6 after again against back before button cd control copy create cvs default development img last m made much name online open place print put release right say section select should show simple someone something store table text those tr type versions width xp
Here is a quick outline to setting up and using CVS on Windows (NT,
2000, XP). Unless you need to do branching and merging this is pretty
much all the commands you need to know about.
- Go to www.cvsnt.org (this
site is run by www.march-hare.com)
and download the binaries for CVS Server and Client for Windows
- Install the binaries somewhere
- Now you need to tell the CVS server where you would like to store
the files (i.e. your repository), go to the Windows Control Panel,
start the CVSNT Server applet. This has a number of tabs, the only one
you need to look at is the "Repository configuration". In that tab
there is an "Add" button, this will open the "Server Settings"
dialogue. In this you need to pick a location for the repository (just
navigate to a directory and select it) something like "c:\cvsfiles"
will do, and you should check the three check boxes: Publish
Repository, Default Repository, Online. Then say OK. CVS will
initialize the repository.
- Now close the applet
- For convenience sake you should add a CVSROOT environment
variable to tell the cvs commands where your default repository is.
Again, go to the control panel and open the "System" applet. Click on
the "Advanced" tab, then click on the "Environment Variables" button.
Now in the "User variables" section press the New button and then enter
"cvsroot" for the variable name and "c:\cvsfiles" for the variable
value. Click OK to close all the dialogues.
- Now start a command prompt and type cvs -v, this should print out
version information about CVS
- Next prepare a directory containing the files for the first
project you want to place into CVS, lets call the project "mole"
- CD into the project directory (note the name of this directory is
not important, also you may want to setup a simplified directory tree
that just contains a portion of the whole project and import that
first, then do a cvs checkout
and then use cvs add to bring
the rest of the files into the project - the import command will place
all files and subdirectories into CVS) and type the command: cvs import -m "" mole mole initial
- The mole project is now in CVS, to check that this is the case CD
to some other directory and type the command: cvs checkout mole
- You should now find a new sub directory called "mole" with the
files you imported in it.
- Now CD into the new mole directory and modify one of the files
- Run the CVS command: cvs -n
- This will tell you that you modified one of the files
- If you do the command: cvs
commit -m "A first test commit"
- the change you made to the file will be committed to CVS
- If you reach a particular point in your development of the files
(say you release a copy to someone else) you can mark the exact version
of those files in CVS for future recall (if needed). To do this
you first do a cvs commit so that version is in CVS, then you do a: cvs tag tag_name
- this will put the symbol "tag_name" (which you put something
useful in) on the current version of all the files in your project,
later you can checkout that particular version (if you need to patch a
bug) or you can diff against it.
- To see what the various versions were on a particular file you
can do a: cvs log filename
- To see what you have changed in your local copy of a file
compared to the previous version of the file you can do a: cvs diff filename
add a new file to your project, just create it in a project directory
you have checked out of CVS and then run the command: cvs add filename, after this you
will need to do a: cvs commit
command before the file actually enters CVS
- If you need to store binary files (like JPG images, Word
documents...) in CVS you need to use the -kb switch when adding them, so a
command like cvs add -kb picture.jpg
would be correct.