> On Sun, 1 Aug 1999, Khimenko Victor wrote:
>
> > > The reason I ask this is that my understanding of the way the VFAT fs
> > > works implies that the two names are effectively independant, and the
> > > only requirement attached to them is that they both point to the same
> > > file.
> >
> > Wrong. LFN is attached to short name.
>
> Please correct me if I'm wrong, but it sounds very like an
> LFN is actually a (rather restricted) symlink.
>
Not exactly. It just placed before files itself in directory and has
simple checksum of short filename inside. It's described in vfat.txt in
Documentation/filesystems directory.
> Perhaps that's a potential solution.
>
> It could make for rather ugly directory listings, but IMO
> that's better than the ugly code that we have now.
>
I'm not sure if it'll be less ugly. It's not untypical for a lot of
programs to create a temporary file and then remove them. With your
symlink proposal just LFN will be deleted and file remains untouched.
The same is true for original hardlink's suggestions.
In short: ANY such scheme will solve some problems and introduce other
nasty problems.
P.S. And no, we can not just ignore short names/long names collisions and
create files with new file with LFN conflicting with short name or other
file. It'll work with Linux, but Windows9X/NT will be confused badly and
if you do not need Windows9X/NT interoperability then better to use ext2fs
or raiserfs anyway... The same goes against the idea of changing short
names when needed...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/