login/logon

James A. Landau JJJRLandau at AOL.COM
Wed May 16 13:12:32 UTC 2001


In a message dated 5/11/01 11:21:26 PM Eastern Daylight Time, pulliam at IIT.EDU
writes:

<< Has anyone else noticed that websites invite us to "login," while
 commercials invite us to "logon"? >>

In a message dated 5/14/01 9:01:00 AM Eastern Daylight Time,
JBaker at STRADLEY.COM writes:

<<         I understand a "logon" to be the process generally of accessing a
 service or web site, while a "login" is specifically the authentication of
 one's identity to a service or web site. >>

In my experience "login" and "login", like "eether" and "eyether", are
synonymous, although they do indicate where you come from.  "Login" vs.
"logon" tell which operating systems you are familiar with.

The OED2 gives the following dates:
- log in   1963 as a verb, 1965 as a noun
- log out 1963 as a noun, 1968 as a verb
- log on  1977 as a noun, 1983 as a verb
- log off   1983 as a verb, no noun citation

Merriam-Webster's 10th Collegiate gives 1962 for "log in" and 1977 for "log
on".
I have found a 1962 citation for both "login" and "logout" [1]

It appears that "login/logout" are the original forms and "logon/logoff"
developed later.

Why did "log on" appear years after "log in"?  I don't know, but the
following theory seems plausible:

While a program is sometimes said to be "in the computer" (or "in the
system"), the person running that program is "on the computer" (or "on the
system").  The preposition distinguishes the metaphorical location of the
person and the software he/she is running.

If I hear someone say "He's in the system" I would think of, say, "he" has
just filed a claim for unemployment and is waiting for his first welfare
check.

As for the date of the usage "on the computer", I remember the following
dialog from 1967:    "He's on the computer"
            "I hope not.  It will bend the panels" (the side panels of the
housing around the CPU)

If someone is "on" the computer, then he/she must have logged "on" rather
than logged "in", hence the phrase "log on" must have originated along with
its companion phrase "log off" rather than "log out".

Project MAC (developed circa 1960), considered to be the first time-sharing
system,
required the user to type in the commands "login" and "logout" [1].   Other
systems copied these commands, until eventually (1977?) somebody decided to
use "logon" and "logoff" instead.

In particular the UNIX operating system, which dates back to the early
1970's, used and still uses "login".  UNIX is the most common operating
system on Internet servers, and there are a good many professional
programmers developing Web sites whose sole experience with operating systems
has been with UNIX (and perhaps Windows and MS-DOS, which do not have
logins/logons).  Hence to many Web site developers, "login" is the correct
phrase to use.  The writers of commercials, however, being more attuned to
popular usage than to operating system idiosyncracies, would tend to use "log
on".

Do you find this theory plausible?

Next question: where did "log" come from?

I can see two possibilities.

1)  The computer has to keep a "log" of when each user signed on and off, so
that the user (or his/her department) can be charged for the time.  Reference
[1] specifically states with reference to "logout":
     "Should be given at end of each user's session.  Rewinds user's private
file tape; punches on-line time accounting cards."
{In this case the log consists of punched cards.)
Now the user may well be "on" the system, but he/she must be entered "in" (or
"into") the log.  Hence "log in" and the corresponding "log out".

2)  Circa 1960 (and today as well) there were many facilities which control
visitor access.  Among the controls are that the visitor must sign a register
upon entering and usually upon leaving.  This register is sometimes called a
"visitor's log" or other variation on "logbook".  Hence to enter the
facility, the visitor had to "log in" and "log out".  Hence a user to access
the Project MAC computers had to "log in" and "log out".

The trouble with 2) is that the OED2 does not recognize this usage of "log
in" and "log out".

                 - Jim Landau

Reference [1]:  "An Experimental Time-Sharing System" in _Proceedings of the
1962 Spring Joint Computer Conference_ published by the AFIPS (American
Federation of Information Processing Societies) and printed by Spartan Books
of Baltimore as vol 21, pp. 335-344.  The paper is reprinted in Saul Rosen,
ed _Programming Systems and Languages_ New York: McGraw-Hill Book Company,
1967, pages 683-698.  The authors of the paper are F. J. Corbato (there is an
acute accent on the second "o"), M. M Daggett, and R. C. Daley.
The applicable quote is on page 688 of Rosen:



More information about the Ads-l mailing list