Messages in this thread | | | From | "Jeff Merkey" <> | Subject | Re: FENRIS & 2.0/2.2 problems | Date | Wed, 23 Jun 1999 16:51:38 -0600 |
| |
Peter,
David Goebel and I have been discussing writing a "universal" interface that will allow linux file systems and NT filesystems to be used interchangably between the two platforms. The benfits to Windows NT are obvious, and David is already looking at putting together an interface that will allow Linux filesystems to be loaded under Windows NT. We are probably about 85% of the way there to have this for Windows NT. Going the other way, and being able to load Windows NT File systems on Linux is much harder -- particularly since the GNU compiler doesn't support the NT syle of structured exception handling, which is what's required to actually do this on Linux. Also, many of the problems with the NTFS "clone" on Linux are due to a general lack of understanding of the way the log file functions in NTFS, and the semantics of structured exception handling that are missing in the Linux world. Based on our agreements with MS, neither David or I can help the NTFS project on Linux -- MS would get very pissed. But we can work in other areas that may be beneficial for both platforms moving forward. An unchanging "universal" file system interface would be very hot in Linux.
Stuff for future development.
Jeff
----- Original Message ----- From: Peter Waltenberg <peterw@dascom.com> To: <jmerkey@timpanogas.com>; <alan@lxorguk.ukuu.org.uk> Sent: Wednesday, June 23, 1999 4:45 PM Subject: FENRIS & 2.0/2.2 problems
> My sympathies on the API changes. > > As one of the people who helped get NCP working again when 2.1 broke > I can say with great confidence that Alan is wrong. It may have caused > HIM very few problems, but it was a PITA for us ;) > > The worst problem was the fact that there were no decent docs on how > to handle the conversion from inodes -> dentries. If Bill Hawes hadn't > been around to help it'd never have got done. > > The improvement in performance almost made the pain worth while... almost. > > What might be worth considering is a synthetic external filesystem interface > module. Make that ALWAYS present a consistant interface to fs's that use it, > that way "externally coded" fs's will always work though maybe with a large > performance hit. That gives a much larger window for any commercial vendors. > > There's the obvious problem with that of course, it reduces the pressure to > produce a system that performs well on Linux. > > Cheers > Peter > ---------------------------------- > E-Mail: Peter Waltenberg <peterw@surf.dascom.com> > Date: 24-Jun-99 > Time: 08:34:08 > > This message was sent by XFMail > ---------------------------------- >
- 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/
| |