How many of you use the lndir command when working in a design database? It is a very handy tool that was created to clean up a database that had a large number of generated files.

Suppose you download the orpsocv2 database and run the simulations. If you do a svn stat then you will see that you created hundreds of generated and log files in the database. If you are adding new files into the database then these files are also buried in this mess of file names and can easily be overlooked when checking in.


A better way is to go to the same directory that holds orpsocv2 and type:

> mkdir orpsocv2_cmp
> cd orpsocv2_cmp
> lndir ../orpsocv2 .


You now have a duplicate directory that contains all the same files as orpsocv2 as symbolic links back to orpsocv2. If you run your sims in this directory then all the generated files stay here an don't show up in the subversion database. If you edit a file then the original will be changed and that will show up with a svn stat command.

Cleaning a database is easy. rm -r * in the _cmp directory and rerun the lndir command. You will need to do this when you add a new file.

I have used this for years and it really helps keep the database clean.

John Eaton