🚀 go-pugleaf

Jason M Baragry (baragry@latcs1.lat.OZ.AU) wrote:
: 

..


: 
: 
: 	Also, the notion of predefined and understood functionality of components
: seems to be an important factor in engineering design. Are there plans for 
: standardizing the functionality of software components rather than the interfaces
: to them? 
: 

Yes, notably the POSIX standards, which standardize both the interface
and software component functionality.  Unfortunately the word
"interface" has different meanings.  Most of the writing about open
system standards, including my own, says that the standards simply
define "an interface rather than an implementation".  But in this case,
"interface" means both the function call syntax and a complete
behavioral description.  POSIX standards are now required to define a
language independent specification (LIS), which defines standard
functionality.  The LIS is then mapped to a call syntax for particular
programming languages (actually just C in almost all cases).

Any component that provides the specified behavior through the
specified call syntax meets the requirements of the standard.  This use
of the term interface is contrasted with a standard "implementation",
which would mean standardizing on a particular collection of source or
executable programs.  It is unfortunate that the word "interface" has
become overloaded in this way.

--

Rick Kuhn                                   Telephone: +1 301 975 3337       
Natl Institute of Standards & Technology    Fax:       +1 301 590 0932      
Technology Bldg. B266                       Internet: kuhn@nist.gov
Gaithersburg, Md.  20899  USA                       DRKuhn@dockmaster.ncsc.mil