For better or worse, I am using rcs to track changes to HTML and JSON files exported from chatGPT. Some of these contain chats with embedded $keyword$'s. These get expanded on ci check in. How do I disble this without stepping on these files?

I want to see exactly what will be committed before I run a cvs commit command. I just recently added a file using cvs add myFile and then committed the file using cvs commit -m "my message" but when I ran the commit command, I ended up modifying other files as well as removing files that I did not intend to change or had no idea were going to be committed and so I had to revert these files back one by one. I believe it is called the staging area? Where files that are added but haven't been committed yet or ready to be commited? I just want to see what changes will be made before I actually run the cvs commit command.

Cleaning up after a former employee, I've found that they accidentally included 800MB of data files in a commit to CVS years ago. Several other commits to the same project were made afterwards. This data is extremely security sensitive and legally restricted and it needs to be removed completely (nothing hanging about in the Attic or filesystem anywhere). How do I get rid of the files without corrupting CVS? (It's a linux host, and I have root access and am CLI fluent, but not a CVS expert!)

When I attempt to log in to a CentOS box via ssh, it hangs after putting in the password. I can get in to the system via console and when i run w I can see "cvs server" running for the user trying to login. USER TTY FROM LOGIN IDLE PCPU WHAT root pts/1 systemname 11:37 1:14m 0.00s cvs server In the secure logs it looks like it was a valid login: Accept password for root from 172.0.0.17 port 53270 ssh2 pam_unix(sshd:session): session opened for user root by (uid=0) It appears like a normal login, but it will never actually give me the prompt. I've tried restarting the SSHD service, same issues. This is a production box so im a little hesitant to reboot it. I can't seem to find anything in the log files that looks suspect. When I run ssh -vvv and try to connect, it also looks fine but I never get the login message after "request accepting on channel.." Here is the paste of the end of -vvv: debug2: channel 1: request shell confirm 1 debug3: send packet: type 98 debug2: callback done debug2: channel 1: open confirm rwindow 0 rmax 32768 debug3: receive packet: type 99 debug2: channel_input_status_confirm: type 99 id 1 debug2: PTY allocation request accepted on channel 1 debug2: channel 1: rcvd adjust 2097152 debug3: receive packet: type 99 debug2: channel_input_status_confirm: type 99 id 1 debug2: shell request accepted on channel 1 Also, I went through the network firewall to see if maybe it was dropping packets or something, nothing like that is happening. I also tried from some other system, same issue. I've never seen this before and can't seem to find any good info on this exact issue. Any help, most appreciated! EDIT: I only have a root account on here and there is no .bashrc or .bash_profile that I can find in root, or anywhere else. However, I do see an cvs.sh and a cvs.csh in /etc/profile.d/. I did see a for loop in /etc/profile/ that is looking for scripts in /etc/profile.d/ to run at startup. Here are the contents of those scripts: cvs.sh # change default from rsh to ssh for cvs command export CVS_RSH=${CVS_RSH-ssh} cvs.csh # change default from rsh to ssh for cvs command if ( "$?CVS_RSH" == 0 ) setenv CVS_RSH ssh I thought maybe it was looking for cvs and also server but I don't see anything with server in the script directory. Is there somewhere else I should be looking? All the normal places ive seen profile files, nothing is there. Thanks!

I'm not sure if this is a cvs bug or what it is, but I'm debugging some crusty scripts and found > cvs history -D "-1 days" -c "s-server-moo-3" spits out > M 2018-03-02 19:06 +0000 user 1.71 s-server-moo-39 Network-Configs/configs == /opt/rancid/customers/CUSTA/var/Network-Configs > > M 2018-03-02 20:06 +0000 user 1.72 s-server-moo-39 Network-Configs/configs == /opt/rancid/customers/CUSTA/var/Network-Configs which is not what I want, it shouldn't find "39", it should only find "3". My old crusty script is totally the number of commits and logging warnings if the total is above 8 per day, therefore if s-server-moo-38 and s-server-moo-39 each have 5 commits, I get warnings. I know I could just grep afterwards, but this seems like a very strange way for any version control system to search. I don't see anything under history in the man page that says it's supposed to behave like this (or a flag to use as a workaround).

I tar'd the repository folder (including the CVSROOT) and copied it to the new server, then un-tarred it. When I connect to CVS via Eclipse, I can get the sourcecode, however the server is not providing the versions, branches or history. Is there some extra step I might have missed?

We just migrated a cvs file repository from one server to another. To test it out, I checked out, changed, then committed a file. The owner / group changed from "cvs / cvs" to "MyAccount / SomeoneElseAccount". I was expecting at least "MyAccount / cvs". 1. What was I supposed to expect? 2. Do I need to resolve this, or is this OK? 3. How might I resolve this if it is not OK? UPDATE Per the comments, I checked user ids. Results on same line after '='. id -u cvs = 60000 id -g cvs = 60000 id -G cvs = 600000 20 40 id -u MyAccount = 200007 id -g MyAccount = 200002 id -G MyAccount = 200002 60000 100005 100025 id -u SomeoneElseAccount = 200002 id -g SomeoneElseAccount = 200002 id -G SomeoneElseAccount = 200002 100678 I also had "AnotherUser" check out and commit; his file also ended up with group "SomeoneElseAccount". id -u AnotherUser = 200018 id -g AnotherUser = 200002 id -G AnotherUser = 200002 60000 100005 100025 Setup on the old server provides a similar id configuration with 200002, though cvs is '100' there, and the items are committed as group 'cvs'. Checking /etc/group OLD SERVER cat /etc/group ... cvs:!:200002:name,name,name,MyAccount,AnotherUser,name,name... ... NEW SERVER cat /etc/group ... cvs:!:60000:name,name,name,MyAccount,AnotherUser,name,name... ...

I have a huge cvs log file which, cleaned from the useless info, reads something like Working file: unmodifiedfile1.c ================ Working file: modifiedfile1.h ---------------------------------- revision 1.3 Fixed some bug ================ Working file: unmodifiedfile2.h ================ Working file: modifiedfile2.h ---------------------------------- revision 1.1 Added some feature ================ Working file: unmodifiedfile3.h I would like to clean the lines related to unmodified files: Working file: modifiedfile1.h ---------------------------------- revision 1.3 Fixed some bug ================ Working file: modifiedfile2.h ---------------------------------- revision 1.1 Added some feature ================ The pattern to match is Working file: FILENAME ================ What i've been able to do up to now is the following: sed '/Working file:/ N ; s/\n/PLACEHOLDER/' changelog.txt | grep -v 'PLACEHOLDER===' | sed 's/PLACEHOLDER/\n/ I'm sure however there is a cleaner solution that my sed ignorance precludes me... ( also, a bonus would be being able to erase the very latest line if necessary) P.S. An output ending with: ================ Working file: unmodifiedfile3.h is also acceptable

Hi I'm trying to logon to cvs automatically every time I start the machine. This works well but I have to manually type cvs login in the shell: setenv CVSROOT :pserver:$USER@vlsj-cccc:/vols/cvsroot So I tried to add the password in my .cshrc file and having trouble when inserting my password to it. cvs -d :pserver:$USER:password@vlsj-cccc:/vols/cvsroot login It works in a normal script file, but not work in a .cshrc file. How to deal with it properly?

I've trying to set up a CVS repository using ACL's (setfacl) and having trouble getting permissions to inherit properly. Namely, when a user commits to the repository, I want the file to inherit the acl's of the parent folder. However, using the -d option on the parent seems to cause issues with the read/write lock files that are automatically created by the CVS. Is it possible to do this with standard acl mechanisms? Thanks

Could you tell me can I add the line in commitinfo file if I want that the hook will be run if I commit an appropriate file? Example: I have *.txt file I added next line: *.txt $CVSROOT/CVSROOT/hook.sh %p %s and it is not working. What should I do ? Thanks.

Recently I am forced to migrate an ancient CVS installation from ubuntu to CentOS. I am using 'cvsnt' from 'March Hare'. I am trying to reproduce this old legacy server's function on my relatively new CentOS box. I have got cvslockd running, everything seems to be listening correctly, but after I do my export like: export CVSROOT=:pserver:username@127.0.0.1:/somerepo cvs login I get the error: > cvs [login aborted]: Unknown command: `pserver' My xinet.d conf is like: server_args = --allow-root=/somerepo -f pserver My question is: how can I resolve the above error and get pserver method working properly? [root@det1svn01 testuser]# cvs info Available protocols: local (internal) server server 2.8.01 (Soolin) Build 5590 pserver pserver 2.8.01 (Soolin) Build 5590 sspi sspi 2.8.01 (Soolin) Build 5590 ext ext 2.8.01 (Soolin) Build 5590

I'm trying to automate version-control repository update in crontab. The following script /home/usr/mrbean/bin/update.sh works in interactive Bash command-line: #!/bin/bash ROOTPATH="/home/user/mrbean/temp/" cd $ROOTPATH SSHTMPFILE="${ROOTPATH}ssh-agent.cf" ssh-agent -s | head -n 2 > "$SSHTMPFILE" function killsshagent { /bin/kill $SSH_AGENT_PID } trap killsshagent EXIT if [ -f "$SSHTMPFILE" ]; then source "$SSHTMPFILE" ssh-add rm "$SSHTMPFILE" fi killsshagent if [ hostname == "mrbeancomp.local" ]; then REPODIR="/home/usr/mrbean/repo/" cd $REPODIR cvs update -dP > "$HOME/log/log.txt" fi I then add @hourly /home/usr/mrbean/bin/update.sh The script works by itseful when invoked in interactive Bash shell but not via crontab. It gives errors and sent to system mail /var/mail/mrbean: Permission denied, please try again. Permission denied, please try again. Permission denied (publickey,gssapi-with-mic,password). cvs [update aborted]: end of file from server (consult above messages if any) ### [Update] killsshagent added in program so ssh-agent is not dangling. ### [Update] Tried an adapted program on a Git repository and experienced same issue. Mr. Bean isn't allowed to immediately abandon this CVS repo and switch to a new Git one. ### [Update] I have multiple identity files in $HOME/.ssh/. Is it possible that when running from crontab, I need to specify which identity file to use? If so, how?

I have a directory and file with these permissions: # file: blue/ # owner: user-a # group: group-a user::rwx user:user-b:rwx group::rwx mask::rwx other::r-x default:user::rwx default:user:user-b:rwx default:group::rwx default:mask::rwx default:other::r-x # file: blue/bluer.php # owner: user-a # group: group-b user::rw- user:user-b:rwx group::rwx mask::rwx other::r-- I run cvs up -d /home/user-a/blue/bluer.php and now the file has the following permissions # file: blue/bluer.php # owner: user-a # group: group-b user::rw- user:user-b:rwx #effective:r-x group::rwx #effective:r-x mask::r-x other::r-- According to [can not explain acl behavior](https://unix.stackexchange.com/questions/56278/can-not-explain-acl-behavior) , the running program has to ask for permissions explicitly. How can I cause cvs update to use the destination folder's default acl mask?

I know CVS is pretty old but have to use it in company. Recently we have upgraded a library to the latest version. Now we want to replace the old library with new library. So basically we want to first rename the old library in CVS and then commit new library. In short 1. In CVS server: Rename folder /.../js/library to /.../js/library_legacy. 2. To CVS server: Commit /.../js/library to /.../js/.

If I remove a file from CVS with rm -v testfile, cvs remove testfile and cvs commit testfile, then the *testfile* is moved to attic. I can restore the file from attic with cvs add testfile and cvs commit testfile commands. However, I need to know the exact file name. How can I list all the files in attic?

At work we have a server which hosts our CVS repository and in order to restrict shell access to this system, I have deployed [rssh](http://www.pizzashack.org/rssh/) which has worked well for a number of years. We are now converting part of our repository to Git and using this existing server to host a centralized bare Git repository. I would like to extend these restrictions to Git commands as well, but rssh does not support Git. I know that Git has [git-shell](http://git-scm.com/docs/git-shell.html) which does a similar thing for Git (though I haven't used it) but I don't think it understands CVS. Given that both approaches use the mechanism of setting the user's shell to one of these, I can't use them both at the same time. We still need to use the CVS repository so **is there any mechanism that would work for both CVS and Git?** My options at the moment are: * Do nothing (not really an option). * Find a way to make them both work (hence this question). * Forget security and just give everyone shell access. * Abandon the idea and put the central Git repository on another server (probably virtual) and use git-shell. * Something else which I haven't thought of.

I'm interested in installing CVS on Scientific Linux. I'm new to Linux and I need to input the command: "cvs co X" My terminal states that what I have typed is an "invalid command," so I'm assuming that the problem is that I do not have CVS installed. After a lot of googling, I downloaded cvs-1.11.23, where the HTTP download link is provided [here](http://www.linuxfromscratch.org/blfs/view/svn/general/cvs.html) . I see that below the download link, there is a section written "Installation of CVS." I'm new to Linux, so I have little idea of the meaning behind these commands it says to type in, but I decide anyway to copy and paste them into terminal and see if they work. (Naturally) just punching in: ./configure --prefix=/usr --docdir=/usr/share/doc/cvs-1.11.23 && make results in the following error message: > "./configure: no such file or directory". Any ideas what I'm doing wrong? I'll admit I have very little comprehension of what this configure command is or how I should be using it. Again my goal is simply installing CVS so that I can use the command "cvs co X".

I did a checkout recently, and one of the files was not the file that I checked in, but rather it was a diff of that file and the previous version. Is cvs known for this kind of error?

I recently read this on the OpenBSD misc mailing list: "John Darrah" writes: > Hi. Would it be possible to get SSL on the OpenBSD website(s)? > It would be just a couple lines to change in nginx.conf/httpd.conf. > SSL certificates are free from Startcom and cheap from other vendors. > It would be really nice to have, even if it's not the default. I feel naked > viewing the site over plain http. Thanks. If you want encrypted access to the OpenBSD web site, do a CVS checkout and view it locally... I'm not a programmer, so that hint doesn't tell me how I can do this.