Re: [9fans] is plan 9 based on XML?

Re: [9fans] is plan 9 based on XML?

[dmr]

 > When did the first hierarchical file system appear?  Multics?  CTSS?

The first I saw was Multics.  CTSS was approximately 2-level,
perhaps with more than one root.

	Dennis

Re: [9fans] is plan 9 based on XML?

[C H Forsyth]

> When did the first hierarchical file system appear?  Multics?  CTSS?

OS/360 had hierarchical names ca. 1964/5, although much
might depend on the interpretation of `appear', given that
system's troubled early history, and i recall someone telling me the
implementation was unusual.  Several groups seemed to be
implementing them about the same time.  None of my old books mention
who suggested it first, or indeed when Microsoft first got the patent.
I've got some more old books at the office, though.

Re: [9fans] is plan 9 based on XML?

[Geoff Collyer]

A google search turned up this claim at
http://turing.cs.camosun.bc.ca/comp112/notes/chapter5.html:

      Among the first such systems (if not the first) was MIT's CTSS
      (Corbató et. al) implemented on an IBM 7090.  This was also the
      first system to implement a hierarchical file store (i.e. one
      with directories - aka folders - and files with them)

Re: [9fans] hierarchical filesystems for non-file things?

[David Tolpin]

Is the use of a uniform hierarchical (filesystem-like) representation
for various non-filesystem things an invention brought with Plan 9?

Re: [9fans] hierarchical filesystems for non-file things?

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 632 bytes --]

Depends on what you want to call that which Unix did.  We stuck the
proc file system there before there was a plan 9.  Also, Unix had
devices as files, though a hierarchy of 1 is not that general.

There were various file systems on Unix, like the semantic file
system, that represented more than just files.  I remember similar
things also presented in Usenix over the years.

Then there was the Mumps computer system at the Mas General Hosp
built in 1969.  It was a system that presented an integrated
language and database where everything available to the language
was a persistent object.  Not hierarchical though.

[-- Attachment #2: Type: message/rfc822, Size: 2205 bytes --]

From: David Tolpin <dvd@davidashen.net>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] hierarchical filesystems for non-file things?
Date: Fri, 26 Mar 2004 13:05:46 +0400 (AMT)
Message-ID: <200403260905.i2Q95kfs076757@adat.davidashen.net>


Is the use of a uniform hierarchical (filesystem-like) representation
for various non-filesystem things an invention brought with Plan 9?

Re: [9fans] is plan 9 based on XML?

[David Presotto]

[-- Attachment #1: Type: text/plain, Size: 577 bytes --]

I used CTSS in the 1971/72, my first year at MIT.  They gave undergrads spare
time accounts on any systems there.  It was at the end of its life, was way
overloaded, and ran on a 7094 that stuck around for quite a few years later.

It didn't have a hierarchical file system per se.  Just a single level of
directories.  I can't remember how you named an other user's directory
or even the system one.

Multics fixed that and added much more.

So, I went to see what I could find on the web...

	http://larch-www.lcs.mit.edu:8001/~corbato/turing91

is a nice read.

[-- Attachment #2: Type: message/rfc822, Size: 2106 bytes --]

From: Geoff Collyer <geoff@collyer.net>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] is plan 9 based on XML?
Date: Fri, 26 Mar 2004 01:00:04 -0800
Message-ID: <378214d4b24f870a683cd4d88ea5b009@collyer.net>

A google search turned up this claim at
http://turing.cs.camosun.bc.ca/comp112/notes/chapter5.html:

      Among the first such systems (if not the first) was MIT's CTSS
      (Corbató et. al) implemented on an IBM 7090.  This was also the
      first system to implement a hierarchical file store (i.e. one
      with directories - aka folders - and files with them)

Re: [9fans] is plan 9 based on XML?

[rog]

> A google search turned up this claim at
> http://turing.cs.camosun.bc.ca/comp112/notes/chapter5.html:

some interesting snippets there.

i particularly liked this throwaway line:
"...  a replica Colossus was completed ....  A Pentium was programmed
to perform the same task and Colossus was found to be just as fast."

shows the value of h/w acceleration, i guess.

Re: [9fans] hierarchical filesystems for non-file things?

[boyd, rounin]

> Depends on what you want to call that which Unix did.  We stuck the
> proc file system there before there was a plan 9.

8th Ed