Subject: Re: "tfs" and other filesystems with very short names
To: None <>
From: Greg A. Woods <>
List: source-changes
Date: 09/23/1998 14:41:39
[ On Wed, September 23, 1998 at 15:06:29 (+0100), Andrew McMurry wrote: ]
> Subject: Re: "tfs" and other filesystems with very short names
> Especially if all the fs's move into a seperate directory, I do not
> see any problem with filesystems with a 'long' name (eg filecore,
> coda, iso9660) just being known by that name (without 'fs' on the
> end). New filesystems should be encouraged to use this
> format. However, the 'fs' on nfs, ffs, mfs etc is an integral part of
> the name of the filesystem and should be kept.

Hmmm....  I think it all depends on your experience and breadth of
knowledge.  For example I "know" that iso9660 is a filesystem standard,
but does everyone?  What about "filecore" and "coda"?  I'd never heard
of them until recently and had no clue what domain they belonged in
(except for the fact that it was actually "filecorefs", which gave me an
important clue).  And then there are overlapping names, such as "msdos"
and "ados", and who knows how many more.  Are those interpreters, or
emulators, or filesystems, or what?  In the kernel namespace these
issues are of course more critical since there is as yet no "sys/fs"
subdirectory in which all filesystems exist, but even if they were such
a directory there are other places where these names are used without

It's all a matter of where you draw the line.  Indeed "Berkely Fast File
System" is the title on a paper very well known to the BSD crowd, so
yes, the "fs" in "FFS" is an integral part of the name (similarly NFS).
However I'd argue the same applies for "MSDOS File System" or even
"ISO9660 File System".  I still don't know/remember what "CODA" stands
for, but I suspect "File System" is an important part of describing it.

Especially with new things there's a "chicken & egg" problem too.  I see
"coda" stuff coming along, but since it's new, and my existing system
has no response to "apropos coda", nor is it in any other on-line
glossary I may have, so if it were called "codafs" I'd be much more
likely to guess it's a "CODA File System", and unless I'm interested in
new filesystems I'd know I can ignore it for now.

							Greg A. Woods

+1 416 218-0098      VE3TCP      <>      <robohack!woods>
Planix, Inc. <>; Secrets of the Weird <>