From: KD

To: Editor

Date: 3 Aug 2004

Date's remarks in A Hypothesis [On Logical Differences 1-4] regarding the object world's failure to distinguish between values and variables should be historically framed.  The object mania of the 90s, to the present, tends to overreach by making applications programming a systems programming chore arising from misuse of exposed system programming capabilities.

Prior to the first object like HOL language, SIMULA, an operating systems or communications network software developer could gain a rudimentary sense of Date's storage model aspect of object oriented programming through the use of prototype data blocks; and rudimentary, function (method) dispatch and inheritance (by selective data structure copy and amendment), capabilities (e.g. Digital Equipment Corporation's PDP-10 TOPS10 or PDP-11 RSX-11M OSs).  These primitive but effective constructs were used to implement behavior and to represent state not only of I/O devices, but also of abstract entities such as computer network nodes and communications protocol virtual-circuit-end state.

But these should be regarded as only narrow capabilities w.r.t. where object capabilities would lead. To the Artificial Intelligence LISP community of the 1980s object capability became a powerful implementation means to realize declarative rule based systems (e.g. the pedagogic FRAMELANG system of MIT's Patrick Winston).

Unfortunately, due to stunted education, many applications programmers fail to see the top of the declarative programming layered systems implementation architecture.  The result is a muddling of object concepts and capabilities application.

C. J. Date Responds: I appreciate the comment and have nothing to add.

Posted: 11/05/04