End of ADS-L Digest - 22 Feb 1998 to 23 Feb 1998
<br>************************************************
<br>
<br>======================================================================
<br>
<br>From: Automatic digest processor (2/23/98)
<br>To: Recipients of ADS-L digests
<br>
<br>         ADS-L Digest - 21 Feb 1998 to 22 Feb 1998             98-02-23 00:00:25
<br>There are 5 messages totalling 357 lines in this issue.
<br>
<br>Topics of the day:
<br>
<br>  1. "cookies"
<br>  2. hurache
<br>  3. Ice cream cone; Pop (soda)
<br>  4. Sky line; Ragged edge; Muckers; Psycho; Yellow dog
<br>  5. Windy City; Magic City
<br>
<br>----------------------------------------------------------------------
<br>
<br>Date:    Sat, 21 Feb 1998 23:09:36 -0800
<br>From:    Gabor Fencsik  gabor[AT SYMBOL GOES HERE]WELL.COM 
<br>Subject: Re: "cookies"
<br>
<br>Greg Downing forwards a question from the Linguist List:
<br>
<br>  can anyone tell me anything about the etymology resp. the sense
<br>  development of the computer term "cookie"? One definition I found in
<br>  "Among The New Words" in American Speech says:
<br>  "Cookies are bits of computer code that allow a Web page's operators
<br>  to collect information about each user for later reference." ,
<br>  but I still can't understand the motivation for calling this a
<br>  cookie. It must be a metaphorical meaning, but what kind of meaning
<br>  elements are being transferred here?
<br>
<br>The word "cookie" in computer jargon is shorthand for "magic cookie".
<br>The UNIX manual pages, published initially in the seventies, and gradually
<br>swelling to massive volumes through the eighties and nineties, were
<br>liberally sprinkled with "magic cookies" and "magic numbers".  I guess the
<br>terminology is rooted in the various computerized versions of dungeons-and-
<br>dragons type games, played by bleary-eyed programmers to fill up the time
<br>while their programs were being compiled, or to relieve the stress after
<br>a night of strenuous hacking.
<br>
<br>Technically speaking, the "magic cookies" and "magic numbers" were bit
<br>patterns found in a certain place within a file.  They were used to
<br>identify the type of the file in question.  The "magic cookie" was
<br>checked by computer programs to make sure the file was of the kind the
<br>program was allowed to operate on.  Thus, a tape archiver program
<br>was only allowed to touch archive files, a loader was only allowed to
<br>load executable files, and so forth.  The bit patterns constituting
<br>the "magic cookie" were chosen arbitrarily, but when they were absent,
<br>nothing worked.  That's why they were "magic".
<br>
<br>Thus, the original meaning of "magic cookie" involved information
<br>passing from a file to a program reading the file.  Later, the concept
<br>was extended to information being passed from a program to another program
<br>running on a different machine.  Programs send each other "magic cookies"
<br>primarily for purposes of authentication.  When you are browsing the
<br>New York Times web page, for example, the web server at the NY Times
<br>sends a cookie to you browser to identify you as a properly logged-on
<br>customer.  Without the cookie, they would have to have you log in every
<br>time you click on a link on their webpage, an annoyance up with which
<br>we would not put.
<br>
<br>I don't know when "magic cookie" began to be shortened to "cookie"
<br>tout simple, but I believe this is a relatively recent development.
<br>The question could be settled by searching the archives of the many
<br>UNIX-related discussion groups on Usenet.
<br>
<br>-----
<br>Gabor Fencsik
<br>