Re: dcache problems with vfat

Jamie Lokier (lkd@tantalophile.demon.co.uk)
Sun, 10 Jan 1999 00:49:34 +0000

On Thu, Jan 07, 1999 at 06:54:22PM -0500, Albert D. Cahalan wrote:
> Gordon Chaffee writes:
> > Yes, we really do need to be able to access by short names.
> > Things like Wine depend on them.
>
> Must they be fast though? You could immediately convert them to
> long form. With all the names cached in long form only, it ought
> to be a bit easier to avoid problems.
>
> BTW, these short names are really common. (vfat, ntfs, smbfs, iso9660)
> Perhaps they ought to be in struct stat. **duck** Well?

IMO, it is extremely desirable that Wine should *not* depend on the
underlying filesystem. Therefore, if Wine really depends on this
behaviour, it would be desirable to fix Wine. (I'm not sure that Wine
really requires this).

We also want Wine to work on other operating systems.

AFAIK, the only other program which uses access to the short names is
Samba. (a) Why? (b) Is there some other approach which Samba can use?

BTW, vfat is the only filesystem which will actually return the short
names to the application, through an ioctl(). ntfs, smbfs and iso9660
don't have such a mechanism, so can it really be that important?

-- Jamie

-
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/