The goal is to provide the same output as StatCVS. There are a few minor differences and one major: Subversion does not provide line count information in its log files. Therefore, StatSVN must have access to the server in order to compute the line count information. Because of this, StatSVN incurs a large performance penalty. However, this information is persisted to disk in order to speed up future computations. Edit
Why does the parsing of the log appear to fail?
- Nothing is generated, you should check that:
- You used the -v option to create the svn log
- the log file is not empty
- that you are running from a checked out directory
- that your checked out directory does not contain any non-committed items.
Why is StatSVN so slow?
- See above.
- Note that speed will pick up over time, as you will only be querying the SVN server for incremental differences.
- If you have full access to the SVN server and have a bit of time to do a manual setup, you can greatly speed up the process by running StatSVN on a local repository.
- To speed up the initial loading process, we recommend you do an 'svnadmin dump' on the remote server, compress and download the dump, and reload it locally via 'svnadmin create' and 'svnadmin load'. You don't even have to run the SVN server on your local machine, just use a filesystem repository ('file:///c:/svn/repo').
- You can then check out your project from this temporary repository, run StatSVN locally and get stats much faster than with a remote server. You can then go to the StatSVN cache directory (which defaults to your user's home directory /.statsvn/) and copy the newly created cache file over the one that is associated with your remote directory. (You'll need to know the UUIDs of your local and remote servers, or you simply infer them from the folder contents.)
- There is even a repository synchronization tool now available with Subversion. You could use this to synch your local copy and always run locally instead of remotely. Reduces the stress on the central server.
- If you're interested in writing a step by step guide, please feel free to edit this page!
Have you considered merging your changes back into StatCVS?
- Yes; and we have been doing so since version 0.2.0. We have contributed the XDOC generation, BugZilla, ViewVc integration amongst others.
Will you be offering XML output instead of HTML?
- The matter needs to be discussed with the StatCVS-XML team, if we don't merge back into StatCVS in a timely fashion.
StatSVN said I needed to upgrade to Subversion 1.3; I did and it still doesn't work. What do I need to do?
- You must checkout your project with the Subversion 1.3 client; this adds additional meta-data to the checked-out version that is required by StatSVN. Simply upgrading the tool is not sufficient.
Why am I getting errors about having invalid file revisions?
- Most probably you are not running StatSVN on a clean version of your checked out folder (must not have files marked as added, modified or deleted).
- Another reason would be your log file is not in synch with your checked out copy. Re-generate the logfile for this new revision.
How to install Java on Linux? Edit
Does StatSVN work in a headless environment?
- Yes, it does. Simply add the following argument to the java executable: -Djava.awt.headless=true
Can I run StatSVN with a custom diff command like vimdiff?
- No, you cannot. StatSVN parses the diff outputs and expects standard diff format.