It seems to create two compressed files; one with a .bz2 extension and one without a .bz2 extension. I don't need two copies of the same thing. So I keep the one with the .bz2 extension and delete the other one. Since it appends the date, I have no easy way of determining (in an automated fashion) what it will append (I'm not a programmer, so while someone else might be able to do this, I cannot), so I was left with trying to figure out a way to delete all files that begin with taudit- but not those that end in .bz2. Assuming that the wild card placeholders would work in Linux the same as they do in DOS (* being a wild card for an unlimited amount of anything and ? being a wildcard for anything but only one character of anything) and given that the date appended will always be eight characters, I simply had it delete taudit-????????. That deletes what I don't want and leaves the files with the .bz2 extension.

But I've run into another problem. Because I'm rotating hourly and there doesn't seem to be an option in logrotate called timeext, the first logrotate for a given day works just fine. But all logrotates that come after that fail because they will have the same name. So, in addition to a postrotate entry, it would seem that I also need a prerotate entry, one that would change the name of the taudit-????????.bz2 file to taudit-????????.<HH:MM>.bz2 (with the time being where <HH:MM> is located). But that creates a problem:

1. I don't know how to go about getting the time inserted into the file as part of an automated process. Since all other files in the directory would already be renamed taudit-????????.<HH:MM>.bz2, there would only be one file that would match the filename pattern of taudit-????????.bz2, so I would hope that I could simply do something like mv taudit-????????.bz2 taudit-????????.$TIME.bz2 (or whatever I would have to do to get the time inserted in the filename.