I uncompressed a tar into a directory that has some files I need. I'd like to just delete those uncompressed files and tried:
tar ftx ft.tgz > rm
and about a million other variants and can't get it to do what I want. Can someone give me a hint on how to delete ONLY the files that are in the tar?
Regards,
Kurt

Your idea is good, but here is my hint:
then double check the resulting script.sh and when you are absolutely sure, execute
I'm not totally sure of $6 in the awk statement, but you can easily check that it matches the filename by issuing the command (alone)
I guess the untarred files do not have a relative path, that is they don't look like "some/dir/filename" but only "filename", otherwise we can parse the output of the above command to change the directory name accordingly.

If the files were extracted into their own directory, then it would be easy to remove the contents. If not, one thing you could do is use the find command with a narrow -cmin range:

find top/extraction/dir -cmin +500 -cmax -504 >filestodeletelist

You could also use the tarball's file listing to compare so you don't delete files that have been replaced since the extraction. ( Using comm or grep -f ).
This would prevent deleting a file that had been subsequently replaced.

You also want to learn about the xargs command. This is the one you would use to take a list from another command and supply them to another command. If some of the arguments contain spaces or other whitespace then use the -print0 option when using find and the corresponding -0 argument for xargs.

For example:
find ./ -cmin +1200 -cmin -1205 -print0 | xargs -0 rm
or
tar tzf tarball.tar.gz | tr '\n' '\000' | xargs -0 rm

The xargs worked like a charm. Thanks! (I didn't get to try the first suggestion since the second one worked the first time.)
Kurt