I believe that I am the one who started the practice of using the 1 CLEANSPACE date for carbon-dating APL. When I was publishing an internal newsletter within IBM (The APL Jot Dot Times), I was faced at one point with the question of how old APL is– when did it start? I’m sure we could never really give an absolute starting date, because a nearly-endless stream of ideas must have been considered before Iverson Notation came into being. The suggestion that we consider the 1962 date of the book publication to be a starting point is certainly valid. But what I was trying to do was to give a date to the starting point of the APL implementation on computers. Remember that APL is unique among computer languages in that it wasn’t originally written as a computer language– it was written as a means of letting two people converse unambiguously about math concepts. But then, if two people could converse, how about one person and a computer….

For discussions in the newsletter, and later for commemorating an APL at our internal (IBM APL ITL) conferences, it made sense to choose a date of when the first on-line implementation began. So, choosing the date that the first workspace was saved seemed like a reasonable approach. And since that workspace was (and is) still there, that seemed like the way to go. And even if, as Dick Lathwell pointed out, that workspace might have been recreated at one point, that still seemed like a reasonable approach.

Because the on-line implementation of APL as an executable computer langauge is what gave many of us our careers, that is the event that we chose to commemorate. So, many of us from IBM have spoken over the years of 1966-11-27 as being the starting point for APL’s on-line life. And all of the hard implementation work that the developers put into APL prior to that first saved workspace– well, that was pre-birth for APL.

But yes, if you are commemorating the origin of the notation itself, that would certainly be earlier.

]]>