From: Claude Morin (klode@syslab.csd.uwo.ca)
Date: 08/17/92


From: klode@syslab.csd.uwo.ca (Claude Morin)
Subject: fsck naming schemes
Date: Mon, 17 Aug 1992 17:18:33 GMT

This discussion probably belongs in the standards mailing list, but...

I've noticed a 'trend', when naming variations on a tool, of prepending a
letter, eg:
                 fsck - check minix filesystems
                efsck - check extfs filesystems
                dfsck - check MSDOS filesystems

I've seen the following on other systems:

                fsck - check *any* filesystem
                fsck.mfs - check minix filesystem
                fsck.efs - check extended filesystem
                fsck.dfs - check MSDOS filesystem

        In this case, fsck collects information (arguments etc) and passes
        them to the appopriate fsck.* If a filesystem is not currently
        mounted, you need to tell fsck which type to use (MIPS uses -T<type>),
        otherwise it figures it out for itself. Hmmm...perhaps fsck can
        be made to examing /etc/fstab to figure out the type of unmounted
        filesystems...

Pros: - ease of use -- usually only need to run 'fsck'
        - seems somehow *cleaner* to my little mind
        - scanning for commands/files starting with fsck is possible

Cons: - fsck must be written, or users are stuck with fsck.efs instead of
          efsck (for example)
        - why not just scan for *fsck* and avoid this whole issue?

This is an *option* offered for the consideration of those who may not
have thought of such a thing. If it's too much trouble, scrap it. If I (or
someone else) really want it, I can always write it myself, right?