If the same utility script is checked in to different SCM directories yet should be the same everywhere, how can I verify they are the same?

For the Makefile example, let's say that someone is working alone on a new and improved variant, projectX, and has updated the Makefile to automate a formerly manual task. You're responsible for making sure that the rest of the older projects catch up, but someone might have already done some of the updates.

You could use the script below to run diffTree.sh Makefile and get output like this:

bash$ diffTree.sh Makefile
diff ./newprojects/projectX/Makefile ./projectB/Makefile
1,2d0
< CONFIGPARAM = "foo"
17c5
<       echo "I did that thing for you"
---
>       echo "You should do that thing"
diff ./projectB/Makefile ./projectC/Makefile
diff ./projectC/Makefile ./projectD/Makefile
end of files
bash$ _

This tells you that projectB and C and D are all identical, but newprojects/projectX/Makefile is different from B, and therefore also equally different from C and D.

If all the files in different locations with the same name are the same, you'll get something like this:

bash$ diffTree.sh sameFile.h
diff ./newprojects/projectX/src/headers/sameFile.h ./projectB/src/headers/sameFile.h
diff ./projectB/src/headers/sameFile.h ./projectC/src/headers/sameFile.h
diff ./projectC/src/headers/sameFile.h ./projectD/src/headers/sameFile.h
end of files
bash$ _

If you specify a filename that doesn't exist at all in the tree, you'll get:

bash$ diffTree.sh foo
end of files
bash$ _

Here's the script:

#!/usr/bin/env bash
set `find . -name $1 -print ; echo no files`
while [[ $# -gt 1 ]]
  do 
    if [[ $2 = no && $3 = files ]]
      then 
        shift
      else 
        echo diff $1 $2
        diff $1 $2
      fi
    shift
  done
echo end of files

It appears to handle wildcards in the filename, but there's probably something to be done to make that safer. Likewise, allowing for diff options like 'ignore whitespace' and so on would be an improvement.

Obviously, it totally breaks if you happen to have a file called 'no' and a second file called 'files' come up in the middle of the 'find' command's output, but since all the filenames are prefixed by './' and 'no' comes after 'files' alphabetically, I think it's safe enough.