unix filename/patch substitution functionality?

Teus Hagen teus at nlnet.nl
Tue Mar 13 09:08:09 GMT 2001


An added functionality as first step to the "mangle filename" samba
functionality could be a first step. It is much easier bit not
as functional as a transludent filesystem. 
E.g. a functionality where the last part of a file path is mangled:
(*/one/two/three/ */%u/three/)
should one have (one/two/three/* %u/three/*) as well?
Take care the %u for user names substitution and other samba substitutions
is an extra feature: e.g. mangled map = (*_%u.html *.htm) does not work as you may expect yet.
The other thing is that mangling a path is on the unix filesystem level.
I see tricky situations for full path name expression substititions, but maybe I'm wrong.
Items as caching and maybe path name mangling database support may pop up.
This unix path name mangling will solve may smelling cheese with holes problem.

Of course (again) this is not the same as a transludent filesystem, which latter maybe
should not be part of samba?

teus
David Collier-Brown wrote:
> 
> Teus Hagen wrote:
> > Many DOS/Windows applications are installed in such a way that there is a need
> > to have some files write accessable for the world. In this way
> > your filesystem is looking like a swiss chees with lots of security holes.
> > It will going to smell well after a while.
> 
>         Yes, that would bge an excellent use of a translucent
>         fiesystem: this could be prototyped using Tim Potters
>         VFS-for-samba.
> 
> --dave (whould like one himself for Solaris) c-b
> --
> David Collier-Brown,           | Always do right. This will gratify
> Performance & Engineering Team | some people and astonish the rest.
> Americas Customer Engineering  |                      -- Mark Twain
> (905) 415-2849                 | davecb at canada.sun.com




More information about the samba-technical mailing list