Re: 1.3.62 and fat/msdos/vfat observations

H. Peter Anvin (hpa@freya.yggdrasil.com)
20 Feb 1996 04:06:03 GMT

Followup to: <971137578CD@rkdvmks1.ngate.uni-regensburg.de>
By author: "Ulrich Windl" <Ulrich.Windl@rz.uni-regensburg.de>
In newsgroup: linux.dev.kernel
>=20
> My original idea was something like: UMSDOS currently does two=20
> things: First it implements UNIX file attributes that MSDOS/FAT does=20
> not have by maintaining EMD files. Secondly it maps long names that=20
> DOS/FAT can't handle to "mostly reasonable" substitutes.
> With VFAT we would have a nicer mapping for the second point.=20
> Admittedly at the first instance I forgot the upper- and lowercasing=20
> problem and I thought we wouldn't have to remap file names at all.
>=20

Also, VFAT as a filesystem is basically begging for corruption. I
would suggest UMSDOS-on-top-of-VFAT should work by treating the VFAT
long names as comments; create them for the benefit of the user, let
umssync use them to name files not present in the EMD file, but don't
use them for name translation.

-hpa

--=20
PGP public key available - finger hpa@terminus.storm.net
"The earth is but one country, and mankind its citizens." -- Bah=E1'u=
'll=E1h
I don't work for Yggdrasil, but they sponsor the linux.* hierarchy.