Reply to WHAT I WANT (Bartels)

utzoo!decvax!ucbvax!mhtsa!harpo!whuxlb!ech utzoo!decvax!ucbvax!mhtsa!harpo!whuxlb!ech
Tue Jan 12 23:32:13 AEST 1982


- Give me a couple of days.  I am not sure I am comfortable being classified
	as an 'a', but I sure have no problems working with what you describe
	as 'b's: you see, I COLLABORATE with the Human Factors folks in trying
	to design, build, and document user-friendly operations systems for
	the phone company.  I will try to come up with a set of 'guidelines'
	for you in what kind of experiments are possible in assessing user
	interfaces.

- Please recognize that I am NOT an expert at measuring the effectiveness/
	user-friendliness/etc. of user interfaces; that's why I am asking
	for time.

- Please ask YOURSELF whether you really believe that there is anything
	better (i.e. more quantitative, more rational) than your 'gut
	feel' about an interface.  I am not trying to bait you here, that
	is a real question.  If user interfaces are, in your estimation,
	simply a matter of taste which defy objective measurement, then
	let's agree that this is a religious argument and end it right here.
	Otherwise, let's try to keep the flame burning lower on both sides
	of the discussion.

- Sorry for what was (apparently) excessive sarcasm in my previous epistle.
	To repeat my problem in less incendiary terms, I will again 
	stipulate that the UNIX interface is inhospitable to many people,
	including me at times.  The question is how to modify that inter-
	face, or better yet provide an alternative, which better suits
	another class of people.  That is not easy, and moreover it is
	likely to attract people who will criticize it without providing
	an alternative.  Again, I am not trying to bait you: just asking
	you to recognize that NO interface will serve EVERYBODY well, and
	those ill-served are likely to feel some hostility toward those
	well-served.  To that end, the first thing you have to do is
	DEFINE the class of people that you wish to serve better, and then
	propose some measures that allow you to measure your success at
	serving them.

- Finally, that whatever its faults, UNIX provides tremendous flexibility
	for tailoring an environment.  It may be that you have to be a
	type 'a' (whatever that is! context-dependent, you know!) to 
	utilize that flexibility, but please recognize that there are
	plenty of type 'a's around who will be pleased as punch to assist
	the effort to do better.  A couple of good suggestions have already
	appeared in this forum, and I would like to see (and will propose)
	more.  T

Keep the faith,
	=Ned Horvath=



More information about the Comp.unix.wizards mailing list