA authentication with key 'ptolemy@maury.eec\
s.berkeley.edu'
carson.eecs.berkeley.edu: Received RSA challenge from server.
carson.eecs.berkeley.edu: Bad passphrase supplied for key file /users/ptolemy/.\
ssh/identity.
Enter passphrase for RSA key 'ptolemy@maury.eecs.berkeley.edu': your CVS passphrase
carson.eecs.berkeley.edu: Sending response to host key RSA challenge.
carson.eecs.berkeley.edu: Remote: RSA authentication accepted.
carson.eecs.berkeley.edu: RSA authentication accepted by server.
carson.eecs.berkeley.edu: Requesting X11 forwarding with authentication spoofin\
g.
carson.eecs.berkeley.edu: Sending command: cvs
carson.eecs.berkeley.edu: Entering interactive session.
Usage: cvs [cvs-options] command [command-options-and-arguments]
where cvs-options are -q, -n, etc.
(specify --help-options for a list of options)
where command is add, admin, etc.
(specify --help-commands for a list of commands
or --help-synonyms for a list of command synonyms)
where command-options-and-arguments depend on the specific command
(specify -H followed by a command name for command-specific help)
Specify --help to receive this message
The Concurrent Versions System (CVS) is a tool for version control.
For CVS updates and additional information, see
Cyclic Software at http://www.cyclic.com/ or
Pascal Molli's CVS site at http://www.loria.fr/~molli/cvs-index.html
carson.eecs.berkeley.edu: Transferred: stdin 0, stdout 716, stderr 0 bytes in 0\
.1 seconds
carson.eecs.berkeley.edu: Bytes per second: stdin 0.0, stdout 12490.8, stderr 0\
.0
carson.eecs.berkeley.edu: Exit status 1
ptolemy@carson 9%
- Check the value of the CVSROOT environment variable. It should
be something like:
:ext:myhost.eecs.berkeley.edu:/users/cvs/Repository
- Be sure that you have created a key on the Unix side and
copied the identity* files from Unix to NT
- Be sure that $HOME is set for your NT account
- Try using the ssh.exe binary that the $CVS_RSH variable refers to.
Below is an example where we rsh over to carson and get the date:
bash-2.02$ echo $CVS_RSH
D:\Program Files\Ptolemy\CVS SSH\ssh.exe
bash-2.02$ /Program\ Files/Ptolemy/CVS\ SSH/ssh carson date
Enter passphrase for RSA key 'cxh@myhost.eecs.berkeley.edu':
ld.so.1: /usr/local/bin/xauth: warning: /usr/4lib/libXmu.so.4.0: has older revision than expected 10
Thu Feb 4 15:35:06 PST 1999
Below is an example that failed because of an incorrect CVS password, note that the password is prompted
for twice:
bash-2.02$ /Program\ Files/Ptolemy/CVS\ SSH/ssh carson date
Enter passphrase for RSA key 'cxh@myhost.eecs.berkeley.edu':
Bad passphrase.
Password:
Permission denied.
bash-2.02$
- Verify that you can use ssh to connect between two Unix boxes.
- Check ssh_config. RSAAuthentication and/or RhostsRSAAuthentication must be turned on. One way to do this locally is by having an /Program\ Files/Ptolemy/CVS\ SSH/ssh_config file that looks like:
Host *
RSAAuthentication yes
RhostsRSAAuthentication yes
- Reboot NT.
- Using CVS from a laptop
-
Using a laptop with CVS
If you are using CVS from a laptop, and you are
not always connected to the net, you may want to
create a shell script that contains a list of
the cvs commands that you want to execute when
you are connected.
For example, if the file doit.sh
contained:
cvs commit -m "Fixed spelling error" $PTII/README
cvs commit -m "Updated version number " $PTII/makefile
then when you connected again, you could run
the script with /bin/sh doit.sh.
One caveat is that this system works best if
you set up ssh to use RSA-style authentication, so
that it does not prompt for a password each time.
See the How do I use CVS without typing my password each time?
question for details.
- How do I change my CVS account password?
-
The cvs account you were given can do only two thing:
run the cvs command and run the passwd command.
To change the password, use ssh to connect to gigasource.eecs.berkeley.edu
and then run passwd.
cxh@maury 139% ssh -l ptolemy gigasource
ptolemy@gigasource's password:
No mail.
$ passwd
passwd: Changing password for ptolemy
Enter login password:
New password:
Re-enter new password:
passwd (SYSTEM): passwd successfully changed for ptolemy
$ exit
Connection to gigasource closed.
cxh@maury 140%
Note that your CVS account is separate from your CHESS
website account. Not all CHESS members will necessarily
have a CVS account.
Having your password changed for you
If you have forgotten your CVS password, then send
email to webmaster
and state your name and the name of your CVS account
in case it is different from the name of the account
you are sending email from
The CHESS staff will do the following:
- Determine the proper account name by searching
/etc/passwd with grep login /etc/passwd
- As root on gigasource, run
passwd login and change the password
- email the new password to the user, prompting
them to change it
- How do I edit pages in a group with the "CVS Authoring" option?
-
The answer is covered in
the
gsrc Rocket-Science FAQ.
- CVS thinks I'm Administrator, not myself
-
If, under Windows, when you run
cvs update
you are
prompted for the Administrator password on the CVS
repository, or you have permission problems writing to
files, then you may need to set up /etc/passwd
on machine to include your name and UID.
Location of /etc/passwd
Note that the actual location of /etc/passwd
varies depending on where you installed Cygwin.
If Cygwin is installed in c:/cygwin, then
/etc/passwd will actually end up in
c:/cygwin/etc/passwd.
mkpasswd
The easiest way to update /etc/passwd
is to run the mkpasswd command that
is included with the Cygwin toolkit.
Usually, when Cygwin is installed, mkpasswd
is run and /etc/passwd is automatically
created. If your account was added after Cygwin was
installed, then you may need to run
mkpasswd by hand.
mkpasswd -help will print out
the help for mkpasswd.
Below are a few scenarios for updating
/etc/passwd:
Your machine is in a windows domain
If your Windows machine is a member of a Windows
domain, use this command to add yourself to /etc/passwd:
mkpasswd -d | grep yourlogin > /etc/passwd
Note that if your domain is rather large, then
mkpasswd -d could take a minute to complete.
Your machine is not in a Windows domain
To generate /etc/passwd for all the local
Windows accounts, try:
mkpasswd -l > /etc/passwd
Your home directory
/etc/passwd also defines where your home
directory is. Usually, Cygwin will set
your initial default home directory to
/home/yourname, which maps
to c:/cygwin/home/yourname.
Unfortunately, tools like Java do not know about
the Cygwin directory mapping, so you may want to
reset your home directory to a location outside of
the Cygwin tree.
For example, if your home directory was at
c:/users/yourname, then Cygwin
would find this as /cygdrive/c/users/yourname because
Cygwin mounts the c: drive as
/cygdrive/c
To set up Cygwin so that it finds your home directory,
edit c:/cygwin/etc/passwd and change
your home directory from /home/yourname
to /cygdrive/c/users/yourname
Set up /etc/passwd by hand
Stephen Neuendorffer suggests
Edit /etc/passwd and add a new line that contains
your information.
username::userid:groupid:fullName:home directory:/bin/sh
Your username and fullname should be obvious.
If you forget these, then CVS probably won't be much use
to you! Your home directory is probably
/home/username. Your groupid should probably be
the same as the administrator and your userid should be
some unique number that is not used elsewhere in the file
and less than 65535. Adding one to the
largest number in the file will probably work.
Another suggestion for obtaining a userid is to look
at your entry in /etc/passwd on a Unix machine
and use that number.
mkgroup
There is also a mkgroup command
that is similar to mkpasswd.
mkgroup can be used to create
/etc/group:
mkgroup -l > /etc/group
Other resources
The CVS Account Name FAQ covers how
to use a CVS repository with a login that is different
than your current login. Note that even if you configure
your system to properly use a different login on the cvs
server, then you may still have permission problems if
you do not have /etc/passwd set up
properly.
Redhat.com Cygwin mkpasswd page
- Why is my CVS account name different from my website account name?
-
CVS accounts are just a restricted form of UNIX
account. Because of this, when you get a CVS user
account, you also get registered as an account-holder
in the EECS departmental database. If you already
have an EECS account, then you will get the account
name that the Department already has you registered
as.
However, if you do not already have a record
in the departmental database, we create a new record
for you. Sometimes, your website username is
already taken, in which case we have to give you a different
one.
If your login name and your CVS account name
are not the same, then you can try one of the methods below
yourname@gigasource
When using the -d cvs option, try
using yourname@gigasource.eecs.berkeley.edu
cxh@maury 29% cvs -d :ext:ptolemy@gigasource.eecs.berkeley.edu:/home/cvs co ptII
ptolemy@gigasource.eecs.berkeley.edu's password: the passwd
cvs server: Updating ptII
ssh_config
You can modify the
ssh_config file so that when you
connect with ssh, you use your CVS account name.
Under Unix, the file to modify is
~/.ssh/ssh_config, under Windows
try
c:sshetcssh_config.
The format of the file is the same under
Unix and Windows, the thing to do is to add a
User your cvs account name:
Host *
ConnectionAttempts 100
RhostsRSAAuthentication yes
User cxh
Note further that to use ssh as
a different use, use the -l username
option, and to use scp, use
-o user=username
- Under Cygwin, I get 'could not create directory /home/myname/.ssh'
-
If, under Windows when running cvs via Cygwin you
get a message like:
Could not create directory '/home/yourname/.ssh'.
The authenticity of host 'gigascale.eecs.berkeley.edu 128.32.171.xxx)' can't be established.
RSA key fingerprint is xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx.
Are you sure you want to continue connecting (yes/no)? yes
Failed to add the host to the list of known hosts (/home/yourname/.ssh/known_hosts).
yourname@maury.eecs.berkeley.edu's password:
then the problem is that Cygwin has the wrong
idea about where your home directory is.
See CVS thinks I'm Administrator, not myself for how to
fix this.
- Cygwin CR/NL problems
-
When Cygwin is installed, you should be sure to select
DOS for the Default Text File Type.
If you do not, then when you check in files, there
will be problems with the DOS CR/NL line endings.
In particular, what will happen is that when you
check in a file and then someone checks it out, it
will have extra \r or ^M characters.
The way to tell if this is a problem is to run
the mount command under Cygwin bash and
note whether the c: drive is mounted
textmode or not. If the c:
drive is mounted textmode, then
your setup is ok:
$ mount
c:\cygwin\bin on /usr/bin type system (textmode)
c:\cygwin\lib on /usr/lib type system (textmode)
c:\cygwin on / type system (textmode)
c: on /cygdrive/c type user (textmode,noumount)
However, if c: is mounted binmode,
then you have problems:
$ mount
c:\cygwin2\bin on /usr/bin type system (binmode)
c:\cygwin2\lib on /usr/lib type system (binmode)
c:\cygwin2 on / type system (binmode)
c: on /cygdrive/c type user (binmode,noumount)
The fastest solution is to reinstall Cygwin,
see How do I edit pages in a group with the "CVS Authoring" option?
and properly select
DOS for the Default Text File Type.
Another possible alternative is to use these mount
commands:
We have tried these commands but they have not
quite worked for us
mount -f -t -s c:\\cygwin\\lib /usr/lib
mount -f -t -s c:\\cygwin\\bin /usr/bin
mount -f -t -s c:\\cygwin /
mount -f -t c: /cygdrive/c
If you have a d: drive,you may need to do:
mount -f -t d: /cygdrive/c
When you are done, run mount, which should
show all the mounts as textmode.
The reason that this is a problem with CVS is that
CVS handles end of line translation between Unix and Windows.
If, under Cygwin, the directories are mounted
binmode, then Cygwin CVS will not do
the proper translations.
For details, see
How is the DOS/Unix CR/LF thing handled? in the Cygwin FAQ
See also
http://cygwin.com/cygwin-ug-net/using-textbinary.html
To replicate the problem, install Cygwin with Unix line
endings, check out a file using CVS, edit the file using
wordpad and check in your changes. When you check out
the file under Unix, it will have \r\n
line endings.
Note that there is some contention about whether
DOS or Unix line endings are correct. The TinyOS community
has required Cygwin to be set up with Unix line endings,
which works great if your users are technically
sophisticated. However, if your windows users do not have
much exposure to Unix, then configuring Cygwin to use
DOS line endings is more correct.
- Who can get a CVS account?
-
Anyone that has a reason to :-)
Seriously, the only requirements are that
you have an account on the CHESS website and are
working on a software development project that
is located on the CHESS CVS server.
There are two types of cvs access:
If you are or intend to be working on a software
project that doesn't yet have a CVS repository
on the CHESS server, then by all means
request a login (if you don't already
have one), and then request
a group so that your project can start using the
CHESS infrastructure. All that we require is that at
least one person actively working on the project
has a bona-fide connection to the CHESS in some way.
- How does anonymous password authentication work?
-
Some of the repositories have anonymous cvs access, which
allows anonymous users read only access
The ptII repository moved to svn in 2008. The cvs
commands below are an example of how anonymous
authentication can work
The Ptolemy II External Developers workgroup (ptexernal)
was set up to allow for anonymous read only access to
the Ptolemy II tree. After getting the password,
the commands to run were:
cd c:/directoryofyourchoice
cvs -d pserver:anon@source.eecs.berkeley.edu:/home/cvs/cvsanon login
# No password needed, hit enter
cvs -d pserver:anon@source.eecs.berkeley.edu:/home/cvs/cvsanon co ptII
The old instructions used a password from
http://chess.eecs.berkeley.edu/ptexternal/private/2.html.
cvs -d :pserver:ptdeveloper@gigasource.eecs.berkeley.edu:/home/cvs/cvsanon login
cvs -d :pserver:ptdeveloper@gigasource.eecs.berkeley.edu:/home/cvs/cvsanon co ptII
However, that repository was moved in July, 2005.
The best reference is
CVS Book: Password Authenticating Server Section
Troubleshooting
Anonymous read only cvs access uses the pserver facility
which is found on port 2401. If you are behind
a firewill, you may have problem connecting to port 2401
on gigasource. The way to test this is to use
telnet gigasource.eecs.berkeley.edu 2401
to try to connect to the pserver port, and then type
help:
cxh@maury 21% telnet gigasource.eecs.berkeley.edu 2401
Trying 128.32.171.225...
Connected to gigasource.EECS.Berkeley.EDU.
Escape character is '^]'.
help
cvs [pserver aborted]: bad auth protocol start: help^
Connection closed by foreign host.
cxh@maury 22%
If you are running under Windows, then you may
find that the Cygwin telnet binary gives better
diagnostics.
Also, the CVSROOT/config file should be set up to
look like
# Set this to "no" if pserver shouldn't check system users/passwords
SystemAuth=yes
Administrivia
Setting up anonymous read only cvs access
As the administrator of a group,
send email to webmaster.
requesting that anonymous read only cvs access be
setup for your cvs repository.
The website maintainers will:
- Create a link for your repository in
/home/cvs/cvsanon
cd /home/cvs/cvsanon
ln -s ../yourrepository .
- Each workgroup should have their own
anonymous cvs account so that password
administration is easier.
The accounts are listed in
/home/cvs/cvsanon/CVSROOT/passwd
and look like
accountname:encrypted password:cvsanon
cvsanon is the name of the account
that we actually log in as.
- To generate an encrypted password, run
cryptout.pl with the password
gigasource:root: %C2> /usr/local/bin/cryptout.pl foobar
oPG6N3As668O2
gigasource:root: %C2>
and then placing that encrypted password in
a new password entry in
/home/cvs/cvsanon/CVSROOT/passwd
Other details
Note that anonymous read only cvs access sends passwords
over in clear text with a very simple encryption scheme.
Thus, this method is really only useful for shared accounts
since with individual accounts users tend to use the same password
for many different accounts so if their cvs password gets
snooped, other accounts could be compromised.
Note further that the lightly encrypted password
is stored in ~/.cvspass.
Readonly access is controlled by adding users
to /home/cvs/cvsanon/CVSROOT/readers
/etc/services and /etc/inetd.conf
were modified with the following:
gigasource:root: %C2> grep pserver /etc/services
cvspserver 2401/tcp
gigasource:root: %C2> grep pserver /etc/inetd.conf
cvspserver stream tcp nowait root /usr/local/bin/cvs cvs --allow-root=/home/cvs
/cvsanon pserver
gigasource:root: %C2>
xinetd setup (for Linux)
To set up xinetd, I created /etc/xinetd.d/cvspserver:
service cvspserver
{
socket_type = stream
protocol = tcp
wait = no
user = root
passenv = PATH
server = /usr/bin/env
server_args = -i cvs -f --allow-root=/usr2/cvsroot --allow-root
=/home/real/cvsroot pserver
}
This also worked:
service cvspserver
{
socket_type = stream
protocol = tcp
wait = no
user = root
passenv = PATH
server = /usr/bin/cvs
server_args = -f --allow-root=/usr2/cvsroot --allow-root=/home/
real/cvsroot pserver
}
Then I did /etc/init.d/xinetd restart
Solaris 10 Notes
Under Solaris 10, I set up /etc/inetd.conf
as before, but then ran
inetconv -e to convert the cvspserver entry.
Add the following rules to /etc/ipf/ipf.conf
pass in quick on bge0 proto tcp from any to 128.32.48.234 port = cvspserver flags S keep state group 100
# CVS
pass out quick on bge0 proto tcp from 128.32.48.234 to any port = cvspserver flags S keep state group 200
Run ipf -F a -f /etc/ipf/ipf.conf
- How do I convert from SCCS or RCS to CVS
-
Resources:
Converting from SCCS to CVS is not that difficult, there
are scripts in the
contrib directory that
is shipped with the cvs source code. The basic idea
is that one backs up the SCCS tree, converts to RCS and then
converts to CVS.
The notes below discuss the conversion of the Ptolemy II SCCS tree
to CVS.
Converting from RCS to CVS
Once you've converted from SCCS to RCS, you need to convert
from RCS to CVS.
The CVS info documentation says:
From RCS
If you have been using RCS, find the RCS files--usually a file
named `foo.c' will have its RCS file in `RCS/foo.c,v' (but it
could be other places; consult the RCS documentation for details).
Then create the appropriate directories in CVS if they do not
already exist. Then copy the files into the appropriate
directories in the CVS repository (the name in the repository must
be the name of the source file with `,v' added; the files go
directly in the appopriate directory of the repository, not in an
`RCS' subdirectory). This is one of the few times when it is a
good idea to access the CVS repository directly, rather than using
CVS commands. Then you are ready to check out a new working
directory.
The RCS file should not be locked when you move it into CVS; if it
is, CVS will have trouble letting you operate on it.
A CVS training manual says:
9.1 Moving from RCS to CVS
Migration from RCS to CVS is a relatively painless process. CVS uses RCS for all of its file level version control, and so RCS files can be directly
copied into a CVS repository.
In converting from RCS to CVS, make sure that you follow these steps:
Steps to convert one directory from SCCS to CVS
-
sccs2rcs
-
co RCS/*
Steps to convert a tree from SCCS to CVS
- Cleanup, removing .class files and codeDoc
make distclean
cd ptolemy
find . -name "*.class" -exec rm {} \;
find . -name "*codeDoc" -exec rmdir {} \;
- Be sure that there are no files checked out
cd ~ptII
make sccsinfo
/users/ptII/adm/cvs/unsccsedit might help for testing
- You may want to remove all the extra sccs files
that are not being used. One way to do this is to
do a
sccs get SCCS in all the directories
and then run make checkjunk and move
the SCCS files to a junk directory.
- Convert from SCCS to RCS:
/users/ptII/adm/sccs2cvs/sccs2rcsall -q
sccs2rcsall is a locally modified
version of contrib/sccs2rcs that
is shipped with the cvs sources. The version
that is shipped with cvs has y2k hacks not
present in the local copy, so use the most
current version from the cvs sources.
- Check that there are no SCCS directories:
find . -name SCCS -ls
If there are any, fix the problems and rerun
sccs2rcsall
- Check out all the files from RCS
/users/ptII/adm/sccs/2cvs/coall
coall looks like:
#! /bin/sh
# Check out all the files from RCS in a tree
dirs=`find . -xdev -type d | egrep -v "SCCS|adm|RCS|CVS|codeDoc"`
homedir=`pwd`
for dir in $dirs
do
echo $dir
cd $dir
co RCS/*
cd $homedir
done
- Initialize the repository
setenv CVSROOT ~/tmp/ptII/Repository
cvs init
- Convert from RCS to CVS
cd ptII
/users/ptII/adm/cvs/ptrcs-to-cvs ptII
ptrcs-to-cvs is a version of
cvs-1.11/contrib/rcs-to-cvs with the following changes:
cxh@maury 67% diff -c ~/src/cvs-1.11/contrib/rcs-to-cvs ptrcs-to-cvs
*** /home/eecs/cxh/src/cvs-1.11/contrib/rcs-to-cvs Mon Sep 25 18:59:01 200\
0
--- ptrcs-to-cvs Thu Nov 12 14:01:25 1998
***************
*** 90,96 ****
if [ -d SCCS ]; then
echo SCCS files detected! >&2
! exit 1
fi
if [ -d RCS ]; then
co RCS/*
--- 90,97 ----
if [ -d SCCS ]; then
echo SCCS files detected! >&2
! #exit 1
! exit 0
fi
if [ -d RCS ]; then
co RCS/*
***************
*** 99,105 ****
for name in * .[a-zA-Z0-9]*
do
case "$name" in
! RCS | *~ | \* | .\[a-zA-Z0-9\]\* ) continue ;;
esac
echo $name
if [ $vbose -ne 0 ]; then
--- 100,106 ----
for name in * .[a-zA-Z0-9]*
do
case "$name" in
! Repository | RCS | SCCS | old-SCCS | adm | codeDoc | *~ | \* | .\[a-zA-Z0\
-9\]\* ) continue ;;
esac
echo $name
if [ $vbose -ne 0 ]; then
***************
*** 158,171 ****
echo "MSG: Copying old rcs file."
cp RCS/"${name}",v "$file"
else
! if [ -n "${comment}" ]; then
! rcs -q -i -c"${comment}" -t${message_file} -m'.' "$file\
"
! fi
! ci -q -u1.1 -t${message_file} -m'.' "$file"
! if [ $? -ne 0 ]; then
! echo "ERROR: Initial check-in of $file failed - abortin\
g" >&2
! exit 1
! fi
fi
else
file="${update_dir}/Attic/${name},v"
--- 159,174 ----
echo "MSG: Copying old rcs file."
cp RCS/"${name}",v "$file"
else
! echo "MSG: not checking in $name, since it is not in RCS"
!
! #if [ -n "${comment}" ]; then
! # rcs -q -i -c"${comment}" -t${message_file} -m'.' "$fil\
e"
! #fi
! #ci -q -u1.1 -t${message_file} -m'.' "$file"
! #if [ $? -ne 0 ]; then
! # echo "ERROR: Initial check-in of $file failed - aborti\
ng" >&2
! # exit 1
! #fi
fi
else
file="${update_dir}/Attic/${name},v"
- I got "Failed to connect after 6 attempts . . ."
-
If you use CVS several times in succession, then you
may see a message like
Failed to connect after 6 attempts, try again in a few minutes or use
'-oConnectionAttempts 1000'
cvs [update aborted]: end of file from server (consult above messages if
any)
The reason that this occurs is because Windoze does not close sockets promptly.
The fix is to create or edit c:/ssh/etc/ssh_config
and add
Host *
ConnectionAttempts 100
- cvs update is slow
-
There are many reasons why
cvs update
could be slow.
Network Connectivity
The most common reason for CVS to misbehave is poor
network connectivity.
If you have DSL, see the
DSL Troubleshooting FAQ.
Below are some things to try:
- Reboot - Often the quickest solution.
- In a separate window, trying pinging gigasource
while running the cvs command, and then hit
Control-C
and look for dropped packets in the output.
ping -t gigasource
SSH has a hard time with packet loss greater than 3%.
- Be sure that names are properly being resolved, try
nslookup gigasource
- Sometimes it seems like starting over and checking
out a new tree is necessary. This is especially true
if you have been having network connectivity problems
because of dropped packets.
- (A long shot): If no packets are getting through and your machine
is usually on the UCB EECS 171 network and now
you are at home connected via modem,
then you might have route problems
where your machine is routing 171 net packets to
the ethernet port instead of the modem.
If this is the case and you are connected via modem, run
route print
and be sure that you have no 171 network entries.
CVS
SSH
- Verify that ssh is working with
ssh -v gigasource cvs --help
- How do I set it up so that I get email whenever someone commits a change?
-
To set up email for CVS or Subversion:
- Email will be set to people with
developer accounts
in a workgroup. To enable the developer choice,
the admin of the workgroup should go to Admin -> Group Profile and select Software Project
- the admin of
the workgroup should go to the Admin->Configure Group page
and click on the
CVS mail list checkbox
- then
hit the
Change group configuration button
at the bottom.
This will send email to the CHESS website managers
who will create the mailing list and edit the
loginfo file.
If a workgroup member would like to get email about
cvs changes, they should go to
Options -> Membership and request a Developer account for
the corresponding workgroup.
(Website managers: the instructions for what to do for CVS are in the Infrax FAQ:
How do I set up cvs mail?
or in the dopsysadmin: How do I set up CVS mail?
For Subversion, see
Setting up SVN Email)
See also How do I send commit log messages to a mailing list?
and
E-mail Notifications
- Cygwin copy and paste
-
You may need to enable copy and paste in the bash shell window.
If copy and paste are working properly, then you should be able to
highlight text by left clicking and dragging the mouse over the text
and then hitting the Enter key to copy the highlighted text.
The Cygwin faq at
http://www.cygwin.com/faq/faq.html#SEC54
says:
How can I copy and paste into Cygwin console windows?
Under Windows NT, open the properties dialog of the console
window. The options contain a toggle button, named "Quick edit
mode". It must be ON. Save the properties.
Under Windows 9x, open the properties dialog of the console
window. Select the Misc tab. Uncheck Fast Pasting. Check QuickEdit.
You can also bind the insert key to paste from the clipboard by adding
the following line to your .inputrc file:
"\e[2~": paste-from-clipboard
More precisely:
- Start up bash by doing Start -> Programs -> Cygwin -> Cygwin Bash\
Shell.
- Right click on the icon in the upper left of the title bar and select
properties
- Under Options, select 'QuickEdit Mode', hit ok
- Select 'Modify shortcut that started this window', hit ok
To copy from a Cygwin Bash window, highlight the text
by left clicking and dragging the mouse, then hit
the enter key to copy the text.
To paste into a Cygwin Bash window,
click on the icon in the upper
left of the window menu bar and then do edit -> paste
or do Alt-Space e p
That is, hold down the Alt key, hit the
space bar, let up on
the Alt key, type an e and then a p.
- Is there a web interface to the CVS repository
-
No, there is no webcvs view.
The basic issue is that it is tricky to integrate a web
server based solution into our current permission scheme on the
website. Using a full blown client that uses ssh to access the
repository is probably better than using a client on the web server.
Jorn Janneck wrote:
it seems that these days there are a number of decent user
interfaces to cvs. intellij, e.g., has a very nice cvs integration,
including visual display of conflicts, and i hear that so does
eclipse (never tried it with cvs myself), and i think jbuilder. of
course, these are programming ides---if that is too heavy, i believe
that there are a number of graphical cvs-only frontends, which i
never used though, because i am very happy with intellij. a list of
those can be found here: http://www.wincvs.org/
- SVN vs CVS
-
I've not set up a SVN repository, I'm sure it is fairly easy.
I have set up a CVS repository. Setting up CVS is very easy.
If people have ssh access to the repository, then you are done.
Setting up anonymous access (pserver) is not that hard.
See anonymous access
Lots of people swear by SVN and swear at CVS. One issue is
that I'm not so sure Eclipse supports SVN out of the box,
though there is the subclipse plugin.
That's what stops me from switching.
In general, one issue I've found is that people tend
to jump to the newest technology for odd reasons.
Yes, the CVS command line and branching mechanism sucks
and SVN is "better". However, are you willing to
toss access via Eclipse out of the box if you go
with SVN? I'm sure there is a SVN interface to
Eclipse, but why make installation even more complicated
for users who are already struggling?
I think it is better to set up CVS or SVN on a Linux box
than a Windows box. One issue is that the Windows so called
file system is case insensitive, case preserving, which
means that if under Unix there are two filenames differing
only in case, then god knows what happens if the
CVS or SVN server is on Windows.
- How do I use CVS on the Mac under OSX
-
To use CVS on the Macintosh, you should install Xcode, which is about 1 gig in size!.
Other resources:
|
