Path: utzoo!utgpu!watmath!att!tut.cis.ohio-state.edu!cica!gatech!prism!loligo!pepke
From: pepke@loligo.cc.fsu.edu (Eric Pepke)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Subtantiatng my criticism (was: simple text interface)
Message-ID: <248@loligo.cc.fsu.edu>
Date: 8 Aug 89 22:00:01 GMT
References: <9674@phoenix.Princeton.EDU> <43528@bbn.COM> <14780@dartvax.Dartmouth.EDU> <183@dbase.UUCP> <14834@dartvax.Dartmouth.EDU> <7241@microsoft.UUCP>
Reply-To: pepke@loligo.UUCP (Eric Pepke)
Organization: Supercomputer Computations Research Institute
Lines: 43

Let's see if I shed a little darkness on this discussion:

There is NO reason to believe that graphical interfaces are inherently less
powerful, flexible, or synchronistic than command-line interfaces.  Paul
Haeberli* describes a visual system that is inherently considerably more
powerful than Unix.  Useful though redirection and pipes may be, the
Unix command line is still limited by the fact that it is basically a
one-dimensional construct, and each tool only has one input port.
Haeberli's _ConMan_ allows many and links them in a two-dimensional 
graphical interface.

With regard to the idea that nobody would ever program with a mouse, well,
there are several software project management systems that use dataflow
concepts and a visual metaphor.  Some of them are very good, far better,
in fact, than giving every member of the software development team a copy
of EMACS and turning them all loose.  These are no longer the days of the
lone hacker heroically kludging away; good software must be designed, and
some of the best tools are inherently graphical. 

Constructing a powerful visual metaphor is harder than throwing together a 
powerful but arcane command line interface (I hope this isn't news to anybody),
but this does not mean that it is impossible, just that more work needs
to be done.  Remember that in 1984 people were saying that nobody would
ever use a Macintosh for word processing because they would have to take
their hands off the keyboard to move the cursor.  Well, there were other
overriding advantages that the visionaries saw that we now think are old
hat, and it's not easy to remember the time when it was different.  

There couldn't possibly be any such advantages lurking waiting to be
discovered in graphical development systems, could there?  :-)

Reference:
*_ConMan: A Visual Programming Language for Interactive Graphics_
 Paul E. Haeberli
 Siggraph '88 conference proceedings.

Eric Pepke                                     INTERNET: pepke@gw.scri.fsu.edu
Supercomputer Computations Research Institute  MFENET:   pepke@fsu
Florida State University                       SPAN:     scri::pepke
Tallahassee, FL 32306-4052                     BITNET:   pepke@fsu

Disclaimer: My employers seldom even LISTEN to my opinions.
Meta-disclaimer: Any society that needs disclaimers has too many lawyers.