Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.rmit.EDU.AU!news.unimelb.EDU.AU!munnari.OZ.AU!news.ecn.uoknor.edu!paladin.american.edu!gatech!news.mathworks.com!uunet!in1.uu.net!news.artisoft.com!usenet
From: Terry Lambert <terry@lambert.org>
Newsgroups: comp.os.linux.development.system,comp.unix.bsd.freebsd.misc
Subject: Re: Ideal filesystem
Date: Fri, 19 Apr 1996 21:31:04 -0700
Organization: Me
Lines: 106
Message-ID: <31786888.87F5E2D@lambert.org>
References: <4hptj4$cf4@cville-srv.wam.umd.edu> <4jpjb6$77c@park.uvsc.edu>
<jlemonDpEw1v.4Ez@netcom.com> <4kfoqd$dgs@coyote.Artisoft.COM>
<DpvCB7.xn@midway.uchicago.edu>
<199604161704.SAA02116@kythera.demon.co.uk>
<31740EB4.20617DC8@lambert.org> <31745617.3E43CCA3@netcom.com>
<3175A76B.363FDA96@lambert.org> <wslojsd6t6.fsf@orcus.ping.at>
NNTP-Posting-Host: hecate.artisoft.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.01 (X11; I; Linux 1.1.76 i486)
Xref: euryale.cc.adfa.oz.au comp.os.linux.development.system:21936 comp.unix.bsd.freebsd.misc:17827
Robert Bihlmeyer wrote:
] Terry> [ ... example of using 'cat' to dump attributes on an
] Terry> attribute by attribute basis instead of using a "properties"
] Terry> command like OS/2 and Windows 95 use ... ]
]
] Terry> Seems like gratuitous complexity to pound a bunch of round
] Terry> pegs into a square hole because you love your square hole and
] Terry> are afraid to pull out the drill and Do The Right Thing...
]
] Just make the round pegs smaller... What does this metaphor
] prove, anyway??
Nothing. Metaphors never prove anything, they are just a way of
giving cannonical examples... in this case, an example of
gratuitous complexity. It's a parable to show "gratuitous
complexity is bad".
] True, wasting inodes for every EA would be excessive. But I don't see
] your point in making access control semantics an issue. You'd have the
] same problems with
Access control semantics are an issue only if the EA nad the file
to which they apply are somehow seperable. Name space exposure
rather than API exposure makes them seperable.
The issue of a seperated attribute is that you need to
"unseperate" it when it occurs.
Consider an attribute which *denies* access to a file if
you aren't root. If you seperate the attribute, access is
no longer denied.
Consider an attachment to a memo stored as an attribute (this is
a big exaggeration):
---------- DATA FORK (Attr: public)
From: The top brass
Memo: All employees
Subject: Hours worked
It has come to our attention that people are
leaving early when top brass is at the trade
show -- your employment contracts require you
to put in a full 40 hour week
...
---------- MANAGEMENT ADDENDUM (Attr: manager only)
You will track your employess time in and out; any
employee with less than 38 hours per week in the
next two weeks will be put on administrative
suspension and sent home for the day as a warning
to other employess
...
----------
Seperation of EA's (and loss of inheritance as a result) is
not a permissable event.
Now we are tasked with kludging a way to automatically
"unseperate" inherently seperable objects. This requires
formal metadata.
] readea -t purpose /etc/passwd
]
] than with
]
] cat /etc/passwd/purpose
]
] Remember that the EA streams are not intended as real files complete
] with inode & such. The system just mocks one up as a file if someone
] tries to open it as one.
This is just one of a subclass of applications of EA's, and can
be achieved without impacting the underlying implementation. A
loopback multiplex (as has been discussed) is one method. Using
a loopback, what is seperable is the name space exposure from the
attribute. This is a much more acceptable failure mode than
seperating the attribute from the file.
] Some people really find it desireable to use
]
] xv /usr/bin/vi/icon
]
] rather than
]
] readea -t icon /usr/bin/vi > /tmp/foo
] xv /tmp/foo
] writeea -t icon /usr/bin/vi < /tmp/foo
] rm /tmp/foo
I'd prefer they use
xv -default_icon /usr/bin/vi
8-).
But, again, a loopback can provide name space exposure without
putting the association between the files and attributes at risk.
Terry Lambert
terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.