Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!ucla-cs!zen!ucbvax!CU20B.COLUMBIA.EDU!OC.GARLAND
From: OC.GARLAND@CU20B.COLUMBIA.EDU (Richard Garland)
Newsgroups: comp.os.vms
Subject: FINGER
Message-ID: <12330910633.165.OC.GARLAND@CU20B.COLUMBIA.EDU>
Date: Mon, 31-Aug-87 12:35:09 EDT
Article-I.D.: CU20B.12330910633.165.OC.GARLAND
Posted: Mon Aug 31 12:35:09 1987
Date-Received: Sat, 5-Sep-87 00:44:59 EDT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 29

I am the author of the version of FINGER that has been mentioned of late,
although I have not worked on it for over 2 years.

The problem indeed exists and I must apologetically take the blame.  The
problem is related to a VMS feature that has long been a source of problems,
both real and conceptual since day one: namely the fact that a directory
entry is not part of the file header, and indeed the whole directory structure
is built on top of the file structure as a separate entity.  This leads to
problems with RENAME, incremental BACKUP, aliases, etc. etc.  This is
particularly severe when the directory has less protection than a file in it.

My feeling on FINGER is to go with Glen Everhardt's fix.  Most systems (such
as TOPS for example) allow the user to set the protection on a file such as
the plan file and that probably should not be overridden by a utility such as
FINGER.

For those who missed Glen's message, just comment out the USEROPEN argument
in the OPEN statements (2 of them).

To fix it the way it is, one would have to get the FIB of the file, then 
read the file header with an ACP QIO and check the file name, ownership, etc.
This is more than I can do and I prefer Glen's solution anyway.

For those who asked, the last version distributed by me was 41.1.10.  I made
a few changes since that but they do not add or change functionality.  It
is also known to work with the newest version of JNET for those who asked.

					Rg
-------