Who introduced named files?

[Thomas Koenig]

I've been looking around a bit, but I cannot find which operating
system introduced named files.

OS/360 certainly had them, as did EXEC II. I would think that
peole started using them as soon as discs with random access made
their appearance - does anybody know more?

[Dallas]

I have wondered about that too.

I remember my transition from using decks of cards with names marked across the tops with a marker
pen to files stored on disk drives that used names for retrieval.

That was when I was a student at UT Austin in 1969 using the CDC-6600 and an operating system named
COS (Chippewa Operating System) .

The access I had to that computer as a student was limited to batch jobs - cards in - paper out.

Later as a staff member I had an account that could store data in named files on tape, and have
them retrieve my files for access via a TTY.

- Larry

Post by Thomas Koenig
I've been looking around a bit, but I cannot find which operating
system introduced named files.
OS/360 certainly had them, as did EXEC II. I would think that
people started using them as soon as discs with random access made
their appearance - does anybody know more?

[Rich Alderson]

Post by Dallas
That was when I was a student at UT Austin in 1969 using the CDC-6600 and an
operating system named COS (Chippewa Operating System) .

I knew that system! I was a freshman in the fall of 1969 at UT; I had a
work-study job with the School of Education's CAI Lab, working with an IBM 1800
in the basement of Sutton Hall and the administration's System/360 Model 50 in
the basement of the Tower.

When interviewing for a work-study job, I got sent over to the computer center
and got my first look at the 6600. The next year, I took a COMPASS class, but
they were running SCOPE by that time.

[Dallas]

You and I shared the same Freshman year at UT Austin.

What a huge enrollment! What a great football team.

Austin Texas and UT was such a fun place compared to my tiny rural high school in East Texas.

I started out as a Chemistry Major, but took a test during admissions that let me opt out of
taking the usual Freshman Chemistry course. So, I took an experimental course for freshmen
by Dr F. A. Matsen named "The Vector Space Theory of Matter", and they taught us FORTRAN.

I was hooked immediately on writing software.

So I changed my major to Math/CS.

I had a part time job coding for graduate students in the Physical Chemistry Dept (mostly FORTRAN
for doing matrix calculations for their Quantum Mechanics models) or various things for the CS Dept
through Dr. Brown all four years I attended as an undergraduate.

I had no idea that having a CDC-6600 as your first computer was quite unusual.

- Larry

Post by Rich Alderson

Post by Dallas
That was when I was a student at UT Austin in 1969 using the CDC-6600 and an
operating system named COS (Chippewa Operating System) .

I knew that system! I was a freshman in the fall of 1969 at UT; I had a
work-study job with the School of Education's CAI Lab, working with an IBM 1800
in the basement of Sutton Hall and the administration's System/360 Model 50 in
the basement of the Tower.
When interviewing for a work-study job, I got sent over to the computer center
and got my first look at the 6600. The next year, I took a COMPASS class, but
they were running SCOPE by that time.

[Rich Alderson]

Post by Dallas

Post by Rich Alderson
I knew that system! I was a freshman in the fall of 1969 at UT; I had a
work-study job with the School of Education's CAI Lab, working with an IBM 1800
in the basement of Sutton Hall and the administration's System/360 Model 50 in
the basement of the Tower.

You and I shared the same Freshman year at UT Austin.

Cool! Hook 'em, 'Horns!

Post by Dallas
What a huge enrollment! What a great football team.
Austin Texas and UT was such a fun place compared to my tiny rural high
school in East Texas.
I started out as a Chemistry Major, but took a test during admissions that
let me opt out of taking the usual Freshman Chemistry course. So, I took an
experimental course for freshmen by Dr F. A. Matsen named "The Vector Space
Theory of Matter", and they taught us FORTRAN.

I learned FORTRAN on an IBM 1401 the last semester of my senior year in a
Chicao suburb, while Dad finished his doctorate at Northwestern. That's why
the work-study folks were sending me to computer-related jobs! (I'm a native
Texan, born in Denison, but that's another story altogether.)

Post by Dallas
I was hooked immediately on writing software.
So I changed my major to Math/CS.

Perfectly reasonable. I was a linguistics major from the get-go.

Post by Dallas
I had a part time job coding for graduate students in the Physical Chemistry
Dept (mostly FORTRAN for doing matrix calculations for their Quantum
Mechanics models) or various things for the CS Dept through Dr. Brown all
four years I attended as an undergraduate.
I had no idea that having a CDC-6600 as your first computer was quite unusual.
- Larry

You've just missed having some fun. I work(ed) at Living Computers: Museum+Labs
in Seattle for the last 17 years. We restored a 6500 (originally at Purdue from
1967-1989, then in the Chippewa Falls Museum of Industry and Technology till
2011) and have given free accounts on it for several years. Unfortunately, the
museum is closing for 12-18 months (due to the effects of the pandemic on any
museum and more especially on "touch everything" museums).

My friend who got me the museum job was an operator at the UTCC, but only after
they had moved from the 6600 onto a Cyber. He's retired back to Austin, which
was home for him all along.

[Peter Flass]

Post by Rich Alderson

Post by Dallas

Post by Rich Alderson
I knew that system! I was a freshman in the fall of 1969 at UT; I had a
work-study job with the School of Education's CAI Lab, working with an IBM 1800
in the basement of Sutton Hall and the administration's System/360 Model 50 in
the basement of the Tower.

You and I shared the same Freshman year at UT Austin.

Cool! Hook 'em, 'Horns!

Post by Dallas
What a huge enrollment! What a great football team.
Austin Texas and UT was such a fun place compared to my tiny rural high
school in East Texas.
I started out as a Chemistry Major, but took a test during admissions that
let me opt out of taking the usual Freshman Chemistry course. So, I took an
experimental course for freshmen by Dr F. A. Matsen named "The Vector Space
Theory of Matter", and they taught us FORTRAN.

I learned FORTRAN on an IBM 1401 the last semester of my senior year in a
Chicao suburb, while Dad finished his doctorate at Northwestern. That's why
the work-study folks were sending me to computer-related jobs! (I'm a native
Texan, born in Denison, but that's another story altogether.)

Post by Dallas
I was hooked immediately on writing software.
So I changed my major to Math/CS.

Perfectly reasonable. I was a linguistics major from the get-go.

Post by Dallas
I had a part time job coding for graduate students in the Physical Chemistry
Dept (mostly FORTRAN for doing matrix calculations for their Quantum
Mechanics models) or various things for the CS Dept through Dr. Brown all
four years I attended as an undergraduate.
I had no idea that having a CDC-6600 as your first computer was quite unusual.
- Larry

You've just missed having some fun. I work(ed) at Living Computers: Museum+Labs
in Seattle for the last 17 years. We restored a 6500 (originally at Purdue from
1967-1989, then in the Chippewa Falls Museum of Industry and Technology till
2011) and have given free accounts on it for several years. Unfortunately, the
museum is closing for 12-18 months (due to the effects of the pandemic on any
museum and more especially on "touch everything" museums).

I saw that somewhere. I wish them, and you, the best of luck!

Post by Rich Alderson
My friend who got me the museum job was an operator at the UTCC, but only after
they had moved from the 6600 onto a Cyber. He's retired back to Austin, which
was home for him all along.

[Dallas]

Post by Rich Alderson
You've just missed having some fun. I work(ed) at Living Computers: Museum+Labs
in Seattle for the last 17 years. We restored a 6500 (originally at Purdue from
1967-1989, then in the Chippewa Falls Museum of Industry and Technology till
2011) and have given free accounts on it for several years. Unfortunately, the
museum is closing for 12-18 months (due to the effects of the pandemic on any
museum and more especially on "touch everything" museums).
My friend who got me the museum job was an operator at the UTCC, but only after
they had moved from the 6600 onto a Cyber. He's retired back to Austin, which
was home for him all along.

That's a great job, and I hope they re-open the museum just as soon as possible.

[John Levine]

Post by Thomas Koenig
I've been looking around a bit, but I cannot find which operating
system introduced named files.

Good question. CTSS had named files, I think in 1962.

Named files almost certainly predate disks. IBM operating systems had
(still have, I guess) an elaborate system for naming files on
magtapes. On OS you could put tape files in the catalog, just like
disk files, and if a job called for a file on a tape, it would tell
the operator to mount the tape.

Poking around at bitsavers, I see that the 7070 had tapes labelled
with filenames by 1960 and I presume it had named files when it had
disks or drums, too.

[Douglas Miller]

Can't say who invented it, but Honeywell had "tape monitors" that used named modules on tape, and the ability to search and load modules by name. I think their "card monitor" also did the same. It was probably something that everyone recognized as being advantageous/necessary.

[Robert Thau]

Post by John Levine
Good question. CTSS had named files, I think in 1962.
Named files almost certainly predate disks. IBM operating systems had
(still have, I guess) an elaborate system for naming files on
magtapes. On OS you could put tape files in the catalog, just like
disk files, and if a job called for a file on a tape, it would tell
the operator to mount the tape.

FLOW-MATIC from Univac (predecessor to COBOL) already had explicit
support for tape labels in 1957, in blocks at the beginning of the
tape drive which the run-time would check to insure that the right
tape was mounted -- and that tapes with portions of a file were
mounted in the right order if a file spanned more than one tape.

So, it's at least that old.

Robert Thau
rst@{ai,alum}.mit.edu

[Quadibloc]

Post by Thomas Koenig
OS/360 certainly had them, as did EXEC II. I would think that
peole started using them as soon as discs with random access made
their appearance - does anybody know more?

If so, the first disk with random access was the IBM RAMAC, so that should narrow
down your search.

John Savard

[Quadibloc]

Post by Quadibloc

Post by Thomas Koenig
OS/360 certainly had them, as did EXEC II. I would think that
peole started using them as soon as discs with random access made
their appearance - does anybody know more?

If so, the first disk with random access was the IBM RAMAC, so that should narrow
down your search.

I checked, and instead people using the RAMAC system had to divide the disk up
into ranges of track and sector numbers by hand.

Since IBM referred to a disk drive as a "disk file", what we call a file they
termed a "data set". Named data sets would certainly have been in use at the
point when disk drives were connected to computers running operating systems
that restricted access to particular data sets to the users who were their
owners. The 1301, an improved version of the RAMAC disk drive that had as many
read heads as surfaces, was used as a peripheral on the 7090, so it is indeed
quite possible IBM originated the named file condept, but I haven't been able to
verify this.

John Savard

[Quadibloc]

One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

John Savard

[Thomas Koenig]

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Which one?

[Quadibloc]

Post by Thomas Koenig

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Which one?

This article:
https://en.wikipedia.org/wiki/Computer_file

Yes, some of the other ones I saw had no information relevant to your query
although I might have expected them to.

John Savard

[Dallas]

Post by Quadibloc

Post by Thomas Koenig

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Which one?

https://en.wikipedia.org/wiki/Computer_file
Yes, some of the other ones I saw had no information relevant to your query
although I might have expected them to.
John Savard

I love that 1950's rhetoric quoted in that Wikipedia article, especially

"speeds intelligent solutions through mazes of mathematics"

- Larry

[Peter Flass]

Post by Dallas

Post by Quadibloc

Post by Thomas Koenig

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Which one?

https://en.wikipedia.org/wiki/Computer_file
Yes, some of the other ones I saw had no information relevant to your query
although I might have expected them to.
John Savard

I love that 1950's rhetoric quoted in that Wikipedia article, especially
"speeds intelligent solutions through mazes of mathematics"
- Larry

Actually , the article dates the “file system” to 1961, which is a slightly
different concept.

[Scott Lurndal]

Post by Thomas Koenig

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Which one?

Large systems (B{567}XXX) most likely. The B5000 preceeded the B3500.

[Peter Flass]

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.
John Savard

Wow, that recently? I assume they have sources, but it seems awfully late
in the game.

[Quadibloc]

Post by Peter Flass

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Wow, that recently? I assume they have sources, but it seems awfully late
in the game.

It does seem to me that named files on disk drives would still have been useful
even on purely batch systems. So I, too, wonder if there may not have been
something around earlier. However, IBM's RAMAC dates from 1956, so there isn't
_too_ much room for earlier.

However, what about named files on tape? While LINCtape, the ancestor of
DECtape, dates from 1962 or thereabouts, a perfectly ordinary computer system
might have one magnetic tape drive with a tape on it that contains, say, an
assembler, a COBOL compiler, and a FORTRAN compiler, and some way to advance the
conventional (7-track!) tape to the right one of those.

John Savard

[John Levine]

Post by Quadibloc
However, what about named files on tape?

IBM had named files on 1/2" magtape around 1960. See the note I sent a
day or two ago. The main point was to be sure that the tape in the
drive was the one the program expected.

I'm pretty sure that LINCtape was the first randomly accessible tape
system so it would have been the first filesystem on tape. DECtape was
probably the last.

[Quadibloc]

Post by John Levine
I'm pretty sure that LINCtape was the first randomly accessible tape
system so it would have been the first filesystem on tape. DECtape was
probably the last.

There was a DECtape II. It used the Philips Compact Cassette.

Also, I think there were filesystems on tape using those cartridges that HP
invented.

John Savard

[Bob Eager]

Post by Quadibloc

Post by John Levine
I'm pretty sure that LINCtape was the first randomly accessible tape
system so it would have been the first filesystem on tape. DECtape was
probably the last.

There was a DECtape II. It used the Philips Compact Cassette.

No, it didn't. It used a form factor that was smaller in area but
thicker, with a metal backplate. Similar to (but not the same as) the
later Travan and other drives.

There *was* also a Compact Cassette based system, but it wasn't branded
DECtape II. It was the TU60.

See: http://www.psych.usyd.edu.au/pdp-11/media.html

(I have a couple of DECtape II cartridges somewhere)

[Gareth Evans]

Post by Quadibloc

Post by John Levine
I'm pretty sure that LINCtape was the first randomly accessible tape
system so it would have been the first filesystem on tape. DECtape was
probably the last.

There was a DECtape II. It used the Philips Compact Cassette.

I'm fairly sure (senior moment?) That I had to write a device
driver for a home grown OS in about 1977 on a PDP11.

ISTR Skip To File Gap ?????

[Peter Flass]

Post by Quadibloc

Post by Peter Flass

Post by Quadibloc
One Wikipedia article dates this to 1961, crediting both the Burroughs MCP and
MIT's CTSS as coming up with it at around the same time.

Wow, that recently? I assume they have sources, but it seems awfully late
in the game.

It does seem to me that named files on disk drives would still have been useful
even on purely batch systems. So I, too, wonder if there may not have been
something around earlier. However, IBM's RAMAC dates from 1956, so there isn't
_too_ much room for earlier.
However, what about named files on tape? While LINCtape, the ancestor of
DECtape, dates from 1962 or thereabouts, a perfectly ordinary computer system
might have one magnetic tape drive with a tape on it that contains, say, an
assembler, a COBOL compiler, and a FORTRAN compiler, and some way to advance the
conventional (7-track!) tape to the right one of those.

Probably, originally the “some way” was by file number on the tape.

[John Levine]

Post by Quadibloc
DECtape, dates from 1962 or thereabouts, a perfectly ordinary computer system
might have one magnetic tape drive with a tape on it that contains, say, an
assembler, a COBOL compiler, and a FORTRAN compiler, and some way to advance the
conventional (7-track!) tape to the right one of those.

Probably, originally the “some way” was by file number on the tape.

Originally, sure, but by 1960-ish tapes had labels with file names.
(These are short files between the data files, not the paper labels.)
See the messages I posted in this thread a few days ago.

[Jon Elson]

Post by Thomas Koenig
I've been looking around a bit, but I cannot find which operating
system introduced named files.
OS/360 certainly had them, as did EXEC II. I would think that
peole started using them as soon as discs with random access made
their appearance - does anybody know more?

The LAP6 OS on the LINC computer had named files on LINCtape (similar to
DECtape) in 1965. There may have been earlier versions of this system
(file manager, editor, assembler and program loader all in one.)

Jon

[Anne & Lynn Wheeler]

Post by Thomas Koenig
I've been looking around a bit, but I cannot find which operating
system introduced named files.
OS/360 certainly had them, as did EXEC II. I would think that
peole started using them as soon as discs with random access made
their appearance - does anybody know more?

at least as soon as there was "system" storage devices where items for
different people & programs were stored ... as opposed to dedicated
storage devices (cards, tapes, etc) for specific program execution
(needed to select the data as opposed to select the device).

7094/CTSS
https://en.wikipedia.org/wiki/Compatible_Time-Sharing_System
The Compatible Time-Sharing System (CTSS) was one of the first
time-sharing operating systems; it was developed at the MIT Computation
Center. CTSS was first demonstrated on MIT's IBM 709 in November 1961;
service to MIT users began in the summer of 1963 and was operated until
1973.[1] During part of this time, MIT's influential Project MAC also
ran a CTSS service, but the system did not spread beyond these two
sites.

some of the 7094/CTSS people went to 5th to do multics
https://en.wikipedia.org/wiki/Multics
others went to IBM science center to do virtual machines, online
applications, performance, work profile and capacity planning, invented
GML in 1969 (morphs into international SGML after decade and after
another decade mophs into HTML at CERN).
https://en.wikipedia.org/wiki/History_of_CP/CMS


Experimental Time-Sharing System
https://en.wikipedia.org/wiki/Compatible_Time-Sharing_System#Experimental_Time-Sharing_System
John Backus said in the 1954 summer session at MIT that "By time
sharing, a big computer could be used as several small ones; there would
need to be a reading station for each user".[2] Computers at that time,
like IBM 704, were not powerful enough to implement such system, but at
the end of 1958, MIT's Computation Center nevertheless added a
typewriter input to its 704 with the intent that a programmer or
operator could "obtain additional answers from the machine on a
time-sharing basis with other programs using the machine
simultaneously".[3]

CTSS filesystem implementation (CP40/CMS & CP67/CMS inherited some
similarities)
https://en.wikipedia.org/wiki/Compatible_Time-Sharing_System#File_system