Re: WevDAV anyone?

Christian Scholz (ruebe@aachen.heimat.de)
Sun, 27 Jun 1999 00:21:34 +0200 (CEST)


Hi!

> > I am wondering if anyone wants to do something like that under
> > Linux (means, writing a webdav filesystem which acts as a client.
> > A webdav server in addition to NFS would maybe also be interesting
> > but maybe not that important..)
>
> A modified user-space NFS server might do what you want (essentially,
> re-export the WebDAV filesystem as an NFS filesystem). But good luck
> supporting all the extra file properties that DAV gives you in a clear
> manner (every file as a directory, perhaps, with properties represented as
> files under there, with a special file named "data" or somesuch that
> represents the data associated with that object? Just a thought...).

Well, that's maybe too confusing to the user. For the simple opening/
writing of files they're not neccessary so one would maybe omit them
for the first try (there is actually also not that many software
which is using it).

But actually the idea about using a modified NFS-Server sounds
interesting..
The question might then be if it's not a good idea to have
a general stub in the kernel for attaching user space filesystems.
There might be some more ideas for filesystems which will be too
much code to be in the kernel (and the DAV-FS would one of these
as also XML has to be parsed inside it, etc.)

> Alternatively, you might look at the virtual filesystem work done with
> Midnight Commander and GNOME; the following URL has a brief overview:
>
> http://www.gnome.org/devel/arch/vfs.shtml

Well, as far as I understand it that's then only for GNOME apps.
Thus not the goal ;-) But maybe at least a good start.

> Those would be good proof-of-concept designs, before going about bloating
> the kernel with something that might not end up being incredibly useful.

I know that it's useful, I see on Windows. But you're right that it will
blow up the kernel too much.

> Show that it works, then integrate with the kernel if a userspace setup
> proves to be unacceptable. Thus I like the NFS-Server idea.

The only problem is that I don't have time to code that and my only question
was also if someone's maybe working on it so that I might test it.
Well, and as the answer is apparently "No" that's all I want to know.
When I have time I might start looking at the NFS-Server and maybe extending
it to make it a general server for user space filesystems (and not only
DAV).

Anyway, have a nice weekend!

Christian

-
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/