Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!umnd-cs!ub.D.UMN.EDU!rhealey
From: rhealey@ub.D.UMN.EDU.UUCP
Newsgroups: comp.sys.tandy
Subject: Re: Tree Directories For The TRS-80
Message-ID: <643@umnd-cs.D.UMN.EDU>
Date: Mon, 8-Jun-87 19:43:14 EDT
Article-I.D.: umnd-cs.643
Posted: Mon Jun  8 19:43:14 1987
Date-Received: Thu, 11-Jun-87 02:17:27 EDT
References: <867@vu-vlsi.UUCP>
Sender: news@cs.D.UMN.EDU
Reply-To: rhealey@ub.UUCP (Rob Healey)
Organization: U. of Minnesota, Duluth - Computing Services
Lines: 65

In article <867@vu-vlsi.UUCP> psuvax1!7GMADISO@POMONA.BITNET writes:
>This discussion on the idea of implementing a tree-type directory
>structure on the TRS-80 has kind of gotten under my skin.  I'm going to
>talk about this as working under LDOS 5.3/TRSDOS 6.3, since that's the
>existing directory structure I know best.  Note, however, that I am making
>**NO** claim to the ability to IMPLEMENT this system; just speculating on
>how such a system would work.
>
.
. Lot's of stuff
.
.
>
>I invite commentary on this idea; I realize that it's pretty preliminary,
>but I thought it might get the ball rolling.

	Wish I could find those damb notes I wrote last summer....

	Anyways, my idea as best I can remember. Bad news for 5.3 people,
	probably wouldn't work because of DOS vector's being non SVC driven.

	1) All extra code, buffers, tables, etc. reside's in high memory.
	2) Current working drive and directory info stored in tables in
	   Driver.
	3) All disks have root directorys, hash and GAT tables just like
	   they do already.
	4) Directory's are indicated by having the DAM of a directory and
	   are called dirnamex/SYS. Recursive directory's and relative
	   paths are possible.
	5) Only ONE GAT is maintained, the root directory contains it.
	   Each sub directory has it's own HASH table.
	   Also, system overlays MUST reside in the root directory so that
	   everything works right.
	6) Each directory has TWO default entry's just like the current one:
	   First entry is either BOOT/SYS for root or a copy of the parent
	   directory's directory entry, this allow's relative paths. The
	   second entry is of course the directory's own entry. The directory's
	   size is determined at creation and is one of the current sizes
	   allowed by the dos, i.e. Single Density Single Sided, SDDS, DDSS,
	   DSDD. 4 sizes depending on what you think will be neccessary.
	 7) Filename structure:

	    dir1,dir2,...,filename/ext.password:d

	    Comma's used to delimit directorys, two dot's stand for
	    a parent, i.e. :    ..,dir2,filename

	  8) cd, mkdir, rmdir, etc. are UTILITIES that handle directory's,
	     directory's are 100% invisable to the file system per say.
	     open's, closes and read/writes are intercepted in the SVC
	     table and all kludging is done there.
	  9) Search path is as follow's:

	     1) Current Working Directory, CWD.
	     2) Root of Current Working Drive CWDV.
	     3) Root directories of all other logged drives.

	 10) The modifying of the SVC and kernal look's messy as well as
	     handling FCB's.

	     This is all I can remember off the top of my head right now,

			-Rob Healey

			rhealey@ub.d.umn.edu